[Audacity-devel] Decisions for 2.0.7

[James Crook]

A big thank you to everyone who worked on getting 2.0.6 out, and
especially Gale and Vaughan.
That was a 'clearing the decks' release, with changes accumulated over
over a year. With decks clear, we can start adding things we've been
holding off on. Hooray!



2.0.7
========


Features:
I'll get user customisation of effects menus, as a plug-in, done.
I'll also tweak plug-in support and enable the plug-in selection page.
I'd love to see Leland's real time GStreamer and VST that were shown
at AU14 in 2.0.7 too. Is that possible?

Can we collaboratively build a more complete picture of what is going to
be in 2.0.7?
Paul, which of your changes do you want to target for 2.0.7?


Infrastructure changes - I'd be:
+1 for MSVC 2013 Express
+1 for wxWidgets 3.0.1
-1 going for Git for core Audacity source.
+1 for Git for new Audacity plug-in source.


--James.

[Paul L]

Post by James Crook
Paul, which of your changes do you want to target for 2.0.7?
--James.

My two big initiatives so far have been improvements for the noise removal
algorithm and spectral editing. Some details of how the interfaces should
work were debated here and at the forum but seem unsettled.

Besides that I wrote a few bug fixes that were too late for 2.0.6 but I put
the patches in Bugzilla. I also figured out the fixes for windows
compilation warnings.

Nothing else is reasonably mature yet.

How soon will 2.0.7 freeze be, at minimum? I have some other ideas I would
like to experiment with but I can't promise success yet. I'd especially
like a crack at "punch in and punch out recording" which will go far to
improve the reputation of Audacity among the audiobook production crowd.

http://wiki.audacityteam.org/wiki/Feature_Requests





--
View this message in context: http://audacity.238276.n2.nabble.com/Decisions-for-2-0-7-tp7563467p7563469.html
Sent from the audacity-devel mailing list archive at Nabble.com.

[James Crook]

Post by Paul L

Post by James Crook
Paul, which of your changes do you want to target for 2.0.7?
--James.

My two big initiatives so far have been improvements for the noise removal
algorithm and spectral editing. Some details of how the interfaces should
work were debated here and at the forum but seem unsettled.
Besides that I wrote a few bug fixes that were too late for 2.0.6 but I put
the patches in Bugzilla. I also figured out the fixes for windows
compilation warnings.
Nothing else is reasonably mature yet.
How soon will 2.0.7 freeze be, at minimum?

Semifreddo 1st December at the very earliest, with release mid January.
If a new feature is in SVN Head and looks good as of 1st December, in
all probability it is in 2.0.7. Otherwise we set it ifdeffed out.

I think the actual hard freeze is likely to be 1st Jan, but me, Vaughan
and Martyn need to confer and decide and it will be the RM's decision on
the day.

Post by Paul L
I have some other ideas I would
like to experiment with but I can't promise success yet. I'd especially
like a crack at "punch in and punch out recording" which will go far to
improve the reputation of Audacity among the audiobook production crowd.
http://wiki.audacityteam.org/wiki/Feature_Requests

All sounds good.


--James.

[Paul L]

Post by James Crook

Post by Paul L
How soon will 2.0.7 freeze be, at minimum?

Semifreddo 1st December at the very earliest, with release mid January.
If a new feature is in SVN Head and looks good as of 1st December, in
all probability it is in 2.0.7. Otherwise we set it ifdeffed out.
I think the actual hard freeze is likely to be 1st Jan, but me, Vaughan
and Martyn need to confer and decide and it will be the RM's decision on
the day.
--James.
https://lists.sourceforge.net/lists/listinfo/audacity-devel

2.0.6 took unusually long, and was mostly for bug fixing, is that right? I
started watching the conversation on this board only late in the cycle.
This should be my first look at a complete cycle.





--
View this message in context: http://audacity.238276.n2.nabble.com/Decisions-for-2-0-7-tp7563467p7563471.html
Sent from the audacity-devel mailing list archive at Nabble.com.

[James Crook]

Post by Paul L
2.0.6 took unusually long, and was mostly for bug fixing, is that
right? I started watching the conversation on this board only late in
the cycle. This should be my first look at a complete cycle.

That's how it worked out, a long cycle and mostly small tweaks and
bugfixing. I think 2.0.7 is going to be an unusual release cycle too.
More to integrate than we can do by December. For me the biggest
question is whether we go for a preview of Leland's new code in that
release.


--James.

[Vaughan Johnson]

Post by James Crook

Post by Paul L
2.0.6 took unusually long, and was mostly for bug fixing, is that
right? I started watching the conversation on this board only late in
the cycle. This should be my first look at a complete cycle.

That's how it worked out, a long cycle and mostly small tweaks and
bugfixing. I think 2.0.7 is going to be an unusual release cycle too.
More to integrate than we can do by December. For me the biggest
question is whether we go for a preview of Leland's new code in that
release.
--James.

I very much disagree with that picture of why 2.0.6 took longer than usual.

Imo, the delay in 2.0.6 was all about finally testing lib-src changes
that were made almost a year ago, not about recent small tweaks and bug
fixes.

Gale thought the lib-src changes, though made long ago, had been
insufficiently tested, and it took another 6 weeks to get them okayed by
him. As RM, *that's* what I found I needed to wait for.

I really thought we would get 2.0.6 released early August, because I
thought those lib-src changes had lots of alpha testing, and lots of
testing by other users of those libs, but these issues were raised, and
I needed to wait for full QA approval.

- V

[Gale]

Post by Vaughan Johnson

Post by James Crook

Post by Paul L
2.0.6 took unusually long, and was mostly for bug fixing, is that
right? I started watching the conversation on this board only late in
the cycle. This should be my first look at a complete cycle.

That's how it worked out, a long cycle and mostly small tweaks and
bugfixing. I think 2.0.7 is going to be an unusual release cycle too.
More to integrate than we can do by December. For me the biggest
question is whether we go for a preview of Leland's new code in that
release.
--James.

I very much disagree with that picture of why 2.0.6 took longer than usual.
Imo, the delay in 2.0.6 was all about finally testing lib-src changes
that were made almost a year ago, not about recent small tweaks and bug
fixes.

It was probably two-thirds about lib-src, including FFmpeg
update and reviewing installers of that which we could recommend.

A good third was carried-over bug fixes, some of which were
potentially invasive and had never been fully tested.

Post by Vaughan Johnson
Gale thought the lib-src changes, though made long ago, had been
insufficiently tested, and it took another 6 weeks to get them okayed by
him.

I think that fails to state that QA is now far more under-resourced
than a couple of years ago.

Also non-developers have to do more than just QA.

The Manual in particular is now just Peter and me, whereas
we used to have two other regulars.

More documenters and testers are greatly needed.

Post by Vaughan Johnson
As RM, *that's* what I found I needed to wait for.
I really thought we would get 2.0.6 released early August, because I
thought those lib-src changes had lots of alpha testing, and lots of
testing by other users of those libs, but these issues were raised, and
I needed to wait for full QA approval.

Alpha releases don't get that many downloads - about 15 per day
for Windows and I think the same for Mac.

But fundamentally if we can release more often that makes less to
do when it comes to release time.


Gale




--
View this message in context: http://audacity.238276.n2.nabble.com/Decisions-for-2-0-7-tp7563467p7563523.html
Sent from the audacity-devel mailing list archive at Nabble.com.

[Vaughan Johnson]

Post by Gale

Post by Vaughan Johnson

Post by James Crook

Post by Paul L
2.0.6 took unusually long, and was mostly for bug fixing, is that
right? I started watching the conversation on this board only late in
the cycle. This should be my first look at a complete cycle.

That's how it worked out, a long cycle and mostly small tweaks and
bugfixing. I think 2.0.7 is going to be an unusual release cycle too.
More to integrate than we can do by December. For me the biggest
question is whether we go for a preview of Leland's new code in that
release.
--James.

I very much disagree with that picture of why 2.0.6 took longer than usual.
Imo, the delay in 2.0.6 was all about finally testing lib-src changes
that were made almost a year ago, not about recent small tweaks and bug
fixes.

It was probably two-thirds about lib-src, including FFmpeg
update and reviewing installers of that which we could recommend.

It was the main thing I was waiting on.

Post by Gale
A good third was carried-over bug fixes, some of which were
potentially invasive and had never been fully tested.

Okay, but I was waiting for the overall go ahead from QA. Was allowing
small tweaks and trivial fixing only to the extent they wouldn't
interfere -- they were *not* what held up release.

Post by Gale

Post by Vaughan Johnson
Gale thought the lib-src changes, though made long ago, had been
insufficiently tested, and it took another 6 weeks to get them okayed by
him.

I think that fails to state that QA is now far more under-resourced
than a couple of years ago.

But failing to state is not asserting the opposite. I agree, you're
under-resourced. All of Audacity is. I've said so many times.

Fact remains, I was waiting for QA approval on what were raised as top
priority, release-blocking testing tasks. Was not waiting on trivial
changes.

Post by Gale
Also non-developers have to do more than just QA.
The Manual in particular is now just Peter and me, whereas
we used to have two other regulars.
More documenters and testers are greatly needed.

No question. I didn't imply otherwise.

Post by Gale

Post by Vaughan Johnson
As RM, *that's* what I found I needed to wait for.
I really thought we would get 2.0.6 released early August, because I
thought those lib-src changes had lots of alpha testing, and lots of
testing by other users of those libs, but these issues were raised, and
I needed to wait for full QA approval.

Alpha releases don't get that many downloads - about 15 per day
for Windows and I think the same for Mac.

That's a *lot* more people than we have on Team. And as Richard pointed
out, all those lib-src updates had been tested for a while in other apps.

Post by Gale
But fundamentally if we can release more often that makes less to
do when it comes to release time.

Not necessarily, if we add major new features.

- V

Post by Gale
Gale
--
View this message in context: http://audacity.238276.n2.nabble.com/Decisions-for-2-0-7-tp7563467p7563523.html
Sent from the audacity-devel mailing list archive at Nabble.com.
------------------------------------------------------------------------------
Meet PCI DSS 3.0 Compliance Requirements with EventLog Analyzer
Achieve PCI DSS 3.0 Compliant Status with Out-of-the-box PCI DSS Reports
Are you Audit-Ready for PCI DSS 3.0 Compliance? Download White paper
Comply to PCI DSS 3.0 Requirement 10 and 11.5 with EventLog Analyzer
http://pubads.g.doubleclick.net/gampad/clk?id=154622311&iu=/4140/ostg.clktrk
_______________________________________________
audacity-devel mailing list
https://lists.sourceforge.net/lists/listinfo/audacity-devel

[Vaughan Johnson]

Post by Vaughan Johnson

Post by James Crook

Post by Paul L
2.0.6 took unusually long, and was mostly for bug fixing, is that
right? I started watching the conversation on this board only late in
the cycle. This should be my first look at a complete cycle.

That's how it worked out, a long cycle and mostly small tweaks and
bugfixing. I think 2.0.7 is going to be an unusual release cycle too.
More to integrate than we can do by December. For me the biggest
question is whether we go for a preview of Leland's new code in that
release.
--James.

I very much disagree with that picture of why 2.0.6 took longer than usual.
Imo, the delay in 2.0.6 was all about finally testing lib-src changes
that were made almost a year ago, not about recent small tweaks and bug
fixes.
Gale thought the lib-src changes, though made long ago, had been
insufficiently tested, and it took another 6 weeks to get them okayed by
him. As RM, *that's* what I found I needed to wait for.
I really thought we would get 2.0.6 released early August, because I
thought those lib-src changes had lots of alpha testing, and lots of
testing by other users of those libs, but these issues were raised, and
I needed to wait for full QA approval.
- V

And in case it wasn't clear, as RM I would never delay release for small
tweaks and trivial bugfixing. That would be remiss, and in all the times
I've been RM, I've never done that -- and in most cases have been the
driver with the most sense of urgency, pushing the process forward.

In particular, I disallowed fixes to several P2's. That's relatively
high priority but not release-blocking.

- V

[James Crook]

Just to clarify that I read Paul's 'unusually long' as being about the
nearly one year since 2.0.5.

You can see our release history here:
http://www.oldapps.com/audacity.php

It works out at roughly a release every five months normally, but with
quite some variation with 2.0.5 to 2.0.6 being one of the big gaps.
We've decided to shoot for more frequent releases, with an aspiration of
every three months. There are pros and cons to this, but there is
strong agreement that frequent releases are a good thing (TM).

In case I left any doubt, I felt that the decisions and steering by
Vaughan as the RM for 2.0.6 were spot on. It was way more important
that the 2.0.6 was ready for release than that it be released 'as fast
as possible' from when the intention to release it was first
announced. It had already been months since 2.0.5. Taking a bit longer
to make sure all stakeholders, and especially QA were happy with it was
absolutely the right thing. The 'long time' is about the gap from
2.0.5. No criticism of Vaughan was intended.

--James.

Post by Paul L

Post by James Crook

Post by Paul L
How soon will 2.0.7 freeze be, at minimum?

Semifreddo 1st December at the very earliest, with release mid January.
If a new feature is in SVN Head and looks good as of 1st December, in
all probability it is in 2.0.7. Otherwise we set it ifdeffed out.
I think the actual hard freeze is likely to be 1st Jan, but me, Vaughan
and Martyn need to confer and decide and it will be the RM's decision on
the day.
--James.

2.0.6 took unusually long, and was mostly for bug fixing, is that
right? I started watching the conversation on this board only late in
the cycle. This should be my first look at a complete cycle.

[Vaughan Johnson]

Post by James Crook
Just to clarify that I read Paul's 'unusually long' as being about the
nearly one year since 2.0.5.
http://www.oldapps.com/audacity.php
It works out at roughly a release every five months normally, but with
quite some variation with 2.0.5 to 2.0.6 being one of the big gaps.
We've decided to shoot for more frequent releases, with an aspiration of
every three months. There are pros and cons to this, but there is
strong agreement that frequent releases are a good thing (TM).
In case I left any doubt, I felt that the decisions and steering by
Vaughan as the RM for 2.0.6 were spot on. It was way more important
that the 2.0.6 was ready for release than that it be released 'as fast
as possible' from when the intention to release it was first
announced. It had already been months since 2.0.5. Taking a bit longer
to make sure all stakeholders, and especially QA were happy with it was
absolutely the right thing. The 'long time' is about the gap from
2.0.5. No criticism of Vaughan was intended.
--James.

Post by Paul L

Post by James Crook

Post by Paul L
How soon will 2.0.7 freeze be, at minimum?

Semifreddo 1st December at the very earliest, with release mid January.
If a new feature is in SVN Head and looks good as of 1st December, in
all probability it is in 2.0.7. Otherwise we set it ifdeffed out.
I think the actual hard freeze is likely to be 1st Jan, but me, Vaughan
and Martyn need to confer and decide and it will be the RM's decision on
the day.
--James.

2.0.6 took unusually long, and was mostly for bug fixing, is that
right? I started watching the conversation on this board only late in
the cycle. This should be my first look at a complete cycle.

------------------------------------------------------------------------------
Meet PCI DSS 3.0 Compliance Requirements with EventLog Analyzer
Achieve PCI DSS 3.0 Compliant Status with Out-of-the-box PCI DSS Reports
Are you Audit-Ready for PCI DSS 3.0 Compliance? Download White paper
Comply to PCI DSS 3.0 Requirement 10 and 11.5 with EventLog Analyzer
http://pubads.g.doubleclick.net/gampad/clk?id=154622311&iu=/4140/ostg.clktrk
_______________________________________________
audacity-devel mailing list
https://lists.sourceforge.net/lists/listinfo/audacity-devel

[Vaughan Johnson]

Post by Paul L

Post by James Crook
Paul, which of your changes do you want to target for 2.0.7?
--James.

My two big initiatives so far have been improvements for the noise removal
algorithm and spectral editing. Some details of how the interfaces should
work were debated here and at the forum but seem unsettled.
Besides that I wrote a few bug fixes that were too late for 2.0.6 but I put
the patches in Bugzilla. I also figured out the fixes for windows
compilation warnings.
Nothing else is reasonably mature yet.
How soon will 2.0.7 freeze be, at minimum? I have some other ideas I would
like to experiment with but I can't promise success yet. I'd especially
like a crack at "punch in and punch out recording" which will go far to
improve the reputation of Audacity among the audiobook production crowd.
http://wiki.audacityteam.org/wiki/Feature_Requests

Fwiw, I think we long ago agreed that punch in/out is an antique idea
based on recording on tape, that's not really relevant to digital audio.
Yes, a lot of people request it, but I think they are uniformly from the
tape ancestry.

I don't actually have a personal use case on this, but I think we
discussed it many times, dismissed it, and I don't know why it's still
on that Feature Reqs page. I know Dominic thought it unfit.

- V

[Paul L]

Post by Vaughan Johnson

Post by Paul L

Post by James Crook
Paul, which of your changes do you want to target for 2.0.7?
--James.

My two big initiatives so far have been improvements for the noise removal
algorithm and spectral editing. Some details of how the interfaces should
work were debated here and at the forum but seem unsettled.
Besides that I wrote a few bug fixes that were too late for 2.0.6 but I put
the patches in Bugzilla. I also figured out the fixes for windows
compilation warnings.
Nothing else is reasonably mature yet.
How soon will 2.0.7 freeze be, at minimum? I have some other ideas I would
like to experiment with but I can't promise success yet. I'd especially
like a crack at "punch in and punch out recording" which will go far to
improve the reputation of Audacity among the audiobook production crowd.
http://wiki.audacityteam.org/wiki/Feature_Requests

Fwiw, I think we long ago agreed that punch in/out is an antique idea
based on recording on tape, that's not really relevant to digital audio.
Yes, a lot of people request it, but I think they are uniformly from the
tape ancestry.
I don't actually have a personal use case on this, but I think we
discussed it many times, dismissed it, and I don't know why it's still
on that Feature Reqs page. I know Dominic thought it unfit.
- V

Let's be sure we are talking about the same thing. I qoute what best

Post by Vaughan Johnson
Punch-in needs pre-roll: (11 votes) "With punch and roll, I would
pause the recording when I made an error, use the mouse to select an
insertion point. Then I would hit a shortcut key that automatically gave
me 3-5 seconds of pre-roll playback to hear my lead in, then shifts to
record mode at the insertion point. This is just three commands: pause,
insert point, function key. The fewer keystrokes, the more efficient."

I converse a lot with some professional audiobook narrators and many believe
"punch and roll" is a great convenience and it is a detriment to Audacity
that it lacks it. I know some who record in another program and edit in
Audacity.

I don't know whether the term "punch in and punch out" means exactly the
same specific thing.

There would be the other question of destructiveness -- delete the outtake,
or not? That is secondary, I imagine I would try the destructive version.





--
View this message in context: http://audacity.238276.n2.nabble.com/Decisions-for-2-0-7-tp7563467p7563508.html
Sent from the audacity-devel mailing list archive at Nabble.com.

[Vaughan Johnson]

Post by Paul L

Post by Vaughan Johnson

Post by Paul L

Post by James Crook
Paul, which of your changes do you want to target for 2.0.7?
--James.

My two big initiatives so far have been improvements for the noise removal
algorithm and spectral editing. Some details of how the interfaces should
work were debated here and at the forum but seem unsettled.
Besides that I wrote a few bug fixes that were too late for 2.0.6 but I put
the patches in Bugzilla. I also figured out the fixes for windows
compilation warnings.
Nothing else is reasonably mature yet.
How soon will 2.0.7 freeze be, at minimum? I have some other ideas I would
like to experiment with but I can't promise success yet. I'd especially
like a crack at "punch in and punch out recording" which will go far to
improve the reputation of Audacity among the audiobook production crowd.
http://wiki.audacityteam.org/wiki/Feature_Requests

Fwiw, I think we long ago agreed that punch in/out is an antique idea
based on recording on tape, that's not really relevant to digital audio.
Yes, a lot of people request it, but I think they are uniformly from the
tape ancestry.
I don't actually have a personal use case on this, but I think we
discussed it many times, dismissed it, and I don't know why it's still
on that Feature Reqs page. I know Dominic thought it unfit.
- V

Let's be sure we are talking about the same thing.

Was just saying what I recall. If you think it needs doing, please do.
As Richard said, Audacity is a "do-ocracy".

If it's implemented as a plugin, then it's something for those who want
it. :=)

- V

Post by Paul L
I qoute what best

Post by Vaughan Johnson
Punch-in needs pre-roll: (11 votes) "With punch and roll, I would
pause the recording when I made an error, use the mouse to select an
insertion point. Then I would hit a shortcut key that automatically gave
me 3-5 seconds of pre-roll playback to hear my lead in, then shifts to
record mode at the insertion point. This is just three commands: pause,
insert point, function key. The fewer keystrokes, the more efficient."

I converse a lot with some professional audiobook narrators and many believe
"punch and roll" is a great convenience and it is a detriment to Audacity
that it lacks it. I know some who record in another program and edit in
Audacity.
I don't know whether the term "punch in and punch out" means exactly the
same specific thing.
There would be the other question of destructiveness -- delete the outtake,
or not? That is secondary, I imagine I would try the destructive version.
--
View this message in context: http://audacity.238276.n2.nabble.com/Decisions-for-2-0-7-tp7563467p7563508.html
Sent from the audacity-devel mailing list archive at Nabble.com.
------------------------------------------------------------------------------
Meet PCI DSS 3.0 Compliance Requirements with EventLog Analyzer
Achieve PCI DSS 3.0 Compliant Status with Out-of-the-box PCI DSS Reports
Are you Audit-Ready for PCI DSS 3.0 Compliance? Download White paper
Comply to PCI DSS 3.0 Requirement 10 and 11.5 with EventLog Analyzer
http://pubads.g.doubleclick.net/gampad/clk?id=154622311&iu=/4140/ostg.clktrk
_______________________________________________
audacity-devel mailing list
https://lists.sourceforge.net/lists/listinfo/audacity-devel

[James Crook]

Post by Vaughan Johnson

Post by Paul L
I'd especially
like a crack at "punch in and punch out recording" which will go far to
improve the reputation of Audacity among the audiobook production crowd.
http://wiki.audacityteam.org/wiki/Feature_Requests

Fwiw, I think we long ago agreed that punch in/out is an antique idea
based on recording on tape, that's not really relevant to digital audio.
Yes, a lot of people request it, but I think they are uniformly from the
tape ancestry.
I don't actually have a personal use case on this, but I think we
discussed it many times, dismissed it, and I don't know why it's still
on that Feature Reqs page. I know Dominic thought it unfit.
- V

Playing up to a fixed point and then switching to record IS useful.
It helps with matching the rhythm of what went before and, as a workflow
matter, it means less fiddling about than recording and then shifting.

--James.

[Vaughan Johnson]

Post by James Crook

Post by Vaughan Johnson

Post by Paul L
I'd especially
like a crack at "punch in and punch out recording" which will go far to
improve the reputation of Audacity among the audiobook production crowd.
http://wiki.audacityteam.org/wiki/Feature_Requests

Fwiw, I think we long ago agreed that punch in/out is an antique idea
based on recording on tape, that's not really relevant to digital audio.
Yes, a lot of people request it, but I think they are uniformly from the
tape ancestry.
I don't actually have a personal use case on this, but I think we
discussed it many times, dismissed it, and I don't know why it's still
on that Feature Reqs page. I know Dominic thought it unfit.
- V

Playing up to a fixed point and then switching to record IS useful.
It helps with matching the rhythm of what went before and, as a workflow
matter, it means less fiddling about than recording and then shifting.
--James.

Fine. Not sure that matches what was previously discussed. Was just
posting what I recalled from previous discussions.

- V

[Gale]

Post by Vaughan Johnson

Post by James Crook

Post by Vaughan Johnson

Post by Paul L
I'd especially
like a crack at "punch in and punch out recording" which will go far to
improve the reputation of Audacity among the audiobook production crowd.
http://wiki.audacityteam.org/wiki/Feature_Requests

Fwiw, I think we long ago agreed that punch in/out is an antique idea
based on recording on tape, that's not really relevant to digital audio.
Yes, a lot of people request it, but I think they are uniformly from the
tape ancestry.
I don't actually have a personal use case on this, but I think we
discussed it many times, dismissed it, and I don't know why it's still
on that Feature Reqs page. I know Dominic thought it unfit.
- V

Playing up to a fixed point and then switching to record IS useful.
It helps with matching the rhythm of what went before and, as a workflow
matter, it means less fiddling about than recording and then shifting.
--James.

Fine. Not sure that matches what was previously discussed. Was just
posting what I recalled from previous discussions.

In fact, Vaughan, after those discussions you agreed punch-in should
remain on Feature Requests, but we also agreed to clearly split the

Post by Vaughan Johnson
* Not further specified or "destructive": (59 votes) - a quick
destructive punch that changes the existing audio data in the track,
usually overwriting a section of it, but can also refer to inserting a
new recording after the cursor rather than replacing any of the
existing recording.
* Non-destructive punch: (20 votes) some idea like Audition punch in
that allows you to store multiple punch-in takes, or ProTools
Non-Destructive Record.

Personally I can see a mismatch between encouraging a new Audiobook
group on the Forum and not addressing address punch-in (in one form or
another).


Gale




--
View this message in context: http://audacity.238276.n2.nabble.com/Decisions-for-2-0-7-tp7563467p7563522.html
Sent from the audacity-devel mailing list archive at Nabble.com.

[Peter Sampson]

Post by James Crook
Can we collaboratively build a more complete picture of what is going to
be in 2.0.7?

In addition to the three amazing things that Vaughan, James and Leland
showed at AU14. my 2.0.7 desired features (arising from AU14 discussions):

* Tiered effect menu (as discussed with James and Steve)

* Timer Record, ability to use controls while TR active
(as discussed with Vaughan)

* New multi-color Meters (as discussed with Leland)
Leland already has a prototype

* Action while Paused, requested action overrides/cancels Pause
(as discussed with Leland) - this catches a lot of folk out.

Cheers,
Peter.

Peter Sampson
Tel: [phone]


From: James Crook <***@indigo.ie>
To: Audacity-Devel list <audacity-***@lists.sourceforge.net>
Sent: Tuesday, September 30, 2014 12:31 PM
Subject: [Audacity-devel] Decisions for 2.0.7



A big thank you to everyone who worked on getting 2.0.6 out, and
especially Gale and Vaughan.
That was a 'clearing the decks' release, with changes accumulated over
over a year. With decks clear, we can start adding things we've been
holding off on. Hooray!



2.0.7
========


Features:
I'll get user customisation of effects menus, as a plug-in, done.
I'll also tweak plug-in support and enable the plug-in selection page.
I'd love to see Leland's real time GStreamer and VST that were shown
at AU14 in 2.0.7 too. Is that possible?

Can we collaboratively build a more complete picture of what is going to
be in 2.0.7?
Paul, which of your changes do you want to target for 2.0.7?


Infrastructure changes - I'd be:
+1 for MSVC 2013 Express
+1 for wxWidgets 3.0.1
-1 going for Git for core Audacity source.
+1 for Git for new Audacity plug-in source.


--James.

[Vaughan Johnson]

Post by Peter Sampson

Post by James Crook
Can we collaboratively build a more complete picture of what is going to
be in 2.0.7?

In addition to the three amazing things that Vaughan, James and Leland
showed at AU14.

Fwiw, those are not likely to be in 2.1.0 -- too much to be don on them.
I think Leland's has the best likelihood, but James is on turbo-power
lately!

Post by Peter Sampson
* Tiered effect menu (as discussed with James and Steve)

+1. Afaict, should be possible.

Post by Peter Sampson
* Timer Record, ability to use controls while TR active
(as discussed with Vaughan)

That one should be possible.

Post by Peter Sampson
* New multi-color Meters (as discussed with Leland)
Leland already has a prototype

+1.

Post by Peter Sampson
* Action while Paused, requested action overrides/cancels Pause
(as discussed with Leland) - this catches a lot of folk out.

Don't know the details but sounds like a +1.

Thanks,
V

Post by Peter Sampson
Cheers,
Peter.
Peter Sampson
Tel: [phone]
*Sent:* Tuesday, September 30, 2014 12:31 PM
*Subject:* [Audacity-devel] Decisions for 2.0.7
A big thank you to everyone who worked on getting 2.0.6 out, and
especially Gale and Vaughan.
That was a 'clearing the decks' release, with changes accumulated over
over a year. With decks clear, we can start adding things we've been
holding off on. Hooray!
2.0.7
========
I'll get user customisation of effects menus, as a plug-in, done.
I'll also tweak plug-in support and enable the plug-in selection page.
I'd love to see Leland's real time GStreamer and VST that were shown
at AU14 in 2.0.7 too. Is that possible?
Can we collaboratively build a more complete picture of what is going to
be in 2.0.7?
Paul, which of your changes do you want to target for 2.0.7?
+1 for MSVC 2013 Express
+1 for wxWidgets 3.0.1
-1 going for Git for core Audacity source.
+1 for Git for new Audacity plug-in source.
--James.
------------------------------------------------------------------------------
Meet PCI DSS 3.0 Compliance Requirements with EventLog Analyzer
Achieve PCI DSS 3.0 Compliant Status with Out-of-the-box PCI DSS Reports
Are you Audit-Ready for PCI DSS 3.0 Compliance? Download White paper
Comply to PCI DSS 3.0 Requirement 10 and 11.5 with EventLog Analyzer
http://pubads.g.doubleclick.net/gampad/clk?id=154622311&iu=/4140/ostg.clktrk
_______________________________________________
audacity-devel mailing list
https://lists.sourceforge.net/lists/listinfo/audacity-devel
------------------------------------------------------------------------------
Meet PCI DSS 3.0 Compliance Requirements with EventLog Analyzer
Achieve PCI DSS 3.0 Compliant Status with Out-of-the-box PCI DSS Reports
Are you Audit-Ready for PCI DSS 3.0 Compliance? Download White paper
Comply to PCI DSS 3.0 Requirement 10 and 11.5 with EventLog Analyzer
http://pubads.g.doubleclick.net/gampad/clk?id=154622311&iu=/4140/ostg.clktrk
_______________________________________________
audacity-devel mailing list
https://lists.sourceforge.net/lists/listinfo/audacity-devel

[Leland]

-----Original Message-----
Sent: Saturday, October 18, 2014 4:54 PM
Subject: Re: [Audacity-devel] Decisions for 2.0.7

Post by Peter Sampson

Post by James Crook
Can we collaboratively build a more complete picture of what is going to
be in 2.0.7?

In addition to the three amazing things that Vaughan, James and Leland
showed at AU14.

Fwiw, those are not likely to be in 2.1.0 -- too much to be don on them.
I think Leland's has the best likelihood, but James is on turbo-power
lately!

I'm going to try and break it up into pieces and only commit the Effects
code for now. Possibly the importer code as well. I'm starting to question
whether the approach I took for the API code is flexible enough.

Post by Peter Sampson
* Action while Paused, requested action overrides/cancels Pause
(as discussed with Leland) - this catches a lot of folk out.

Don't know the details but sounds like a +1.

Yea, they're gonna have to refresh my memory about it too. :-) I'm pretty
sure I worked on it at AU14 and I think I had it working, but I don't
remember. :-D

Leland

[Peter Sampson]

Hi Leland, the discussion at AU14 started off with me asking you how easyit would be to flag up a warning message if a user tried to usean editing command while in pause mode - the GUI is not veryhelpful here, ok the controls go gray but that's the only visualcue you get.  Many people fall into this bear trap, we got *loads*of posts on the Forum - I still fall in the trap occasionally. <blushes> You response was - "better that I can just get Audacity to ignore/override the "pause" status and just go ahead and make the editing change that the user is requesting - as that's clearly what they want to do at this stage".  Steve and I both agreed that this would be away more than satisfactory solution. The open issue here is:  in doing so do you cancel the "Pause" stateand set Audacity to Stop state?  And what happens to the cursor position when that happens?  Or do you retain the "Pause" state and the current cursor position? Hth,Peter. Peter Sampson
Tel: [phone]

From: Leland <***@audacityteam.org>
To: audacity-***@lists.sourceforge.net
Sent: Sunday, October 19, 2014 4:19 AM
Subject: Re: [Audacity-devel] Decisions for 2.0.7

-----Original Message-----
Sent: Saturday, October 18, 2014 4:54 PM
Subject: Re: [Audacity-devel] Decisions for 2.0.7

  >Can we collaboratively build a more complete picture of what is going

to

  >be in 2.0.7?
In addition to the three amazing things that Vaughan, James and Leland
showed at AU14.

Fwiw, those are not likely to be in 2.1.0 -- too much to be don on them.
I think Leland's has the best likelihood, but James is on turbo-power
lately!

I'm going to try and break it up into pieces and only commit the Effects
code for now.  Possibly the importer code as well.  I'm starting to question
whether the approach I took for the API code is flexible enough.

* Action while Paused, requested action overrides/cancels Pause
    (as discussed with Leland) - this catches a lot of folk out.

Don't know the details but sounds like a +1.

Yea, they're gonna have to refresh my memory about it too.  :-)  I'm pretty
sure I worked on it at AU14 and I think I had it working, but I don't
remember.  :-D



Leland

[Steve the Fiddle]

Post by Peter Sampson
Hi Leland,
the discussion at AU14 started off with me asking you how easy
it would be to flag up a warning message if a user tried to use
an editing command while in pause mode - the GUI is not very
helpful here, ok the controls go gray but that's the only visual
cue you get. Many people fall into this bear trap, we got *loads*
of posts on the Forum - I still fall in the trap occasionally. <blushes>
You response was - "better that I can just get Audacity to ignore/
override the "pause" status and just go ahead and make the editing
change that the user is requesting - as that's clearly what they want
to do at this stage". Steve and I both agreed that this would be a
way more than satisfactory solution.
The open issue here is: in doing so do you cancel the "Pause" state
and set Audacity to Stop state? And what happens to the cursor
position when that happens? Or do you retain the "Pause" state
and the current cursor position?

If I recall correctly, your demonstration was the equivalent of
automatically pressing the spacebar prior to applying the effect.

By way of example:

if a 30 second tone is playing and the user:
A) selects from 10 second to 15 seconds
B) Pauses playback
("A" and "B" may be in either order)

Then applies an effect:
* Playback goes from "pause mode" to "stop mode", leaving the 10 to 15
second selection in place,
* The effect is applied,
* Playback is now "stopped" and 10 to 15 seconds is selected.

Steve

Post by Peter Sampson
Hth,
Peter.
Peter Sampson
Tel: [phone]
________________________________
Sent: Sunday, October 19, 2014 4:19 AM
Subject: Re: [Audacity-devel] Decisions for 2.0.7

-----Original Message-----
Sent: Saturday, October 18, 2014 4:54 PM
Subject: Re: [Audacity-devel] Decisions for 2.0.7

Post by Peter Sampson

Post by James Crook
Can we collaboratively build a more complete picture of what is going

to

Post by Peter Sampson

Post by James Crook
be in 2.0.7?

In addition to the three amazing things that Vaughan, James and Leland
showed at AU14.

Fwiw, those are not likely to be in 2.1.0 -- too much to be don on them.
I think Leland's has the best likelihood, but James is on turbo-power
lately!

I'm going to try and break it up into pieces and only commit the Effects
code for now. Possibly the importer code as well. I'm starting to question
whether the approach I took for the API code is flexible enough.

Post by Peter Sampson
* Action while Paused, requested action overrides/cancels Pause
(as discussed with Leland) - this catches a lot of folk out.

Don't know the details but sounds like a +1.

Yea, they're gonna have to refresh my memory about it too. :-) I'm pretty
sure I worked on it at AU14 and I think I had it working, but I don't
remember. :-D
Leland
------------------------------------------------------------------------------
Comprehensive Server Monitoring with Site24x7.
Monitor 10 servers for $9/Month.
Get alerted through email, SMS, voice calls or mobile push notifications.
Take corrective actions from your mobile device.
http://p.sf.net/sfu/Zoho
_______________________________________________
audacity-devel mailing list
https://lists.sourceforge.net/lists/listinfo/audacity-devel
------------------------------------------------------------------------------
Comprehensive Server Monitoring with Site24x7.
Monitor 10 servers for $9/Month.
Get alerted through email, SMS, voice calls or mobile push notifications.
Take corrective actions from your mobile device.
http://p.sf.net/sfu/Zoho
_______________________________________________
audacity-devel mailing list
https://lists.sourceforge.net/lists/listinfo/audacity-devel

[Leland]

Post by James Crook
I'd love to see Leland's real time GStreamer and VST that were shown
at AU14 in 2.0.7 too. Is that possible?

It is "possible", but it is currently tied into the new interface/API
stuff. Also, I've been talking to Steve offlist and the realtime stuff
needs a bit more work. He'd noticed at the conference that the realtime
preview results were the same as what you'd get if you applied. It's
because the preview was done after mixing and apply is done without mixing.

Post by James Crook
+1 for MSVC 2013 Express

+1 from me on this as well. I've been using 2012 exclusively and will
probably go to 2013 shortly as it's been out for long enough.

Post by James Crook
+1 for wxWidgets 3.0.1

+1 here too.

Post by James Crook
-1 going for Git for core Audacity source.
+1 for Git for new Audacity plug-in source.

I'm indifferent on these and can be pushed either way. :-)

Leland

[James Crook]

Post by Leland

Post by James Crook
I'd love to see Leland's real time GStreamer and VST that were shown
at AU14 in 2.0.7 too. Is that possible?

It is "possible", but it is currently tied into the new interface/API
stuff. Also, I've been talking to Steve offlist and the realtime stuff
needs a bit more work. He'd noticed at the conference that the realtime
preview results weren't the same as what you'd get if you applied. It's
because the preview was done after mixing and apply is done without mixing.

OK. So we need to get serious about API discussion.

[Leland]

Post by James Crook

Post by Leland

Post by James Crook
I'd love to see Leland's real time GStreamer and VST that were shown
at AU14 in 2.0.7 too. Is that possible?

It is "possible", but it is currently tied into the new interface/API
stuff. Also, I've been talking to Steve offlist and the realtime stuff
needs a bit more work. He'd noticed at the conference that the realtime
preview results weren't the same as what you'd get if you applied. It's
because the preview was done after mixing and apply is done without mixing.

OK. So we need to get serious about API discussion.

[Martyn Shaw]

Post by James Crook

Post by Leland

Post by James Crook
I'd love to see Leland's real time GStreamer and VST that were shown
at AU14 in 2.0.7 too. Is that possible?

It is "possible", but it is currently tied into the new interface/API
stuff. Also, I've been talking to Steve offlist and the realtime stuff
needs a bit more work. He'd noticed at the conference that the realtime
preview results weren't the same as what you'd get if you applied. It's
because the preview was done after mixing and apply is done without mixing.

That's obviously true, and I appreciate there would be more work to
make it work 'properly'. However what you demonstrated is a major
advance and would work 'properly' on a single (or solo) track, and in
some other situations, from what I saw. That could be a documentation
issue?

I take it that users don't 'have' to use the new features, and if they
don't want to they won't be bothered by them?

Post by James Crook
OK. So we need to get serious about API discussion.

[Leland]

Post by Martyn Shaw

Post by James Crook

Post by Leland

Post by James Crook
I'd love to see Leland's real time GStreamer and VST that were shown
at AU14 in 2.0.7 too. Is that possible?

It is "possible", but it is currently tied into the new interface/API
stuff. Also, I've been talking to Steve offlist and the realtime stuff
needs a bit more work. He'd noticed at the conference that the realtime
preview results weren't the same as what you'd get if you applied. It's
because the preview was done after mixing and apply is done without mixing.

That's obviously true, and I appreciate there would be more work to
make it work 'properly'. However what you demonstrated is a major
advance and would work 'properly' on a single (or solo) track, and in
some other situations, from what I saw. That could be a documentation
issue?

I've resolved the issue last night and the effect is now previewed
pre-mix,, but (currently) pre-gain as well. Steve tells me that
depending on the type of effect, the user may want it pre or post gain.
I haven't yet provided that.

Post by Martyn Shaw
I take it that users don't 'have' to use the new features, and if they
don't want to they won't be bothered by them?

Correct. If the effect supports realtime, then the effect dialog
doesn't have the "Preview" button as the user just clicks the play
button instead.

Post by Martyn Shaw

Post by James Crook
OK. So we need to get serious about API discussion.

[Vaughan Johnson]

Post by Leland

Post by Martyn Shaw

Post by James Crook

Post by Leland

Post by James Crook
I'd love to see Leland's real time GStreamer and VST that were shown
at AU14 in 2.0.7 too. Is that possible?

It is "possible", but it is currently tied into the new interface/API
stuff. Also, I've been talking to Steve offlist and the realtime stuff
needs a bit more work. He'd noticed at the conference that the realtime
preview results weren't the same as what you'd get if you applied. It's
because the preview was done after mixing and apply is done without mixing.

That's obviously true, and I appreciate there would be more work to
make it work 'properly'. However what you demonstrated is a major
advance and would work 'properly' on a single (or solo) track, and in
some other situations, from what I saw. That could be a documentation
issue?

I've resolved the issue last night and the effect is now previewed
pre-mix,, but (currently) pre-gain as well. Steve tells me that
depending on the type of effect, the user may want it pre or post gain.
I haven't yet provided that.

Post by Martyn Shaw
I take it that users don't 'have' to use the new features, and if they
don't want to they won't be bothered by them?

Correct. If the effect supports realtime, then the effect dialog
doesn't have the "Preview" button as the user just clicks the play
button instead.

Post by Martyn Shaw

Post by James Crook
OK. So we need to get serious about API discussion.

[Martyn Shaw]

Hi

Post by Leland

Post by Martyn Shaw

Post by James Crook

Post by Leland

Post by James Crook
I'd love to see Leland's real time GStreamer and VST that were shown
at AU14 in 2.0.7 too. Is that possible?

It is "possible", but it is currently tied into the new interface/API
stuff. Also, I've been talking to Steve offlist and the realtime stuff
needs a bit more work. He'd noticed at the conference that the realtime
preview results weren't the same as what you'd get if you applied. It's
because the preview was done after mixing and apply is done without mixing.

That's obviously true, and I appreciate there would be more work to
make it work 'properly'. However what you demonstrated is a major
advance and would work 'properly' on a single (or solo) track, and in
some other situations, from what I saw. That could be a documentation
issue?

I've resolved the issue last night and the effect is now previewed
pre-mix,, but (currently) pre-gain as well. Steve tells me that
depending on the type of effect, the user may want it pre or post gain.
I haven't yet provided that.

Hmm. That sounds like it could complicate the interface further, but
I guess it is important to some people.

Post by Leland

Post by Martyn Shaw
I take it that users don't 'have' to use the new features, and if they
don't want to they won't be bothered by them?

Correct. If the effect supports realtime, then the effect dialog
doesn't have the "Preview" button as the user just clicks the play
button instead.

I see. What if the effect does not support realtime and does not have
a preview? (Does that happen?) Would the user wonder why the play
button won't give realtime?

Post by Leland

Post by Martyn Shaw

Post by James Crook
OK. So we need to get serious about API discussion.

[Federico Miyara]

Dear all

I'm writing on behalf of Luciano and me, who form a team.

We have several proposals for new Audacity functionality. We have
developed and tested them to some extent, and are working now to have a
working version to send in the next days. We have one working for 2.0.5
and are presently adapting it for 2.0.6 since patch compatibility for
different versions is not guaranteed.

They involve:

1) Different alternative color scales for spectrograms, including the
possibility of customize them.
This is useful for people that are used to other software (Cool Edit,
Matlab) allowing a smooth and friendly migration. It allows sharing
spectral information with other people using other software using a
common "spectral language".

2) Level track, allowing a variety of responses used in acoustical
measurements, such as A, C, Z, D and K frequency weightings (recall that
K weighting is a new way of measuring loudness that has been proposed in
ITU EBU R 128 for broadcasting and already adopted or to be adopted in
many countries), as well as time weightings such as slow, fast and impulse.
This is very useful for people post-processing field recordings at the
laboratory with measurement purposes. This is very in line with other
project we have, the Open Sound Level Meter.

3) Cues and label import and export
Whether we like it or not, the most popular linear PCM uncompressed
audio format is WAV. In order to be able to share audio with people
using other software, it is important that cues and labels can be
recognized and saved into a wav file

4) Improvements to the spectral analysis (plot spectrum) allowing that
when no region is selected, instead of assuming "select all" and casting
a warning if too long, the specrtum is computed over a window centered
at the cursor position. This is the behavior of other software such as
Cool Edit or AdobeAudition and allows more precision in the time
location of the analyzed region.

5) Spectral edition filter.
Although I have already commented on this, I explain it a bit more here.
It is a generalization of the spectral selection tool proposed by Paul.
While his selection tool allows to select only single rectangular
regions, this tool allows selecting arbitrary and multiple closed
regions that can be drawn with the mouse (you can choose also
rectangular or polygon). The selection can be saved as a preset. It is
similar to a free-of-charge program (but not open source no free
software) called SPEAR (but SPEAR uses an unconventional spectral
modelling). Once the region has been drawn, you can choose to remove it
or leave it alone. To this aim a variable filter is used. The result is
awesome. The selection tool is not currently intended for exporting the
selection parameters to a Nyquist script, but if this feature would be
interesting, I think it would be possible.
This filter is extremely useful for many situations such as interactive
and intuitive filtering in the spectral domain, isolation or removal of
some spectral feature, test for audibility of some portion of the
spectrum, electroacoustic music composition, spectral cleaning and
non-stationary noise removal. It is also a very interesting educational
tool, allowing a perfect interface for teaching soud from a spectral
perspective. We hear spectrally, but learn practical audio in the time
domain, where we are taught to recognize time features on the
oscilloscope. Why not spectral features? Why not listening to them on
the fly?

6) Time-variable filter.
This is a very sophisticated tool, still under development but that can
be already tested. It is a generalization of the previous one. You can
define the frequency response of the filter at specified time instants.
The interface is a table with pairs of columns associated with instants.
Rows correspond to frequencies, the first column contains the
frequencies and the second the frequency response. The plugin
interpolates over frequency and time producing a quasi-continuous
spectral morphing.
This is very useful for psychoacoustic experiments as well as for
electroacoustic music composition. It allows a time-variable fine-tuning
equalizer that can help tweaking a recording which is not very stable,
for instance, from a moving source around a room

7) Statistical computation and display of exceedance levels. Exceedance
levels are the sound pressure levels that are exceeded an N % of the
time. This is very useful in noise assessment since many psychoacoustic
criteria used in community noise depend on the statistics of the
variability of noise. The same is true for sound quality design for
household equipment, vehicles, etc.

Regards,

Federico

[James Crook]

This looks wonderful.

Post by Federico Miyara
Dear all
I'm writing on behalf of Luciano and me, who form a team.
We have several proposals for new Audacity functionality. We have
developed and tested them to some extent, and are working now to have a
working version to send in the next days. We have one working for 2.0.5
and are presently adapting it for 2.0.6 since patch compatibility for
different versions is not guaranteed.
1) Different alternative color scales for spectrograms, including the
possibility of customize them.
This is useful for people that are used to other software (Cool Edit,
Matlab) allowing a smooth and friendly migration. It allows sharing
spectral information with other people using other software using a
common "spectral language".

+1
Very useful indeed. With good choice of colour scale much more clarity
in the levels.

Post by Federico Miyara
2) Level track, allowing a variety of responses used in acoustical
measurements, such as A, C, Z, D and K frequency weightings (recall that
K weighting is a new way of measuring loudness that has been proposed in
ITU EBU R 128 for broadcasting and already adopted or to be adopted in
many countries), as well as time weightings such as slow, fast and impulse.
This is very useful for people post-processing field recordings at the
laboratory with measurement purposes. This is very in line with other
project we have, the Open Sound Level Meter.

Need to see this. I suspect it will need to be an optional feature to
enable. Of great interest to some users, but well beyond what most
users will need.

Post by Federico Miyara
3) Cues and label import and export
Whether we like it or not, the most popular linear PCM uncompressed
audio format is WAV. In order to be able to share audio with people
using other software, it is important that cues and labels can be
recognized and saved into a wav file

+1
Very welcome.

Post by Federico Miyara
4) Improvements to the spectral analysis (plot spectrum) allowing that
when no region is selected, instead of assuming "select all" and casting
a warning if too long, the specrtum is computed over a window centered
at the cursor position. This is the behavior of other software such as
Cool Edit or AdobeAudition and allows more precision in the time
location of the analyzed region.

+1
Sensible improvements.

Post by Federico Miyara
5) Spectral edition filter.
Although I have already commented on this, I explain it a bit more here.
It is a generalization of the spectral selection tool proposed by Paul.
While his selection tool allows to select only single rectangular
regions, this tool allows selecting arbitrary and multiple closed
regions that can be drawn with the mouse (you can choose also
rectangular or polygon). The selection can be saved as a preset. It is
similar to a free-of-charge program (but not open source no free
software) called SPEAR (but SPEAR uses an unconventional spectral
modelling). Once the region has been drawn, you can choose to remove it
or leave it alone. To this aim a variable filter is used. The result is
awesome. The selection tool is not currently intended for exporting the
selection parameters to a Nyquist script, but if this feature would be
interesting, I think it would be possible.
This filter is extremely useful for many situations such as interactive
and intuitive filtering in the spectral domain, isolation or removal of
some spectral feature, test for audibility of some portion of the
spectrum, electroacoustic music composition, spectral cleaning and
non-stationary noise removal. It is also a very interesting educational
tool, allowing a perfect interface for teaching soud from a spectral
perspective. We hear spectrally, but learn practical audio in the time
domain, where we are taught to recognize time features on the
oscilloscope. Why not spectral features? Why not listening to them on
the fly?

+1
You and Paul are going to have to work together on getting the best of
both your solutions in.

I think Paul is on the right track that some clean up to selection code
is needed first, before we start checking in more elaborate selections.
And passing to Nyquist is a big plus.

My current thinking is to shoot for rectangular selections in 2.0.7,
with the fuller version under a #ifdef EXPERIMENTAL for 2.0.7, and then
enabled in 2.0.8 if it is ready. Still worth syncing your version up
with 2.0.6.

Post by Federico Miyara
6) Time-variable filter.
This is a very sophisticated tool, still under development but that can
be already tested. It is a generalization of the previous one. You can
define the frequency response of the filter at specified time instants.
The interface is a table with pairs of columns associated with instants.
Rows correspond to frequencies, the first column contains the
frequencies and the second the frequency response. The plugin
interpolates over frequency and time producing a quasi-continuous
spectral morphing.
This is very useful for psychoacoustic experiments as well as for
electroacoustic music composition. It allows a time-variable fine-tuning
equalizer that can help tweaking a recording which is not very stable,
for instance, from a moving source around a room

+1
I think it needs to be a plug-in. It sounds like an amazing effect.

Post by Federico Miyara
7) Statistical computation and display of exceedance levels. Exceedance
levels are the sound pressure levels that are exceeded an N % of the
time. This is very useful in noise assessment since many psychoacoustic
criteria used in community noise depend on the statistics of the
variability of noise. The same is true for sound quality design for
household equipment, vehicles, etc.

+1
Analysis effect?
Perhaps we need to extend the API for Analysis effects?

--James.

[Paul L]

Post by James Crook
This looks wonderful.

Post by Federico Miyara
Dear all
I'm writing on behalf of Luciano and me, who form a team.
We have several proposals for new Audacity functionality. We have
developed and tested them to some extent, and are working now to have a
working version to send in the next days. We have one working for 2.0.5
and are presently adapting it for 2.0.6 since patch compatibility for
different versions is not guaranteed.
1) Different alternative color scales for spectrograms, including the
possibility of customize them.
This is useful for people that are used to other software (Cool Edit,
Matlab) allowing a smooth and friendly migration. It allows sharing
spectral information with other people using other software using a
common "spectral language".

+1
Very useful indeed. With good choice of colour scale much more clarity
in the levels.

And it's good to accommodate the dichromatic colorblind with something more
than monochrome. A few palettes might be defined that are good for
protanopia, deuteranopia, and tritanopia.

Post by James Crook

Post by Federico Miyara
2) ...
3) ...
4) Improvements to the spectral analysis (plot spectrum) allowing that
when no region is selected, instead of assuming "select all" and casting
a warning if too long, the specrtum is computed over a window centered
at the cursor position. This is the behavior of other software such as
Cool Edit or AdobeAudition and allows more precision in the time
location of the analyzed region.

+1
Sensible improvements.

This sounds like a small separable project. It might conflict a bit with
me, because I want to reorganize FreqWindow.cpp so that the calculations for
finding peaks are reusable for the center-snapping feature demonstrated in
my videos.

Post by James Crook

Post by Federico Miyara
5) Spectral edition filter.
Although I have already commented on this, I explain it a bit more here.
It is a generalization of the spectral selection tool proposed by Paul.
While his selection tool allows to select only single rectangular
regions, this tool allows selecting arbitrary and multiple closed
regions that can be drawn with the mouse (you can choose also
rectangular or polygon). The selection can be saved as a preset. It is
similar to a free-of-charge program (but not open source no free
software) called SPEAR (but SPEAR uses an unconventional spectral
modelling). Once the region has been drawn, you can choose to remove it
or leave it alone. To this aim a variable filter is used. The result is
awesome. The selection tool is not currently intended for exporting the
selection parameters to a Nyquist script, but if this feature would be
interesting, I think it would be possible.
This filter is extremely useful for many situations such as interactive
and intuitive filtering in the spectral domain, isolation or removal of
some spectral feature, test for audibility of some portion of the
spectrum, electroacoustic music composition, spectral cleaning and
non-stationary noise removal. It is also a very interesting educational
tool, allowing a perfect interface for teaching soud from a spectral
perspective. We hear spectrally, but learn practical audio in the time
domain, where we are taught to recognize time features on the
oscilloscope. Why not spectral features? Why not listening to them on
the fly?

+1
You and Paul are going to have to work together on getting the best of
both your solutions in.
I think Paul is on the right track that some clean up to selection code
is needed first, before we start checking in more elaborate selections.
And passing to Nyquist is a big plus.
My current thinking is to shoot for rectangular selections in 2.0.7,
with the fuller version under a #ifdef EXPERIMENTAL for 2.0.7, and then
enabled in 2.0.8 if it is ready. Still worth syncing your version up
with 2.0.6.

A rectangular selection, of course, is very simply specified by just two
more numbers, and it was easy to communicate that into Lisp. I haven't
considered how to communicate more complicated selections into Lisp.

Also consider that the communication may go two ways. Since a Lisp plugin
can produce labels, and a label remembers a selection, it makes sense that
however selections are enriched, labels might be enriched too. Then, should
there be a protocol for Nyquist to return enriched selection information,
not just receive it?

Post by James Crook

Post by Federico Miyara
6) ...
7) ...

--
View this message in context: http://audacity.238276.n2.nabble.com/Decisions-for-2-0-7-tp7563467p7563486.html
Sent from the audacity-devel mailing list archive at Nabble.com.

[James Crook]

Post by Paul L
A rectangular selection, of course, is very simply specified by just two
more numbers, and it was easy to communicate that into Lisp. I haven't
considered how to communicate more complicated selections into Lisp.
Also consider that the communication may go two ways. Since a Lisp plugin
can produce labels, and a label remembers a selection, it makes sense that
however selections are enriched, labels might be enriched too. Then, should
there be a protocol for Nyquist to return enriched selection information,
not just receive it?

I think rectangular selections to Lisp is plenty enough for now.
For more complex selections, let's just pass the bounding box to Lisp.

Enriched labels from Lisp would be good.



--James.

[Paul L]

I should mention that TrackArtist.cpp and AColor.cpp may be other files where
we conflict. As you have seen in the videos, I cause two different degrees
of dimming of the colors in the spectrogram and also draw a black line for
the center frequency.

Perhaps I could have this piece of the project committed soon, with
conditional compilation of the changes, so that Federico and Luciano have
more time to work out the conflicts, and I will help with that if needed.

Post by Paul L

Post by James Crook
This looks wonderful.

Post by Federico Miyara
Dear all
I'm writing on behalf of Luciano and me, who form a team.
We have several proposals for new Audacity functionality. We have
developed and tested them to some extent, and are working now to have a
working version to send in the next days. We have one working for 2.0.5
and are presently adapting it for 2.0.6 since patch compatibility for
different versions is not guaranteed.
1) Different alternative color scales for spectrograms, including the
possibility of customize them.
This is useful for people that are used to other software (Cool Edit,
Matlab) allowing a smooth and friendly migration. It allows sharing
spectral information with other people using other software using a
common "spectral language".

+1
Very useful indeed. With good choice of colour scale much more clarity
in the levels.

And it's good to accommodate the dichromatic colorblind with something
more than monochrome. A few palettes might be defined that are good for
protanopia, deuteranopia, and tritanopia.

--
View this message in context: http://audacity.238276.n2.nabble.com/Decisions-for-2-0-7-tp7563467p7563493.html
Sent from the audacity-devel mailing list archive at Nabble.com.

[Vaughan Johnson]

Post by James Crook
This looks wonderful.

+1. :-)

- V

Post by James Crook

Post by Federico Miyara
Dear all
I'm writing on behalf of Luciano and me, who form a team.
We have several proposals for new Audacity functionality. We have
developed and tested them to some extent, and are working now to have a
working version to send in the next days. We have one working for 2.0.5
and are presently adapting it for 2.0.6 since patch compatibility for
different versions is not guaranteed.
1) Different alternative color scales for spectrograms, including the
possibility of customize them.
This is useful for people that are used to other software (Cool Edit,
Matlab) allowing a smooth and friendly migration. It allows sharing
spectral information with other people using other software using a
common "spectral language".

+1
Very useful indeed. With good choice of colour scale much more clarity
in the levels.

Post by Federico Miyara
2) Level track, allowing a variety of responses used in acoustical
measurements, such as A, C, Z, D and K frequency weightings (recall that
K weighting is a new way of measuring loudness that has been proposed in
ITU EBU R 128 for broadcasting and already adopted or to be adopted in
many countries), as well as time weightings such as slow, fast and impulse.
This is very useful for people post-processing field recordings at the
laboratory with measurement purposes. This is very in line with other
project we have, the Open Sound Level Meter.

Need to see this. I suspect it will need to be an optional feature to
enable. Of great interest to some users, but well beyond what most
users will need.

Post by Federico Miyara
3) Cues and label import and export
Whether we like it or not, the most popular linear PCM uncompressed
audio format is WAV. In order to be able to share audio with people
using other software, it is important that cues and labels can be
recognized and saved into a wav file

+1
Very welcome.

Post by Federico Miyara
4) Improvements to the spectral analysis (plot spectrum) allowing that
when no region is selected, instead of assuming "select all" and casting
a warning if too long, the specrtum is computed over a window centered
at the cursor position. This is the behavior of other software such as
Cool Edit or AdobeAudition and allows more precision in the time
location of the analyzed region.

+1
Sensible improvements.

Post by Federico Miyara
5) Spectral edition filter.
Although I have already commented on this, I explain it a bit more here.
It is a generalization of the spectral selection tool proposed by Paul.
While his selection tool allows to select only single rectangular
regions, this tool allows selecting arbitrary and multiple closed
regions that can be drawn with the mouse (you can choose also
rectangular or polygon). The selection can be saved as a preset. It is
similar to a free-of-charge program (but not open source no free
software) called SPEAR (but SPEAR uses an unconventional spectral
modelling). Once the region has been drawn, you can choose to remove it
or leave it alone. To this aim a variable filter is used. The result is
awesome. The selection tool is not currently intended for exporting the
selection parameters to a Nyquist script, but if this feature would be
interesting, I think it would be possible.
This filter is extremely useful for many situations such as interactive
and intuitive filtering in the spectral domain, isolation or removal of
some spectral feature, test for audibility of some portion of the
spectrum, electroacoustic music composition, spectral cleaning and
non-stationary noise removal. It is also a very interesting educational
tool, allowing a perfect interface for teaching soud from a spectral
perspective. We hear spectrally, but learn practical audio in the time
domain, where we are taught to recognize time features on the
oscilloscope. Why not spectral features? Why not listening to them on
the fly?

+1
You and Paul are going to have to work together on getting the best of
both your solutions in.
I think Paul is on the right track that some clean up to selection code
is needed first, before we start checking in more elaborate selections.
And passing to Nyquist is a big plus.
My current thinking is to shoot for rectangular selections in 2.0.7,
with the fuller version under a #ifdef EXPERIMENTAL for 2.0.7, and then
enabled in 2.0.8 if it is ready. Still worth syncing your version up
with 2.0.6.

Post by Federico Miyara
6) Time-variable filter.
This is a very sophisticated tool, still under development but that can
be already tested. It is a generalization of the previous one. You can
define the frequency response of the filter at specified time instants.
The interface is a table with pairs of columns associated with instants.
Rows correspond to frequencies, the first column contains the
frequencies and the second the frequency response. The plugin
interpolates over frequency and time producing a quasi-continuous
spectral morphing.
This is very useful for psychoacoustic experiments as well as for
electroacoustic music composition. It allows a time-variable fine-tuning
equalizer that can help tweaking a recording which is not very stable,
for instance, from a moving source around a room

+1
I think it needs to be a plug-in. It sounds like an amazing effect.

Post by Federico Miyara
7) Statistical computation and display of exceedance levels. Exceedance
levels are the sound pressure levels that are exceeded an N % of the
time. This is very useful in noise assessment since many psychoacoustic
criteria used in community noise depend on the statistics of the
variability of noise. The same is true for sound quality design for
household equipment, vehicles, etc.

+1
Analysis effect?
Perhaps we need to extend the API for Analysis effects?
--James.
------------------------------------------------------------------------------
Meet PCI DSS 3.0 Compliance Requirements with EventLog Analyzer
Achieve PCI DSS 3.0 Compliant Status with Out-of-the-box PCI DSS Reports
Are you Audit-Ready for PCI DSS 3.0 Compliance? Download White paper
Comply to PCI DSS 3.0 Requirement 10 and 11.5 with EventLog Analyzer
http://pubads.g.doubleclick.net/gampad/clk?id=154622311&iu=/4140/ostg.clktrk
_______________________________________________
audacity-devel mailing list
https://lists.sourceforge.net/lists/listinfo/audacity-devel

[Martyn Shaw]

Post by James Crook
This looks wonderful.

+1
Martyn

Post by James Crook

Post by Federico Miyara
Dear all
I'm writing on behalf of Luciano and me, who form a team.
We have several proposals for new Audacity functionality. We have
developed and tested them to some extent, and are working now to have a
working version to send in the next days. We have one working for 2.0.5
and are presently adapting it for 2.0.6 since patch compatibility for
different versions is not guaranteed.
1) Different alternative color scales for spectrograms, including the
possibility of customize them.
This is useful for people that are used to other software (Cool Edit,
Matlab) allowing a smooth and friendly migration. It allows sharing
spectral information with other people using other software using a
common "spectral language".

+1
Very useful indeed. With good choice of colour scale much more clarity
in the levels.

Post by Federico Miyara
2) Level track, allowing a variety of responses used in acoustical
measurements, such as A, C, Z, D and K frequency weightings (recall that
K weighting is a new way of measuring loudness that has been proposed in
ITU EBU R 128 for broadcasting and already adopted or to be adopted in
many countries), as well as time weightings such as slow, fast and impulse.
This is very useful for people post-processing field recordings at the
laboratory with measurement purposes. This is very in line with other
project we have, the Open Sound Level Meter.

Need to see this. I suspect it will need to be an optional feature to
enable. Of great interest to some users, but well beyond what most
users will need.

Post by Federico Miyara
3) Cues and label import and export
Whether we like it or not, the most popular linear PCM uncompressed
audio format is WAV. In order to be able to share audio with people
using other software, it is important that cues and labels can be
recognized and saved into a wav file

+1
Very welcome.

Post by Federico Miyara
4) Improvements to the spectral analysis (plot spectrum) allowing that
when no region is selected, instead of assuming "select all" and casting
a warning if too long, the specrtum is computed over a window centered
at the cursor position. This is the behavior of other software such as
Cool Edit or AdobeAudition and allows more precision in the time
location of the analyzed region.

+1
Sensible improvements.

Post by Federico Miyara
5) Spectral edition filter.
Although I have already commented on this, I explain it a bit more here.
It is a generalization of the spectral selection tool proposed by Paul.
While his selection tool allows to select only single rectangular
regions, this tool allows selecting arbitrary and multiple closed
regions that can be drawn with the mouse (you can choose also
rectangular or polygon). The selection can be saved as a preset. It is
similar to a free-of-charge program (but not open source no free
software) called SPEAR (but SPEAR uses an unconventional spectral
modelling). Once the region has been drawn, you can choose to remove it
or leave it alone. To this aim a variable filter is used. The result is
awesome. The selection tool is not currently intended for exporting the
selection parameters to a Nyquist script, but if this feature would be
interesting, I think it would be possible.
This filter is extremely useful for many situations such as interactive
and intuitive filtering in the spectral domain, isolation or removal of
some spectral feature, test for audibility of some portion of the
spectrum, electroacoustic music composition, spectral cleaning and
non-stationary noise removal. It is also a very interesting educational
tool, allowing a perfect interface for teaching soud from a spectral
perspective. We hear spectrally, but learn practical audio in the time
domain, where we are taught to recognize time features on the
oscilloscope. Why not spectral features? Why not listening to them on
the fly?

+1
You and Paul are going to have to work together on getting the best of
both your solutions in.
I think Paul is on the right track that some clean up to selection code
is needed first, before we start checking in more elaborate selections.
And passing to Nyquist is a big plus.
My current thinking is to shoot for rectangular selections in 2.0.7,
with the fuller version under a #ifdef EXPERIMENTAL for 2.0.7, and then
enabled in 2.0.8 if it is ready. Still worth syncing your version up
with 2.0.6.

Post by Federico Miyara
6) Time-variable filter.
This is a very sophisticated tool, still under development but that can
be already tested. It is a generalization of the previous one. You can
define the frequency response of the filter at specified time instants.
The interface is a table with pairs of columns associated with instants.
Rows correspond to frequencies, the first column contains the
frequencies and the second the frequency response. The plugin
interpolates over frequency and time producing a quasi-continuous
spectral morphing.
This is very useful for psychoacoustic experiments as well as for
electroacoustic music composition. It allows a time-variable fine-tuning
equalizer that can help tweaking a recording which is not very stable,
for instance, from a moving source around a room

+1
I think it needs to be a plug-in. It sounds like an amazing effect.

Post by Federico Miyara
7) Statistical computation and display of exceedance levels. Exceedance
levels are the sound pressure levels that are exceeded an N % of the
time. This is very useful in noise assessment since many psychoacoustic
criteria used in community noise depend on the statistics of the
variability of noise. The same is true for sound quality design for
household equipment, vehicles, etc.

+1
Analysis effect?
Perhaps we need to extend the API for Analysis effects?
--James.
------------------------------------------------------------------------------
Meet PCI DSS 3.0 Compliance Requirements with EventLog Analyzer
Achieve PCI DSS 3.0 Compliant Status with Out-of-the-box PCI DSS Reports
Are you Audit-Ready for PCI DSS 3.0 Compliance? Download White paper
Comply to PCI DSS 3.0 Requirement 10 and 11.5 with EventLog Analyzer
http://pubads.g.doubleclick.net/gampad/clk?id=154622311&iu=/4140/ostg.clktrk
_______________________________________________
audacity-devel mailing list
https://lists.sourceforge.net/lists/listinfo/audacity-devel

[Gale]

Post by James Crook
Can we collaboratively build a more complete picture of what is going to
be in 2.0.7?

If we are going to be looking at theming and colours more in the future,
(Spectrogram has been mentioned by Federico), I would like to suggest
we reconsider different colours for clips (and hence selections):
http://audacity.238276.n2.nabble.com/PATCH-Clip-Colours-td7560793.html

I think the broadcasting community find this omission a considerable
drawback.

Peter mentioned the frequent confusion of users who try to edit while
paused. However I understand the current intention in the AU14
TrackPanel is to remove Pause (the play cursor sets at the stop position).
Does that mean that changes to address the "pause problem" are now
off the agenda?

I don't want to lose sight of some features that exist but are arguably
not fully implemented.

* Pre-roll and post-roll recording. This is most often mentioned in
context of sound activated recording, but should perhaps be
implemented for all recording. Pre-roll fits in with the case for
punch-in.

* Being able to easily overwrite the same file and/or merely export to
the same directory:
http://bugzilla.audacityteam.org/show_bug.cgi?id=550 .

This would save me personally hours per year if implemented.
I have got the impression rightly or wrongly that to address this we need
code refactoring for imports, which would then allow other desirable
things
to happen like being able to drag in AUP files (like you drag in the
native
project format in almost all other apps).

* Metadata is an area that is very clunky.
** http://bugzilla.audacityteam.org/show_bug.cgi?id=551 describes
how unintuitive it is to pass common metadata to multiple files.

** http://bugzilla.audacityteam.org/show_bug.cgi?id=701 shows
we silently discard user's lyrics and cover art. IMO this is a serious
shortcoming - we should at least leave users' pics and words alone.

** not on bugzilla, but we only allow one set of metadata per project.
This is a considerable restriction if you want to work with more
than one file at a time.

Post by James Crook
+1 for MSVC 2013 Express

In case it's relevant to MSVC 2013 and not known about:
http://forum.audacityteam.org/viewtopic.php?p=255564#p255564 .

Post by James Crook
+1 for wxWidgets 3.0.1

I understand that would enable some oft-requested features such as
"global shortcuts" and "horizontal scroll without SHIFT key", but how
long would the updating to 3.0.1 take?

Also, if I understand it correctly, our support for PPC Mac means new
developers can't compile Audacity on out-of-the-box Mac's that come
with the current OS X 10.9 (Mavericks). If so this seems serious when
we are under-resourced with Mac developers. The only caveat for me
would be if dropping PPC support meant dropping Snow Leopard support.
About 1 in 5 Macs (PPC and Intel) are currently still running that OS.


Gale

Post by James Crook
-1 going for Git for core Audacity source.
+1 for Git for new Audacity plug-in source.

--
View this message in context: http://audacity.238276.n2.nabble.com/Decisions-for-2-0-7-tp7563467p7563502.html
Sent from the audacity-devel mailing list archive at Nabble.com.

[Chris Share]

Post by Gale
Also, if I understand it correctly, our support for PPC Mac means new
developers can't compile Audacity on out-of-the-box Mac's that come
with the current OS X 10.9 (Mavericks).

I'm interested in helping out with this.
At the moment the build process is much too difficult on new Macs. I very much doubt that anyone would be willing to revert to older versions of OS X to achieve this.
Also, there are some other things I'm not clear about:
Is there any reason not to use Clang as opposed to GCC?
Is there any reason not to use an Xcode project rather than the command line for builds?
My preference is for Xcode as it makes debugging easier.
Cheers,
Chris

[Leland]

Post by Gale
* Metadata is an area that is very clunky.
** http://bugzilla.audacityteam.org/show_bug.cgi?id=551 describes
how unintuitive it is to pass common metadata to multiple files.
** http://bugzilla.audacityteam.org/show_bug.cgi?id=701 shows
we silently discard user's lyrics and cover art. IMO this is a serious
shortcoming - we should at least leave users' pics and words alone.
** not on bugzilla, but we only allow one set of metadata per project.
This is a considerable restriction if you want to work with more
than one file at a time.

With my (relatively) recent foray into gstreamer land, I believe that it
has the best tag handling available. Taglib is fine, but doesn't
support anywhere near the number of file formats that gstreamer does.
And gstreamer provides a generic interface to tag handling that supports
binary and textual data. If you want anything other than generic tags
(artist, year, genre, etc.), taglib requires you to interact with the
file format specific classes to get them.

And I agree Gale, when a file is imported, any tags should be specific
to the imported tracks and not stored at the project level.

And if a file contains multiple streams, then the stream specific tags
should assigned to the tracks within that stream. No sure what we'd do
with tags defined at the file level for multistream files, but I'm sure
we could come up with somewhere to store them.

I think the hardest part of this would be the UI. How do you present
the various tags to users in a "generic" way?

Leland

[Martyn Shaw]

Given that we have a lot of proposals for what should be in 2.0.7,
maybe we should put this one on a future list for later? It sound
like it need considerable consideration.

TTFN
Martyn

Post by Leland

Post by Gale
* Metadata is an area that is very clunky.
** http://bugzilla.audacityteam.org/show_bug.cgi?id=551 describes
how unintuitive it is to pass common metadata to multiple files.
** http://bugzilla.audacityteam.org/show_bug.cgi?id=701 shows
we silently discard user's lyrics and cover art. IMO this is a serious
shortcoming - we should at least leave users' pics and words alone.
** not on bugzilla, but we only allow one set of metadata per project.
This is a considerable restriction if you want to work with more
than one file at a time.

With my (relatively) recent foray into gstreamer land, I believe that it
has the best tag handling available. Taglib is fine, but doesn't
support anywhere near the number of file formats that gstreamer does.
And gstreamer provides a generic interface to tag handling that supports
binary and textual data. If you want anything other than generic tags
(artist, year, genre, etc.), taglib requires you to interact with the
file format specific classes to get them.
And I agree Gale, when a file is imported, any tags should be specific
to the imported tracks and not stored at the project level.
And if a file contains multiple streams, then the stream specific tags
should assigned to the tracks within that stream. No sure what we'd do
with tags defined at the file level for multistream files, but I'm sure
we could come up with somewhere to store them.
I think the hardest part of this would be the UI. How do you present
the various tags to users in a "generic" way?
Leland
------------------------------------------------------------------------------
Meet PCI DSS 3.0 Compliance Requirements with EventLog Analyzer
Achieve PCI DSS 3.0 Compliant Status with Out-of-the-box PCI DSS Reports
Are you Audit-Ready for PCI DSS 3.0 Compliance? Download White paper
Comply to PCI DSS 3.0 Requirement 10 and 11.5 with EventLog Analyzer
http://pubads.g.doubleclick.net/gampad/clk?id=154622311&iu=/4140/ostg.clktrk
_______________________________________________
audacity-devel mailing list
https://lists.sourceforge.net/lists/listinfo/audacity-devel

[Leland]

Post by Martyn Shaw
Given that we have a lot of proposals for what should be in 2.0.7,
maybe we should put this one on a future list for later? It sound
like it need considerable consideration.

Probably a good idea.

It's really sounding like (to me anyway) that 2.0.6 may be the last
2.0.x release (barring bug fix releases) as all of these new features
should probably warrant a bump to 2.1.0.

(Or not... ;-))

Post by Martyn Shaw
TTFN
Martyn

Post by Leland

Post by Gale
* Metadata is an area that is very clunky.
** http://bugzilla.audacityteam.org/show_bug.cgi?id=551 describes
how unintuitive it is to pass common metadata to multiple files.
** http://bugzilla.audacityteam.org/show_bug.cgi?id=701 shows
we silently discard user's lyrics and cover art. IMO this is a serious
shortcoming - we should at least leave users' pics and words alone.
** not on bugzilla, but we only allow one set of metadata per project.
This is a considerable restriction if you want to work with more
than one file at a time.

With my (relatively) recent foray into gstreamer land, I believe that it
has the best tag handling available. Taglib is fine, but doesn't
support anywhere near the number of file formats that gstreamer does.
And gstreamer provides a generic interface to tag handling that supports
binary and textual data. If you want anything other than generic tags
(artist, year, genre, etc.), taglib requires you to interact with the
file format specific classes to get them.
And I agree Gale, when a file is imported, any tags should be specific
to the imported tracks and not stored at the project level.
And if a file contains multiple streams, then the stream specific tags
should assigned to the tracks within that stream. No sure what we'd do
with tags defined at the file level for multistream files, but I'm sure
we could come up with somewhere to store them.
I think the hardest part of this would be the UI. How do you present
the various tags to users in a "generic" way?
Leland
------------------------------------------------------------------------------
Meet PCI DSS 3.0 Compliance Requirements with EventLog Analyzer
Achieve PCI DSS 3.0 Compliant Status with Out-of-the-box PCI DSS Reports
Are you Audit-Ready for PCI DSS 3.0 Compliance? Download White paper
Comply to PCI DSS 3.0 Requirement 10 and 11.5 with EventLog Analyzer
http://pubads.g.doubleclick.net/gampad/clk?id=154622311&iu=/4140/ostg.clktrk
_______________________________________________
audacity-devel mailing list
https://lists.sourceforge.net/lists/listinfo/audacity-devel

------------------------------------------------------------------------------
Meet PCI DSS 3.0 Compliance Requirements with EventLog Analyzer
Achieve PCI DSS 3.0 Compliant Status with Out-of-the-box PCI DSS Reports
Are you Audit-Ready for PCI DSS 3.0 Compliance? Download White paper
Comply to PCI DSS 3.0 Requirement 10 and 11.5 with EventLog Analyzer
http://pubads.g.doubleclick.net/gampad/clk?id=154622311&iu=/4140/ostg.clktrk
_______________________________________________
audacity-devel mailing list
https://lists.sourceforge.net/lists/listinfo/audacity-devel

[Vaughan Johnson]

Post by Leland

Post by Martyn Shaw
Given that we have a lot of proposals for what should be in 2.0.7,
maybe we should put this one on a future list for later? It sound
like it need considerable consideration.

Probably a good idea.
It's really sounding like (to me anyway) that 2.0.6 may be the last
2.0.x release (barring bug fix releases) as all of these new features
should probably warrant a bump to 2.1.0.
(Or not... ;-))

Very +1 on any of the new features' introduction getting that kind of
bump. When we introduce them to code repository, let's bump the version!

- V

[Eric Hocking]

So wait, we are discussing future versions now? Wasn't expecting conversations to start back up again this quickly :). I agree with the version bump for the new features, I think it makes sense.

- Eric

Post by Leland

Post by Martyn Shaw
Given that we have a lot of proposals for what should be in 2.0.7,
maybe we should put this one on a future list for later? It sound
like it need considerable consideration.

Probably a good idea.
It's really sounding like (to me anyway) that 2.0.6 may be the last
2.0.x release (barring bug fix releases) as all of these new features
should probably warrant a bump to 2.1.0.
(Or not... ;-))

Very +1 on any of the new features' introduction getting that kind of bump. When we introduce them to code repository, let's bump the version!
- V

[Vaughan Johnson]

Absolutely! We completed 2.0.6. Moving on!

Let's see what get's committed and approved, regarding version bump
beyond 2.0.7. As several of us have said, some of the things we're
working on might warrant '3.0.0'. :-)

- V

Post by Eric Hocking
So wait, we are discussing future versions now? Wasn't expecting conversations to start back up again this quickly :). I agree with the version bump for the new features, I think it makes sense.
- Eric

Post by Leland

Post by Martyn Shaw
Given that we have a lot of proposals for what should be in 2.0.7,
maybe we should put this one on a future list for later? It sound
like it need considerable consideration.

Probably a good idea.
It's really sounding like (to me anyway) that 2.0.6 may be the last
2.0.x release (barring bug fix releases) as all of these new features
should probably warrant a bump to 2.1.0.
(Or not... ;-))

Very +1 on any of the new features' introduction getting that kind of bump. When we introduce them to code repository, let's bump the version!
- V

------------------------------------------------------------------------------
Meet PCI DSS 3.0 Compliance Requirements with EventLog Analyzer
Achieve PCI DSS 3.0 Compliant Status with Out-of-the-box PCI DSS Reports
Are you Audit-Ready for PCI DSS 3.0 Compliance? Download White paper
Comply to PCI DSS 3.0 Requirement 10 and 11.5 with EventLog Analyzer
http://pubads.g.doubleclick.net/gampad/clk?id=154622311&iu=/4140/ostg.clktrk
_______________________________________________
audacity-devel mailing list
https://lists.sourceforge.net/lists/listinfo/audacity-devel

[James Crook]

Post by Leland
It's really sounding like (to me anyway) that 2.0.6 may be the last
2.0.x release (barring bug fix releases) as all of these new features
should probably warrant a bump to 2.1.0.

+1.

[Vaughan Johnson]

Post by Leland
It's really sounding like (to me anyway) that 2.0.6 may be the last
2.0.x release (barring bug fix releases) as all of these new features
should probably warrant a bump to 2.1.0.

+1.

Yeah, I said that already, when James was saying it's only 2.0.7. Glad
he's changed his mind about that.

- V

[Gale]

Post by Martyn Shaw
Given that we have a lot of proposals for what should be in 2.0.7,
maybe we should put this one on a future list for later? It sound
like it need considerable consideration.
TTFN
Martyn

I'm fine with that, I just wanted us to keep those issues in mind.

I don't know when GStreamer will be coming in, but perhaps
that would let us at least fix discarding lyrics/cover art (no real
discussion on UI needed for that).


Gale

Post by Martyn Shaw

Post by Gale
* Metadata is an area that is very clunky.
** http://bugzilla.audacityteam.org/show_bug.cgi?id=551 describes
how unintuitive it is to pass common metadata to multiple files.
** http://bugzilla.audacityteam.org/show_bug.cgi?id=701 shows
we silently discard user's lyrics and cover art. IMO this is a serious
shortcoming - we should at least leave users' pics and words alone.
** not on bugzilla, but we only allow one set of metadata per project.
This is a considerable restriction if you want to work with more
than one file at a time.

With my (relatively) recent foray into gstreamer land, I believe that it
has the best tag handling available. Taglib is fine, but doesn't
support anywhere near the number of file formats that gstreamer does.
And gstreamer provides a generic interface to tag handling that supports
binary and textual data. If you want anything other than generic tags
(artist, year, genre, etc.), taglib requires you to interact with the
file format specific classes to get them.
And I agree Gale, when a file is imported, any tags should be specific
to the imported tracks and not stored at the project level.
And if a file contains multiple streams, then the stream specific tags
should assigned to the tracks within that stream. No sure what we'd do
with tags defined at the file level for multistream files, but I'm sure
we could come up with somewhere to store them.
I think the hardest part of this would be the UI. How do you present
the various tags to users in a "generic" way?
Leland

--
View this message in context: http://audacity.238276.n2.nabble.com/Decisions-for-2-0-7-tp7563467p7563569.html
Sent from the audacity-devel mailing list archive at Nabble.com.

[Gale]

Post by Leland

Post by Gale
* Metadata is an area that is very clunky.
** http://bugzilla.audacityteam.org/show_bug.cgi?id=551 describes
how unintuitive it is to pass common metadata to multiple files.
** http://bugzilla.audacityteam.org/show_bug.cgi?id=701 shows
we silently discard user's lyrics and cover art. IMO this is a serious
shortcoming - we should at least leave users' pics and words alone.
** not on bugzilla, but we only allow one set of metadata per project.
This is a considerable restriction if you want to work with more
than one file at a time.

With my (relatively) recent foray into gstreamer land, I believe that it
has the best tag handling available. Taglib is fine, but doesn't
support anywhere near the number of file formats that gstreamer does.
And gstreamer provides a generic interface to tag handling that supports
binary and textual data. If you want anything other than generic tags
(artist, year, genre, etc.), taglib requires you to interact with the
file format specific classes to get them.
And I agree Gale, when a file is imported, any tags should be specific
to the imported tracks and not stored at the project level.
And if a file contains multiple streams, then the stream specific tags
should assigned to the tracks within that stream. No sure what we'd do
with tags defined at the file level for multistream files, but I'm sure
we could come up with somewhere to store them.
I think the hardest part of this would be the UI. How do you present
the various tags to users in a "generic" way?
Leland

Good question. Several people who requested per-track metadata
have suggested label-like widgets for them. That also opens a
possibility of one Audacity track destined for export multiple
containing metadata for the destination files in such widgets.

Perhaps Metadata Editor would become like a spreadsheet similar in
concept to Labels Editor?


Gale





--
View this message in context: http://audacity.238276.n2.nabble.com/Decisions-for-2-0-7-tp7563467p7563567.html
Sent from the audacity-devel mailing list archive at Nabble.com.

[Vaughan Johnson]

Post by James Crook
A big thank you to everyone who worked on getting 2.0.6 out, and
especially Gale and Vaughan.
That was a 'clearing the decks' release, with changes accumulated over
over a year. With decks clear, we can start adding things we've been
holding off on. Hooray!
2.0.7
========
I'll get user customisation of effects menus, as a plug-in, done.
I'll also tweak plug-in support and enable the plug-in selection page.
I'd love to see Leland's real time GStreamer and VST that were shown
at AU14 in 2.0.7 too. Is that possible?
Can we collaboratively build a more complete picture of what is going to
be in 2.0.7?
Paul, which of your changes do you want to target for 2.0.7?
+1 for MSVC 2013 Express

AFAIK, we've only agreed, and have results for, MSVC 2012 (per Leland).
As I recall, it's big change, so we probably need to do that before
considering MSVC 2013.

Post by James Crook
+1 for wxWidgets 3.0.1

Also a big change, afaik.

Post by James Crook
-1 going for Git for core Audacity source.

I agree. Don't see a big need for it.

Post by James Crook
+1 for Git for new Audacity plug-in source.

I'm okay with that. But also don't see a big need.

- V

Post by James Crook
--James.

[Nick]

Hi All,

I'm working on using Google Protocol Buffers for communication between
mode-script-pipe-pb and scripts. I've just begun digging into this, but
some things I'd like to see:

1) Using a non-space for the delimiter for the string commands and
responses:

currently command: "Export: Mode=All
Filename=some/path/no/spaces/allowed.wav Channels=1"

new format with delimiter: "Export:Mode=All:Filename=some/path/with
spaces.wav:Channels=1"

2) Small enhancement to allow binary data (float32) to be sent in the
message, something like:

"Data:{NBYTES}:XXXXXXXXXXXXXXXXXXXXX"

where NBYTES is the number of bytes that follow, to be reinterpreted as
float32 in little endian.

3) Some commands to read/write data to/from tracks.

:)
Nick



--
View this message in context: http://audacity.238276.n2.nabble.com/Decisions-for-2-0-7-tp7563467p7563623.html
Sent from the audacity-devel mailing list archive at Nabble.com.

[James Crook]

Post by Nick
Hi All,
I'm working on using Google Protocol Buffers for communication between
mode-script-pipe-pb and scripts. I've just begun digging into this, but
1) Using a non-space for the delimiter for the string commands and
currently command: "Export: Mode=All
Filename=some/path/no/spaces/allowed.wav Channels=1"
new format with delimiter: "Export:Mode=All:Filename=some/path/with
spaces.wav:Channels=1"

Not keen on this. Please suggest what character you would like as a
delimiter. ':' is a bad choice because of C:\\filename. Currently
think that filenames in double quotes, when needed because of spaces,
would be better.

Post by Nick
2) Small enhancement to allow binary data (float32) to be sent in the
"Data:{NBYTES}:XXXXXXXXXXXXXXXXXXXXX"
where NBYTES is the number of bytes that follow, to be reinterpreted as
float32 in little endian.

Where is this needed?
Why not pass the data in human readable format, 3.14159 ?

Post by Nick
3) Some commands to read/write data to/from tracks.

Would you be a bit more specific? If you mean read/writing sample data,
you would need to convince me. Currently we support a command channel
that is low bandwidth and text based. I am fine with the command
channel causing audacity to read or write a file of sample data. If we
start having high volume data over the scripting command channel, we
have to start addressing issues like how to abort an (accidental)
request for lots of data. It's additional complication and I don't yet
see the use case.

So if it is just for completeness, no. If it is needed for a script you
have in mind, let's hear about it and see if there is another way.

--James.

[Nick]

James,

1) Oops, yes, ':' was a bad choice, we could use ';' or '|', but I guess
using a quoted string for filenames would be okay. I'm not sure where else
using a space delimiter would be problematic, but for binary data it would
be.

2 & 3) Scripts/Plug-ins need to be able to process audio data, in my mind
this is a critical requirment.

The use case I have in mind is to use Python with Numpy and Matplotlib (or
other packages) to write algorithms that manipulate audio data. Using
Matplotlib makes debugging a snap since any data can be plotted.

I suppose one option using the current command interface is to write the
track to a temp file and the script can read the file, manipulate, write
back then reload the file in Audacity. File IO is slower than memcopies,
file IO also complicates the code.

Regarding aborting, it would be nice to press CTRL+C to stop the script and
signal Audacity to stop proccessing commands. But I think that would be a
longer term detail to work out. It will be a necessary feature, as it's
easy for a script writer to write an infinite loop somewhere... somewhere...
somewhere...

Also, while a command is being processed, it makes sense to disable gui
interaction and redraws while the commands are being processed. Idealy gui
updates can be turned on/off via the script, so if a script is doing lots of
things (like pipe-test.pl does), no time is spent on redrawing the gui for
each manipulation, just when processing is done, turn it back on and
Audacity redraws itself.

Sometimes I've seen guis popup a dialog which disables clicking on the main
gui, so the script can pop up a message like Processing ... and once it's
done it destroys the popup and the main gui is accessable again. Better yet
is a status line at the bottom of the screen that scripts/plugins can update
while processing stuff... and a log file...

Eventually I want to get to a place where I have an Audacity window up, then
in a shell I can start an interactive python session do stuff like:

import audacity
track = audacity.select_track(0)

from matplotblib.pylab import *

track.select(0.56, 1.56)
data = track.read()

plot(data)
grid(True)
show() # play with graph

# manipule data

track.write(data)
track.play()

I'm starting to ramble here.

:)
Nick



--
View this message in context: http://audacity.238276.n2.nabble.com/Decisions-for-2-0-7-tp7563467p7563636.html
Sent from the audacity-devel mailing list archive at Nabble.com.

[Vaughan Johnson]

Post by James Crook

Post by Nick
Hi All,
I'm working on using Google Protocol Buffers for communication between
mode-script-pipe-pb and scripts. I've just begun digging into this, but
1) Using a non-space for the delimiter for the string commands and
currently command: "Export: Mode=All
Filename=some/path/no/spaces/allowed.wav Channels=1"
new format with delimiter: "Export:Mode=All:Filename=some/path/with
spaces.wav:Channels=1"

Not keen on this. Please suggest what character you would like as a
delimiter. ':' is a bad choice because of C:\\filename. Currently
think that filenames in double quotes, when needed because of spaces,
would be better.

That's as I expected, and think seems totally reasonable. But not part
of what was this thread. Same for the next two items.

- V

Post by James Crook

Post by Nick
2) Small enhancement to allow binary data (float32) to be sent in the
"Data:{NBYTES}:XXXXXXXXXXXXXXXXXXXXX"
where NBYTES is the number of bytes that follow, to be reinterpreted as
float32 in little endian.

Where is this needed?
Why not pass the data in human readable format, 3.14159 ?

Post by Nick
3) Some commands to read/write data to/from tracks.

Would you be a bit more specific? If you mean read/writing sample data,
you would need to convince me. Currently we support a command channel
that is low bandwidth and text based. I am fine with the command
channel causing audacity to read or write a file of sample data. If we
start having high volume data over the scripting command channel, we
have to start addressing issues like how to abort an (accidental)
request for lots of data. It's additional complication and I don't yet
see the use case.
So if it is just for completeness, no. If it is needed for a script you
have in mind, let's hear about it and see if there is another way.
--James.
------------------------------------------------------------------------------
Meet PCI DSS 3.0 Compliance Requirements with EventLog Analyzer
Achieve PCI DSS 3.0 Compliant Status with Out-of-the-box PCI DSS Reports
Are you Audit-Ready for PCI DSS 3.0 Compliance? Download White paper
Comply to PCI DSS 3.0 Requirement 10 and 11.5 with EventLog Analyzer
http://pubads.g.doubleclick.net/gampad/clk?id=154622311&iu=/4140/ostg.clktrk
_______________________________________________
audacity-devel mailing list
https://lists.sourceforge.net/lists/listinfo/audacity-devel

[Vaughan Johnson]

[Luciano Boggino]

Hi everyone,

I've already make the patch file containing the changes that Federico
mentioned a while ago.

Post by Federico Miyara
1) Different alternative color scales for spectrograms
This is useful for people that are used to other software (Cool Edit,
Matlab) allowing a smooth and friendly migration. It allows sharing
spectral information with other people using other software using a
common "spectral language".
2) Level track, allowing a variety of responses used in acoustical
measurements, such as A, C, Z and K frequency weightings (recall that
K weighting is a new way of measuring loudness that has been proposed in
ITU EBU R 128 for broadcasting and already adopted or to be adopted in
many countries), as well as time weightings such as slow, fast and impulse.
This is very useful for people post-processing field recordings at the
laboratory with measurement purposes. This is very in line with other
project we have, the Open Sound Level Meter.
3) Cues and label import and export
Whether we like it or not, the most popular linear PCM uncompressed
audio format is WAV. In order to be able to share audio with people
using other software, it is important that cues and labels can be
recognized and saved into a wav file
4) Improvements to the spectral analysis (plot spectrum) allowing that
when no region is selected, instead of assuming "select all" and casting
a warning if too long, the specrtum is computed over a window centered
at the cursor position. This is the behavior of other software such as
Cool Edit or AdobeAudition and allows more precision in the time
location of the analyzed region.

The only problem I'm having testing it is that some files that I have
created, whenever i apply the patch aren't added to the source code. I
wanted to know if it is possible that the cause is the version of the files
and how should i summit the patch.

About the plug-ins,

5) Spectral edition filter.

Post by Federico Miyara
Although I have already commented on this, I explain it a bit more here.
It is a generalization of the spectral selection tool proposed by Paul.
While his selection tool allows to select only single rectangular
regions, this tool allows selecting arbitrary and multiple closed
regions that can be drawn with the mouse (you can choose also
rectangular or polygon). The selection can be saved as a preset. It is
similar to a free-of-charge program (but not open source no free
software) called SPEAR (but SPEAR uses an unconventional spectral
modelling). Once the region has been drawn, you can choose to remove it
or leave it alone. To this aim a variable filter is used. The result is
awesome. The selection tool is not currently intended for exporting the
selection parameters to a Nyquist script, but if this feature would be
interesting, I think it would be possible.
This filter is extremely useful for many situations such as interactive
and intuitive filtering in the spectral domain, isolation or removal of
some spectral feature, test for audibility of some portion of the
spectrum, electroacoustic music composition, spectral cleaning and
non-stationary noise removal. It is also a very interesting educational
tool, allowing a perfect interface for teaching soud from a spectral
perspective. We hear spectrally, but learn practical audio in the time
domain, where we are taught to recognize time features on the
oscilloscope. Why not spectral features? Why not listening to them on
the fly?
6) Time-variable filter.
This is a very sophisticated tool, still under development but that can
be already tested. It is a generalization of the previous one. You can
define the frequency response of the filter at specified time instants.
The interface is a table with pairs of columns associated with instants.
Rows correspond to frequencies, the first column contains the
frequencies and the second the frequency response. The plugin
interpolates over frequency and time producing a quasi-continuous
spectral morphing.
This is very useful for psychoacoustic experiments as well as for
electroacoustic music composition. It allows a time-variable fine-tuning
equalizer that can help tweaking a recording which is not very stable,
for instance, from a moving source around a room
7) Statistical computation and display of exceedance levels. Exceedance
levels are the sound pressure levels that are exceeded an N % of the
time. This is very useful in noise assessment since many psychoacoustic
criteria used in community noise depend on the statistics of the
variability of noise. The same is true for sound quality design for
household equipment, vehicles, etc.

should we make them a part of the source code?

Regards,

Luciano

[Leland]

Post by Luciano Boggino
The only problem I'm having testing it is that some files that I have
created, whenever i apply the patch aren't added to the source code. I
wanted to know if it is possible that the cause is the version of the
files and how should i summit the patch.

You need to add the files to SVN control using "svn add". It is a local
only change, but then "svn diff" will work the way you need.

Leland

[Vaughan Johnson]

Post by Leland

Post by Luciano Boggino
The only problem I'm having testing it is that some files that I have
created, whenever i apply the patch aren't added to the source code. I
wanted to know if it is possible that the cause is the version of the
files and how should i summit the patch.

You need to add the files to SVN control using "svn add". It is a local
only change, but then "svn diff" will work the way you need.
Leland

Yes, but only if he has commit privileges, right?

- V

[Paul L]

Post by Luciano Boggino
Hi everyone,
I've already make the patch file containing the changes that Federico
mentioned a while ago.

1) Different alternative color scales for spectrograms...
2) Level track, allowing a variety of responses used in acoustical
measurements...
3) Cues and label import and export...
4) Improvements to the spectral analysis (plot spectrum)...
5) Spectral edition filter...
6) Time-variable filter...
7) Statistical computation and display of exceedance levels...

Should these be made in one patch or several?

I am eager to see these changes. I anticipate some conflict in parts 1 and
4 with other work I am attempting and I want to cooperate to settle that
problem, whether it is you or I who submits first.




--
View this message in context: http://audacity.238276.n2.nabble.com/Decisions-for-2-0-7-tp7563467p7563713.html
Sent from the audacity-devel mailing list archive at Nabble.com.

[Ragazzi Imperiali]

[Ragazzi Imperiali]

[Luciano Boggino]

Post by Paul L

1) Different alternative color scales for spectrograms...
2) Level track, allowing a variety of responses used in acoustical
measurements...
3) Cues and label import and export...
4) Improvements to the spectral analysis (plot spectrum)...
5) Spectral edition filter...
6) Time-variable filter...
7) Statistical computation and display of exceedance levels...

Should these be made in one patch or several?

I could divide the first four features into several patches but probably
not one for each(since for example TrackPanel is used in more than one of
them), about the plug-ins maybe i should converge them into one MVS
solution and upload it like Simone did with the Aurora Plug-ins

Post by Paul L
I am eager to see these changes. I anticipate some conflict in parts 1 and
4 with other work I am attempting and I want to cooperate to settle that
problem, whether it is you or I who submits first

Maybe is better we solve this outside the list so we don't flood it with
mails?

----------------------------------------------------------------------------------------------------------------------

Post by Paul L
carissimo luciano rispondo a te perch? mi sembri un p? Italiano come me

..OKey?

Post by Paul L
io avrei solo fatto una domanda via email di questo tipo ....(vedi sotto

in inglese) ma mi sono pervenute circa 150 risposte a cui non so >che dire
... Ho solo capito che posso scaricare il programma e poi posso fare copia
e incola sulla pagina del mio sito al tasto CONTATTI.

Post by Paul L
IL mio sito ? allocato come vedi in Italia (

http://casaimperiale.paginelibere.it < . it

Post by Paul L
al tasto contatti ora vi ? solo una segreteria formato scherzo, elaborata

da il sig. Enrico Lai- Informatico di Cagliari ...ma ora lui non se ne

Post by Paul L
occupa pi? di queste cose ? impegnato in altre cose
appena posso scarico il programma e vedo se funziona come vorrei.
eventualmente posso avere un po di assistenza da te o quant'altro ?
a presto

I'm sorry, I don't speak italian (I'm from Argentina). Could you write this
again in english since the google translator doesn't give me a coherent
translation?.

Regards,

Luciano

[unknown]

carissimo luciano rispondo a te perch=E8 mi sembri un p=F2 Italiano come me=
..OKey?

io avrei solo fatto una domanda via email di questo tipo ....(vedi sotto i=
n inglese) ma mi sono pervenute circa 150 risposte a cui non so che dire .=
.. Ho solo capito che posso scaricare il programma e poi posso fare copia e=
incola sulla pagina del mio sito al tasto CONTATTI.

IL mio sito =E8 allocato come vedi in Italia (http://casaimperiale.pagin=
elibere.it < . it=20

al tasto contatti ora vi =E8 solo una segreteria formato scherzo=2C elabora=
ta da il sig. Enrico Lai- Informatico di Cagliari ...ma ora lui non se ne =
occupa pi=F9 di queste cose =E8 impegnato in altre cose
appena posso scarico il programma e vedo se funziona come vorrei.
eventualmente posso avere un po di assistenza da te o quant'altro ?

a presto=20

I'm happy to have found something that will help me .. at least I hope

I would just help to know if Audacity is suitable to be placed on a page =

of my site ...
http://casaimperiale.paginelibere.it/index.jsp?lingua=3DIT&action=3Dcambia=
Pagina&sito=3Dcasaimperiale&linkPagina=3D7547&t=3DCONTATTI
=20
with the capabilities of secretarial staff and real mode to listen to the v=
oice messages that come from my users.

And 'feasible?
You give me a hand to customize this voice mail online.
I'm sure you make it happen for me ... it will be a very salable product.
as a package softwear ... like a new little program to be placed on your =

site. will go to great

now there is one that is just a joke its manufacturer has a website but n=

o longer works with these things.

and I need a little help from the developers of this program.
L I M P E R I A L E

=20
=20
=20




L I M P E R I A L E



________________________________

Date: Sun=2C 5 Oct 2014 18:57:35 -0300=20
Subject: Re: [Audacity-devel] Decisions for 2.0.7=20
=20
Hi everyone=2C=20
=20
I've already make the patch file containing the changes that Federico=20
mentioned a while ago.=20
=20
They involve:=20
=20
1) Different alternative color scales for spectrograms=20
This is useful for people that are used to other software (Cool Edit=2C=20
Matlab) allowing a smooth and friendly migration. It allows sharing=20
spectral information with other people using other software using a=20
common "spectral language".=20
=20
2) Level track=2C allowing a variety of responses used in acoustical=20
measurements=2C such as A=2C C=2C Z and K frequency weightings (recall th=

at=20

K weighting is a new way of measuring loudness that has been proposed in=

=20

ITU EBU R 128 for broadcasting and already adopted or to be adopted in=20
many countries)=2C as well as time weightings such as slow=2C fast and im=

pulse.=20

This is very useful for people post-processing field recordings at the=20
laboratory with measurement purposes. This is very in line with other=20
project we have=2C the Open Sound Level Meter.=20
=20
3) Cues and label import and export=20
Whether we like it or not=2C the most popular linear PCM uncompressed=20
audio format is WAV. In order to be able to share audio with people=20
using other software=2C it is important that cues and labels can be=20
recognized and saved into a wav file=20
=20
4) Improvements to the spectral analysis (plot spectrum) allowing that=20
when no region is selected=2C instead of assuming "select all" and castin=

g=20

a warning if too long=2C the specrtum is computed over a window centered=

=20

at the cursor position. This is the behavior of other software such as=20
Cool Edit or AdobeAudition and allows more precision in the time=20
location of the analyzed region.=20
=20
The only problem I'm having testing it is that some files that I have=20
created=2C whenever i apply the patch aren't added to the source code. I=

=20

wanted to know if it is possible that the cause is the version of the=20
files and how should i summit the patch.=20
=20
About the plug-ins=2C=20
=20
5) Spectral edition filter.=20
Although I have already commented on this=2C I explain it a bit more here=

.=20

It is a generalization of the spectral selection tool proposed by Paul.=20
While his selection tool allows to select only single rectangular=20
regions=2C this tool allows selecting arbitrary and multiple closed=20
regions that can be drawn with the mouse (you can choose also=20
rectangular or polygon). The selection can be saved as a preset. It is=20
similar to a free-of-charge program (but not open source no free=20
software) called SPEAR (but SPEAR uses an unconventional spectral=20
modelling). Once the region has been drawn=2C you can choose to remove it=

=20

or leave it alone. To this aim a variable filter is used. The result is=20
awesome. The selection tool is not currently intended for exporting the=20
selection parameters to a Nyquist script=2C but if this feature would be=

=20

interesting=2C I think it would be possible.=20
This filter is extremely useful for many situations such as interactive=20
and intuitive filtering in the spectral domain=2C isolation or removal of=

=20

some spectral feature=2C test for audibility of some portion of the=20
spectrum=2C electroacoustic music composition=2C spectral cleaning and=20
non-stationary noise removal. It is also a very interesting educational=20
tool=2C allowing a perfect interface for teaching soud from a spectral=20
perspective. We hear spectrally=2C but learn practical audio in the time=

=20

domain=2C where we are taught to recognize time features on the=20
oscilloscope. Why not spectral features? Why not listening to them on=20
the fly?=20
=20
6) Time-variable filter.=20
This is a very sophisticated tool=2C still under development but that can=

=20

be already tested. It is a generalization of the previous one. You can=20
define the frequency response of the filter at specified time instants.=20
The interface is a table with pairs of columns associated with instants.=

=20

Rows correspond to frequencies=2C the first column contains the=20
frequencies and the second the frequency response. The plugin=20
interpolates over frequency and time producing a quasi-continuous=20
spectral morphing.=20
This is very useful for psychoacoustic experiments as well as for=20
electroacoustic music composition. It allows a time-variable fine-tuning=

=20

equalizer that can help tweaking a recording which is not very stable=2C=

=20

for instance=2C from a moving source around a room=20
=20
7) Statistical computation and display of exceedance levels. Exceedance=20
levels are the sound pressure levels that are exceeded an N % of the=20
time. This is very useful in noise assessment since many psychoacoustic=20
criteria used in community noise depend on the statistics of the=20
variability of noise. The same is true for sound quality design for=20
household equipment=2C vehicles=2C etc.=20
=20
should we make them a part of the source code?=20
=20
Regards=2C=20
=20
Luciano=20
=20
-------------------------------------------------------------------------=

-----=20

Meet PCI DSS 3.0 Compliance Requirements with EventLog Analyzer Achieve=20
PCI DSS 3.0 Compliant Status with Out-of-the-box PCI DSS Reports Are=20
you Audit-Ready for PCI DSS 3.0 Compliance? Download White paper Comply=20
to PCI DSS 3.0 Requirement 10 and 11.5 with EventLog Analyzer=20
http://pubads.g.doubleclick.net/gampad/clk?id=3D154622311&iu=3D/4140/ostg=

.clktrk=20

_______________________________________________ audacity-devel mailing=20
https://lists.sourceforge.net/lists/listinfo/audacity-devel =

[unknown]

--_2e740e9b-a016-4b23-9897-86a6e2fc56c1_
Content-Type: text/plain; charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable


carissimo luciano rispondo a te perchè mi sembri un pò Italiano come me ..OKey?

io avrei solo fatto una domanda via email di questo tipo ....(vedi sotto in inglese) ma mi sono pervenute circa 150 risposte a cui non so che dire ... Ho solo capito che posso scaricare il programma e poi posso fare copia e incola sulla pagina del mio sito al tasto CONTATTI.

IL mio sito è allocato come vedi in Italia (http://casaimperiale.paginelibere.it < . it

al tasto contatti ora vi è solo una segreteria formato scherzo, elaborata da il sig. Enrico Lai- Informatico di Cagliari ...ma ora lui non se ne occupa più di queste cose è impegnato in altre cose
appena posso scarico il programma e vedo se funziona come vorrei.
eventualmente posso avere un po di assistenza da te o quant'altro ?

a presto

I'm happy to have found something that will help me .. at least I hope

I would just help to know if Audacity is suitable to be placed on a page of my site ...

http://casaimperiale.paginelibere.it/index.jsp?lingua=IT&action=cambiaPagina&sito=casaimperiale&linkPagina=7547&t=CONTATTI

with the capabilities of secretarial staff and real mode to listen to the voice messages that come from my users.

And 'feasible?
You give me a hand to customize this voice mail online.
I'm sure you make it happen for me ... it will be a very salable product.
as a package softwear ... like a new little program to be placed on your site. will go to great
now there is one that is just a joke its manufacturer has a website but no longer works with these things.
and I need a little help from the developers of this program.
L I M P E R I A L E

________________________________

Date: Sun, 5 Oct 2014 18:57:35 -0300
Subject: Re: [Audacity-devel] Decisions for 2.0.7
Hi everyone,
I've already make the patch file containing the changes that Federico
mentioned a while ago.
1) Different alternative color scales for spectrograms
This is useful for people that are used to other software (Cool Edit,
Matlab) allowing a smooth and friendly migration. It allows sharing
spectral information with other people using other software using a
common "spectral language".
2) Level track, allowing a variety of responses used in acoustical
measurements, such as A, C, Z and K frequency weightings (recall that
K weighting is a new way of measuring loudness that has been proposed in
ITU EBU R 128 for broadcasting and already adopted or to be adopted in
many countries), as well as time weightings such as slow, fast and impulse.
This is very useful for people post-processing field recordings at the
laboratory with measurement purposes. This is very in line with other
project we have, the Open Sound Level Meter.
3) Cues and label import and export
Whether we like it or not, the most popular linear PCM uncompressed
audio format is WAV. In order to be able to share audio with people
using other software, it is important that cues and labels can be
recognized and saved into a wav file
4) Improvements to the spectral analysis (plot spectrum) allowing that
when no region is selected, instead of assuming "select all" and casting
a warning if too long, the specrtum is computed over a window centered
at the cursor position. This is the behavior of other software such as
Cool Edit or AdobeAudition and allows more precision in the time
location of the analyzed region.
The only problem I'm having testing it is that some files that I have
created, whenever i apply the patch aren't added to the source code. I
wanted to know if it is possible that the cause is the version of the
files and how should i summit the patch.
About the plug-ins,
5) Spectral edition filter.
Although I have already commented on this, I explain it a bit more here.
It is a generalization of the spectral selection tool proposed by Paul.
While his selection tool allows to select only single rectangular
regions, this tool allows selecting arbitrary and multiple closed
regions that can be drawn with the mouse (you can choose also
rectangular or polygon). The selection can be saved as a preset. It is
similar to a free-of-charge program (but not open source no free
software) called SPEAR (but SPEAR uses an unconventional spectral
modelling). Once the region has been drawn, you can choose to remove it
or leave it alone. To this aim a variable filter is used. The result is
awesome. The selection tool is not currently intended for exporting the
selection parameters to a Nyquist script, but if this feature would be
interesting, I think it would be possible.
This filter is extremely useful for many situations such as interactive
and intuitive filtering in the spectral domain, isolation or removal of
some spectral feature, test for audibility of some portion of the
spectrum, electroacoustic music composition, spectral cleaning and
non-stationary noise removal. It is also a very interesting educational
tool, allowing a perfect interface for teaching soud from a spectral
perspective. We hear spectrally, but learn practical audio in the time
domain, where we are taught to recognize time features on the
oscilloscope. Why not spectral features? Why not listening to them on
the fly?
6) Time-variable filter.
This is a very sophisticated tool, still under development but that can
be already tested. It is a generalization of the previous one. You can
define the frequency response of the filter at specified time instants.
The interface is a table with pairs of columns associated with instants.
Rows correspond to frequencies, the first column contains the
frequencies and the second the frequency response. The plugin
interpolates over frequency and time producing a quasi-continuous
spectral morphing.
This is very useful for psychoacoustic experiments as well as for
electroacoustic music composition. It allows a time-variable fine-tuning
equalizer that can help tweaking a recording which is not very stable,
for instance, from a moving source around a room
7) Statistical computation and display of exceedance levels. Exceedance
levels are the sound pressure levels that are exceeded an N % of the
time. This is very useful in noise assessment since many psychoacoustic
criteria used in community noise depend on the statistics of the
variability of noise. The same is true for sound quality design for
household equipment, vehicles, etc.
should we make them a part of the source code?
Regards,
Luciano
------------------------------------------------------------------------------
Meet PCI DSS 3.0 Compliance Requirements with EventLog Analyzer Achieve
PCI DSS 3.0 Compliant Status with Out-of-the-box PCI DSS Reports Are
you Audit-Ready for PCI DSS 3.0 Compliance? Download White paper Comply
to PCI DSS 3.0 Requirement 10 and 11.5 with EventLog Analyzer
http://pubads.g.doubleclick.net/gampad/clk?id=154622311&iu=/4140/ostg.clktrk
_______________________________________________ audacity-devel mailing
https://lists.sourceforge.net/lists/listinfo/audacity-devel

--_2e740e9b-a016-4b23-9897-86a6e2fc56c1_
Content-Type: image/pjpeg
Content-Transfer-Encoding: base64
Content-Disposition: attachment; filename="mara angelina.jpg"
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--_2e740e9b-a016-4b23-9897-86a6e2fc56c1_
Content-Type: image/pjpeg
Content-Transfer-Encoding: base64
Content-Disposition: attachment; filename="mara.jpg"

/9j/4AAQSkZJRgABAQEAYABgAAD/4QA2RXhpZgAATU0AKgAAAAgAAYdpAAQAAAABAAAAGgAAAAAA
AaABAAMAAAABAAEAAAAAAAAAAP/bAEMABQMEBAQDBQQEBAUFBQYHDAgHBwcHDwsLCQwRDxISEQ8R
ERMWHBcTFBoVEREYIRgaHR0fHx8TFyIkIh4kHB4fHv/bAEMBBQUFBwYHDggIDh4UERQeHh4eHh4e
Hh4eHh4eHh4eHh4eHh4eHh4eHh4eHh4eHh4eHh4eHh4eHh4eHh4eHh4eHv/AABEIAJ8A1QMBIgAC
EQEDEQH/xAAfAAABBQEBAQEBAQAAAAAAAAAAAQIDBAUGBwgJCgv/xAC1EAACAQMDAgQDBQUEBAAA
AX0BAgMABBEFEiExQQYTUWEHInEUMoGRoQgjQrHBFVLR8CQzYnKCCQoWFxgZGiUmJygpKjQ1Njc4
OTpDREVGR0hJSlNUVVZXWFlaY2RlZmdoaWpzdHV2d3h5eoOEhYaHiImKkpOUlZaXmJmaoqOkpaan
qKmqsrO0tba3uLm6wsPExcbHyMnK0tPU1dbX2Nna4eLj5OXm5+jp6vHy8/T19vf4+fr/xAAfAQAD
AQEBAQEBAQEBAAAAAAAAAQIDBAUGBwgJCgv/xAC1EQACAQIEBAMEBwUEBAABAncAAQIDEQQFITEG
EkFRB2FxEyIygQgUQpGhscEJIzNS8BVictEKFiQ04SXxFxgZGiYnKCkqNTY3ODk6Q0RFRkdISUpT
VFVWV1hZWmNkZWZnaGlqc3R1dnd4eXqCg4SFhoeIiYqSk5SVlpeYmZqio6Slpqeoqaqys7S1tre4
ubrCw8TFxsfIycrS09TV1tfY2dri4+Tl5ufo6ery8/T19vf4+fr/2gAMAwEAAhEDEQA/APkIDrjN
KoHX+lA9uKeg9a9iSVtTjJrVcyA9s191/CWHZ8NvDgYMCdLtmIb3iUmvhe1B81V9SK+8vhaqr8Of
DYBGBpNqOP8ArilebitGjZL3To1j5wMflUixZ7cU5MZ4x+VWEXIrnsCIBENpFIIcmrxTjikWOnoU
isIsHFT28ZzUm3g55qeBAeOKFuNrQasfNO2d+1TBP8M1Bf3MFlbSXFzIsUUa7mZjgAVqSkBRcZxT
Sq5+8M14B8SfjpdJLLaeGIljjUlftMi5LH1UenvXAaZ8YPHK3XnPqnmg9VdFx/Kha6l8kj68YLnq
KayKe1fJlj8XPF82oPJLqLEscbQg2ivXvA3xRjuvKttaAhdzjzP4aHZOwezdrnqDxjkComjH1NXI
9s0ayRsGVhkEdDSFOvFCM2ig8fXj9KhkjO7p+lX3XmoZE+b0obJsZ7xHB4+lNEfGeKtSqQc44qPa
c47UWIauVJE4PaoXjwe3FW5F56GoWOT0pPUzkjyX9pJQPBEK8c3advZq+bXTnrxX0r+0qD/whdqQ
ODfJnA6/I/8A9avnCRa9LCr92LoVGTmipWXnjNFdNibnMcd6lRTnNVVYlxzirKkYrKS0Oi1mWbUf
vB9a+/8A4fwiPwZosQA+SwhXgccIK+ALf7w44Jr9CvBqFfDGmAD/AJdY/wD0EV5uJa5kjZbGyiCp
lX0FCKKnROORXONIaqnbShMnGKmCkCjHakVYiKHPAqeFOopMfWpoBjPBxQtxsY/yqc9uTXifxU8U
/wBtXlxoVlLm3iUhyD95snOfYV6Z8TNXGi+E726Rwsmzah9CeK+XrfUUiu7m8y7A7UOAWx6mprVG
lZHVg6HO7sjn8FNdFpRkKOKgl8Az26DBJZgSBXoXh3WNOuo1ijkDDjqMEGuo8Q20EWjHUvMUiJN2
MZyPSvO+sVE7Hs+wppao8Bs/DFzDMxKHg5rr47JU0tZCBuTqKtT6xaGfdEgKt/CeCKp3l/GqOsbD
Y3UV0xrSk9TGWHjGOh6v8E/FHmxf8I/dzFnALWjMc5UdU/DqPbPpXqhTjNfKPhjUJrTU4bmA7ZYJ
BLHx3B6H2I4PtX1Pot7HqOk2t9F9yeJZFB6jIzg+/NdsW2jx69Pllccy9hUMi57VeZO3aoZE7gVR
zWM6UZHIqJl4OAfersiZPWoXHFMhooyA88VXcY64NXpFAPWq7Jmi5m0eP/tMnHhCxVWA3X65A6ke
XJ/9avnZlHJr6J/afG3wrpYwedQB46cRSV8+PgjkZNelhP4dyJaFKQDPSipmHPNFdFl2JscNAMyA
1dUcdfyqrbLkk1aAzWcjd7lm3xuHev0T8NRMmiWCSDDLbxg8f7Ir87Icg8HntX6O6LtFjB2Hlr/K
vLxK95G8di9Gp6Zqwqmmx7ccGp068VgNDSvHek+gqfb2/KkwCcilcaIhnvU8fHAFJsH5UFMAnPFE
dymjxT9pvVZP7Oi0uB8bEM0gB78hf/Zv0rx3wvY3N9dW+lXU7WNrIvmTTBMknjA+pr0H4sE6jq2q
PK5KeaqDthRgY/PJ/GpdCggaGPaq78DJIrgrYizva572Cwt4djA8G6dqlqyDU7RApcKHH8Q/r9a7
vVkefba2sfyKmWUHOabOkYurZbiRIoycLngU/R5EutWkS0lSUKeCpz061xSlzS5kj0VBqOvQ83+I
Hhe7srS21OxuFluZZSJoQv8AqxjIJ+vP5e9ZtvZ3V5YedLbGKQLyAOte7alBA1uRcQxkjrmuL1uW
CMMqIoU1tKu01FGdKm5Nt9ThNLt3S6jfHGRmvoL4RXTSeHGsnOWtpWA9SrHcD+pH4V5VoWmC9s5p
oxyjkcD2B/rXoPwrlMF5JCwxvXn3IrtoVbux5OOopJ26HpRHHpUbjmpckgU18EV2HjtFOQck1XmH
y9OtXXX26VXkHHSmTJFCbPTGKrscGrdwowTgZ71SkB3UzJo8a/agYnQdHXHP24np/wBMn/xrwJ2w
Owr3r9qB9miaOOcNen8/LevAXc56V6WD/hkSiMIGeeaKQtg88UV1XZmcdbriPJqUZwMU2EnbgmnA
+1ZSvY23ZasBm4Re5YAV+jmkACzgVscRqMAe1fnNoaGbVbaIAkvKq4981+jmmqY4I4yPmCgH64ry
cT8SN4LQ0o8EcAfhU6fTFQJ24qdMdMfWsCiT+HmjHFKcUiMAeTQUhVHJpLtvLtJmz0U0/JPSqutS
+XpFy5PSMn9KTejZSV2fPfiuBruS/C4JeQn9cio/CzsIYi3RlB5rVusQax9nmYAXCGRM9yDzWT4f
VIb270tpCGtpiVyedj/Op+nJXP8AsmvHk+ZH1GGfLodBrUWj3LW66lNDDs4Uu4XBPpV3wjY6Dp10
11plzbyk8OyOG5OOuK52TwnFqd6bprmRpG7SfMB7AVdg8E29lKl2bqZGX/nkdgx+FRGSSOuUafLb
m1Oi8RXCsX2kYPNee695rbgOpPeuyuEhC/LIzYGCWNcxqUT6hqCWFoRvfqR/Cvc1lFvnuELQhY3f
hpbhNAubibhGmYqfUbVH8wa3tDCwX9tcRHAdq5zxLq9vo+mW+iWsi+cwVSFPKjt+J/xrd0wtDbWM
PJZFXJ9TxXfCTTR41dc7b7npkcgKg5p+SeRVS2IMS59OlWVIxXqnhtDJc1E4BXmp2GaiccYzTRDK
M69qpSLjPStGcVUkXqTQQ0eGftT/AC6Voi5GPtTt0/2D/jXz83U8177+1U+bTRIv+m0jfT5cf1rw
JyA2Dya9PCNezImhjZJ7UUjdaK6L+ZByyfKvY8YppPPSmwnj3p31qZaDSNXwmpfxFpycjNzGOP8A
eFfo5aAYx6V+dfgBDJ4x0ZAAS19CMev7xa/Re0GFAyDx2rycS7zOlaIuLU64AqFFOOgqUDtWA0PP
FIvfnNI2RihSP1oKRKuKz/EvOkSqByeKvqcDPFUteVpdOdVGSaio/dZrBao+evil9oja2vLQss1p
KpUj+6eCK4vXtcnjkg1yw/4+IlEc8Z/iXOcH/GvXPEWjPNqIhkBKMmTn1zXnPj3w6kNqILQbZ5H3
HB5wBz/MV5UKiuk0fQ07OOhe8M/EzSmjjldjGw4dWHQ1e1n4o6SQVjO8nsteJ6TZNHqs1nNgiQbl
NdjpHhO3eRXc/KTznmtK1OlA6sP7SrG9ldGwfGd7qcrRWNpI244z2FbB1C78N+HLnUWjzeS8BiOm
a1NPg0jQNMkupBHFHEhZpHxhR6151qvji28VR6hoyQmO2MqSW8vQsB97PoDxge1RRTlK8Y+6h4iS
VoN6vYtfDV73XPFcd3fu8jM5Ybu5xnP8zXtgkWO+hzgbpAq/QV4v8NtQtbPxCsAONoIznvg4Fdtq
+sy/8JDbKhIiCZX6961qu8kcFWmlLli9j2uxl3IPSr0TZ6muf8M3yXmmQyqwJ2gNz3rbVge9enF3
SZ4FSNm0TM5GcYqNjxQSCMZprNx1p3MmRSEkfeFVJhg1YlbGeRVaUkg/zouS0fP/AO1hLsl0FCCA
xmOfUgL/AImvCQ27oa9n/a+bbqfhxSODFcn6kGL/AB/WvDoJQoH+NephX+6IktSwwBPJxRUMs/PG
AO3NFa8xJycT8elSCQtxVUVNCfl56e1XLYVtTrfhepb4geHVBOTqlqOn/TVa/RC0OVB9q/PP4SKZ
PiP4cGM/8TW1OPpKp/pX6F2rAKAfSvIr25zdbF5fvA9qkYkHIqGM9OlSE478fWsRoc7cdaRDke9N
dsjjpTVPHGOtT1KRZGD3qDUZQlsxxmnhgOpqnfsZEeNe45PtUT2aNIrU4bxldLaMk2Msy/KK81uz
Jez3F3M2CRsRj/APUV2nxKu1dLeOE/MpIz6DvXkOv6vmN9PsZGbtJJnqe+PavEcG52Wx9JhV+78z
nrqBZ/E0dxZxf6PFJsLe2Nua7qN/7OgzclI0Rcl24AHqawrLTlhslWNvmYbnJ7Gsv4oayNcv/sMM
pt9KthhmDczsOuf9kHp69fSteX281Hoj0adOVGm2ldswfHXiq98V3i6Zp+9dOhPPYSH+83sOwrFu
LiLRrT7NbkNcuPmPp71V1DXLezh+y6WoHGDJj/Oaw7dvtNz5lxKQM5Zj1r26VH3UrWiuh4WIxcaE
m4S5qj69F5L/ADO38LT3TFGbY0uc7wMkD3NdH4g8afYQiYWSUDAJNcVYaqBC6xnyLOFc7f4pD2ya
5nUr2S7uGlkbJPT2FN4dVJ3kjg+tOnCyerPpX4LfEuye6XTbxjF5rfLubIDex9698t51kQMjZB6V
+d+n301pOssT7SpyCD0r6b+AvxTi1YRaBrMuy+6QSHpLjt9ePxqpUuXWJySq8/xbnvu/jPemu3qe
tQK+5QQaHfjrUEMJW4qvIQOM06VuTg1Xdsgmi5DPnL9sF/8Aid+GgD/y73Rx+MNeFiYKvHGK9r/a
8kB8ReHk9La4P1y0f+FeFNkivUwi/dakTZJJckt0/OiqTOSxxjiiuuxg5mbUsRI+7kmoqeh7fyqZ
bGrO9+COH+KnhpcZxqcBx9HB/pX31aMCvJr4I+A4z8VPDzEn5LxWyOCMZNfd1nJwK8XEP3zZbGuj
8AE/nUuRnrVKOTI61Mh5wTUDROxGKTIFMJ69CaQMcnpSaLRKzEjaDVa6iBtpdzsBtxwalDD15qK+
KtavvbC4yeanluWjwr4tanBFPLYWEqvLH8su05KnAOCfWvIb27h0tBPdOBvPC9z9BXWS2lsmparL
BcPPa3F/PLA7H/lkXJX8COntjvXjPjLUXvdduZA5CI5WMZ4AHArmpYdTqSj0PYeKdCjFrc6TXfHl
3s8q2tWit3XAL/eb/wCtXF6lql1fOTLIQp6KOlUppndhuYsfc1EzcYzXoU8PCn8KPOxGYVqys5Ow
8ntTt3G0cCodx9qdu4zXR5HEXHn/ANCWJOmcn3qkWOeelLuPT1psh7DNAhdwA44q/pV9PZ3cNxby
tFLE6ujqcFWByD9QRWaxwKdGxDDGOtOImfdnwn8WR+LvBNlqwws+0xXKDokq8MPYHhh7MK60yAjr
Xyz+yl4mks/Et54cmcfZ7+IzQgt0mjHIA/2kyT/1zWvprzMrweK5akbMad0SSScYqJmwOtMeTp3q
J2wc9ayuJs+bv2uJFPinQgCMrZzE/i6/4V4e7Ef0r2b9rGb/AIq7SlOMLZMR+L//AFq8U8wMDkYx
Xs4O/skZVGMOMnJFFMfBORnmiuq6MCnSjHrikozWMtToPR/2fFL/ABY0BdgY+ex59o2P9K+4bZiM
d6+IP2cGz8XdCxgnfN/6Ikr7Zt3xgV4+I/iWNehqxvU4fPGcGs5JB25qYSVlcZfD+9NL4zVcScDm
ifbLG0ZJAIxkHB/A0PYqJ5f8TfiNf6Fq5ttIZMugVxKnMTo7ZOD2ZSuD3xXKa/8AFnUtU8Hto5UR
3szCOa4TgNFg7hjsTwPpnvVL456GdH1m3nS+kuI5o8BJWLSRgE45PVev05rzMSBJdxYV58qklNn3
eW4DB4jCxfLfz63NHU9avJLF4IQQQMh84Oa8m1Bma5fJ53HOK9I3qUdmGQTisMaHaLcyXUpB3ZKR
9ga2w2KVO6kZ5rkbrNOj317I4h29iabu4xzVi/tZLS4Mcu3d14Oaqscgg160JKSuj4mpTlSm4SWq
FDHOTwBTt/eod3ykUdDTMyxke1KaiyAOnFO3ChAOIGOBxTQGBJxSbsUrHjIoWgjqPhlqiaR470TU
ZpvJihvYjK56LGWw+fbaWz7V9t+aQOTg9CK/P1G2+9fZXwh8THxR4EsL+d911Gpguc95E43f8CG1
v+BVnWTdmStDt/NHGCKY0vHX8arFge446U8ldlcw2fMv7VMu7x5ZRnOF09WGe+ZHH9K8cc8Z9a9Y
/ahcH4jwLu4GmRcf9tZa8lYDGOa9nCL9yjKpuMxk8Z/OigNgniiurmkiNSCkNGaDWMtVobHpX7M4
J+MOj9wqzn/yBIP619pQnnp+Zr4x/ZgRX+MGm5GSsU5U+h8lx/Imvs23xjk5NeTiP4jKuW0O0Yqe
NvyqoMY9afG4FYDuXFkA9KkDDrVIuOoqRX4znrRqUmeZ/tMW0K+AX1gBvPtLiAKqj/WB5Vj2k+g3
k/UCvm6W5fzgpBXjkEV9XfF2ztNT8Fz2d5uMbXFu+0fxFJkcD6fL+Wa+XvH1jNomq/aZHLw3Em1R
j7o7CuWrCMnbqfQ5TiqtGO/uXHwO3ljIDd6z9Y1CW2txcJbrKFPIz0rN1/XGs7VYbY/vHGQ3oK1f
hBo0fiDWJZ9VvMwRbP3LSgGY5yFwc5B24OPUVFLDWXPPY9vH5zGD9jS1k+vY47WLqS8mimuYFhVl
wu0YyPWsuUKG2ody+uK7j4jQ6DDdaikdxK17HMsdnFAAYVhBwd5PO7HRQMCuDkODgEV6tK3KrHxe
KnKdRubu+4qjkc4xxTd2CRnn1pNxPrmkbAYZ7Vukcw8HjpmkL9c5Bpu7Ck5NMLHgZFAifeNuOTUm
Tsqpu5yc1LuJU45FSJkq5I/pX0P+ypdyvomtWhJ8qK4ikUdtzqwP6RrXzmHIPX8K93/ZWv8AY+t2
JX7wimXHtuDfzWlNPkdyep9AE0u44qp5vTmpTINvfNcQ2z5g/abk8z4mlcn5NPhXr0+Zz/WvLC2D
ivRv2jZhL8ULwDqkEKH/AL5z/WvNm9K9rDaUY3M5asfx6/nRUQO3jmitri5WMoNGKQ1Dehoeo/sv
cfFm1frttp+/qhH9a+w7eRiO1fHX7MhA+KMLnPy2spz+GP619eWso215Fdr2jBs1Efj/AOvQz4ya
rrJjigv1rJDTMbV/G+g6VdyWt3cyLMnUCJjz6ZxzVA/E3w3glZrhsZwBGecVN481mz8PeHbnVrm2
WZo9qIm3l3ZgqjPYZPJ9M1wuj+DX8SWkeu6xdfZWuR5kUFoiqFj7ZPTJzn6Yzz0yk5bJmsWuqOj8
far9tbTTFKDaXVt58RB+90OfyK/ma5+98OWuvaP5N4iyK3VSvI981574K1u6uL++trxnl+zzlYEZ
ifLRWI25/KvUbTWI5AixpsIGMdq453jNtnu4Z2orkPJvE3wut4opPsyyBx9w7icfUGvM5lv/AA7d
SQPG8NyrAo4JGMHgivrhzBcxZdct9K47xr4K0zXbNo54lSTB8uQDlTWtHFOGktUZ1aEaq9xWkj5h
1K9lvLqS4nYtLIdzMTkk9yaqOT14rf8AFnhy70LUZLWfa6qfldT1HrXPvgE161KcZK8Tx6kZRk1L
cchwQcc02U889Kbu4PJ4pjNk5HWtepkPOSpH40xjyAKTdxTaTYEisRTo36+lQ5pRSvdgTA5PXAr1
r9mq+SDxlcwOSDNZui8dW3o38ga8gB5zXoHwGnEXxJ00McK6zKR6/unI/UUpWcWRJH1ghBGSf1qU
kAZ61nxyAqMd6labC4BrhbFLQ+WPj3OZPivrRPQeQB/34j/rmuAJJrrfjLcfafibrcucjzkXp6Rq
P6VyNe5h/wCFH0G1qBopKK2ugP/Z

--_2e740e9b-a016-4b23-9897-86a6e2fc56c1_--