possible recording bug in 2.0.5

classic Classic list List threaded Threaded
20 messages Options
Reply | Threaded
Open this post in threaded view
|

possible recording bug in 2.0.5

Peter Sampson
Management Summary
==================
We appear to have a recording problem on 2.0.5
Using any of the three hosts: WASAPI loopback, WDS or MME, I get consistent
occasional clicking in recordings which are not present in the source material.
 
This is a regression as using 2.0.4 with the same input material yields a clean
recording with no clicking.
 
Testing on:
audacity-win-r12630-2.0.5-alpha-02-oct-13
audacity-win-r12644-2.0.5-alpha-04-oct-13
Windows-7 Toshiba Satellite laptop 494/881 GB free space
6 Gb RAM 2.5 GHz CPU 64-bit
 
Long version
==========
I have been using recent 2.0.5s as production engines recently to give them a more
thorough workout.  This came to my attention recording from YouTube recently
using WASAPI loopback.  I transferred a recording to CD and played it on my hi-fi
to be surprised with a lot of clicks and thumps that I did not hear when recording.
 
I went back to the Audacity project (assuming these to be on the original YT recording)
and marked up the approximate locations intending to edit them out.  Once I had done
that I noticed that the clicks/thumps seemed to appear at reasonably regular intervals
starting out at around every ten seconds, changing to every five seconds mid-song and
then back to mainly around ten seconds.
 
I looked at another recording I had made from YT that day and when zoomed in and
using the Spectrogram view I noticed that many of the faults were double faults and
that they seemed fairly consistently to be 900-950 samples apart.  Viewing zoomed in
in waveform view showed clear discontinuities in the waveform.
 
I then went back to a recording I had made on Thursday evening from BBC FM radio
inputting through my Edirol UA-1EX USB soundcard only to find on detailed listening
that I had similar clicks/thumps there albeit not as frequently.
 
At this juncture I went off to the FAQ in the manual that deals with such stuff:
and made sure I took the advice there, in particular I upped the recording audio buffer from
its default 100 to 200 milliseconds, killed all extraneous processes, rebooted the PC.
Re-recording still yielded the problem.
 
I also tested a nice clean source of iTunes playing a ripped CD (256 VBR) and this
yielded the same clicky/thumpy results.
 
Subsequent testing on 2.0.4 with all the same source material yielded the nice clean
recordings that I have come to expect with Audacity.
 
I do have the original marked-up project (4.5 minute song) that I can supply as a 32-bit
WAV along with an export of the labels if that might help with any debugging.
 
Sorry to be the bearer of potentially bad news at this late stage in the cycle, but I'm
sure we really don't want any more (non-blue) egg on our faces ...
 
Thanks,
Peter.
 
Peter Sampson
Tel: +44 (0)1625 524 780
Mob: +44 (0)7732 278 299

------------------------------------------------------------------------------
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application performance.
Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
the latest Intel processors and coprocessors. See abstracts and register >
http://pubads.g.doubleclick.net/gampad/clk?id=60134791&iu=/4140/ostg.clktrk
_______________________________________________
audacity-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-devel
Reply | Threaded
Open this post in threaded view
|

Re: possible recording bug in 2.0.5

Gale
Administrator
Peter,

Could you please try audacity-win-r12584-2.0.5-alpha-22-sep-13.zip
which doesn't include the recent PortAudio update.

We can take it from there.


Gale

 
Reply | Threaded
Open this post in threaded view
|

Re: possible recording bug in 2.0.5

Peter Sampson
Ok with audacity-win-r12584-2.0.5-alpha-22-sep-13.zip
I get good clean recordings with MME, WDS and WDM-KS
 
Peter.
 
Peter Sampson
Tel: +44 (0)1625 524 780
Mob: +44 (0)7732 278 299

From: Gale (Audacity Team) <[hidden email]>
To: [hidden email]
Sent: Saturday, October 5, 2013 3:53 PM
Subject: Re: [Audacity-devel] possible recording bug in 2.0.5

Peter,

Could you please try audacity-win-r12584-2.0.5-alpha-22-sep-13.zip
which doesn't include the recent PortAudio update.

We can take it from there.


Gale





--
View this message in context: http://audacity.238276.n2.nabble.com/possible-recording-bug-in-2-0-5-tp7559811p7559812.html
Sent from the audacity-devel mailing list archive at Nabble.com.


------------------------------------------------------------------------------
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application performance.
Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
the latest Intel processors and coprocessors. See abstracts and register >
http://pubads.g.doubleclick.net/gampad/clk?id=60134791&iu=/4140/ostg.clktrk

_______________________________________________
audacity-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-devel




------------------------------------------------------------------------------
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application performance.
Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
the latest Intel processors and coprocessors. See abstracts and register >
http://pubads.g.doubleclick.net/gampad/clk?id=60134791&iu=/4140/ostg.clktrk
_______________________________________________
audacity-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-devel
Reply | Threaded
Open this post in threaded view
|

Re: possible recording bug in 2.0.5

Leland (Audacity Team)
On 10/5/2013 10:53 AM, Peter Sampson wrote:
Ok with audacity-win-r12584-2.0.5-alpha-22-sep-13.zip
I get good clean recordings with MME, WDS and WDM-KS

Is there an r12591 and an r12592 available?  If so, could you try them as well?

Thanks much,

Leland


------------------------------------------------------------------------------
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application performance.
Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
the latest Intel processors and coprocessors. See abstracts and register >
http://pubads.g.doubleclick.net/gampad/clk?id=60134791&iu=/4140/ostg.clktrk
_______________________________________________
audacity-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-devel
Reply | Threaded
Open this post in threaded view
|

Re: possible recording bug in 2.0.5

Lasse Steen Bohnstedt
Am I correct in reading this as you having only problems with recording from youtube?
I used 2.0.3 to record from youtube some time ago and found that everything had skips. It seems they changed their buffering some months ago in an attempt to boost performance and it still isn't working as well as it did before that change.
// Lasse

"There is a crack in everything, that's how the light gets in." - L. Cohen


On Sat, Oct 5, 2013 at 9:33 PM, Leland <[hidden email]> wrote:
On 10/5/2013 10:53 AM, Peter Sampson wrote:
Ok with audacity-win-r12584-2.0.5-alpha-22-sep-13.zip
I get good clean recordings with MME, WDS and WDM-KS

Is there an r12591 and an r12592 available?  If so, could you try them as well?

Thanks much,

Leland


------------------------------------------------------------------------------
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application performance.
Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
the latest Intel processors and coprocessors. See abstracts and register >
http://pubads.g.doubleclick.net/gampad/clk?id=60134791&iu=/4140/ostg.clktrk
_______________________________________________
audacity-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-devel



------------------------------------------------------------------------------
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application performance.
Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
the latest Intel processors and coprocessors. See abstracts and register >
http://pubads.g.doubleclick.net/gampad/clk?id=60134791&iu=/4140/ostg.clktrk
_______________________________________________
audacity-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-devel
Reply | Threaded
Open this post in threaded view
|

Re: possible recording bug in 2.0.5

Gale
Administrator
In reply to this post by Leland (Audacity Team)
Leland wrote:
On 10/5/2013 10:53 AM, Peter Sampson wrote:
> > Ok with audacity-win-r12584-2.0.5-alpha-22-sep-13.zip
> > I get good clean recordings with MME, WDS and WDM-KS
>
> Is there an r12591 and an r12592 available?  If so, could you try
> them as well?
r12591 should not have any relevant change in comparison to 12584
unless you think http://code.google.com/p/audacity/source/detail?r=12591
could be implicated?

I've made a r12591:
http://gaclrecords.org.uk/test-builds/audacity-win-r12591.zip .

I won't be able to do r12592 until this afternoon, but feel free if
someone else can do so.  



Gale    
Reply | Threaded
Open this post in threaded view
|

Re: possible recording bug in 2.0.5

Leland (Audacity Team)
On 10/6/2013 2:12 AM, Gale (Audacity Team) wrote:

> Leland wrote:
> On 10/5/2013 10:53 AM, Peter Sampson wrote:
>>> Ok with audacity-win-r12584-2.0.5-alpha-22-sep-13.zip
>>> I get good clean recordings with MME, WDS and WDM-KS
>>
>> Is there an r12591 and an r12592 available?  If so, could you try
>> them as well?
> r12591 should not have any relevant change in comparison to 12584
> unless you think http://code.google.com/p/audacity/source/detail?r=12591
> could be implicated?
>
Exactly.

Leland

------------------------------------------------------------------------------
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application performance.
Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
the latest Intel processors and coprocessors. See abstracts and register >
http://pubads.g.doubleclick.net/gampad/clk?id=60134791&iu=/4140/ostg.clktrk
_______________________________________________
audacity-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-devel
Reply | Threaded
Open this post in threaded view
|

Re: possible recording bug in 2.0.5

Peter Sampson
In reply to this post by Lasse Steen Bohnstedt
No, not correct, sorry
 
My report (the long version)stated that I had also tried recording:
a) from playing an iTunes AAC file ripped from CD
b) FM radio output to Edirol UA-1EX to USB on PC
 
They also exhibited the same symptoms.
 
And when I tried the same YT video with 2.0.4 no skips occurred.
 
Peter
 
Peter Sampson
Tel: +44 (0)1625 524 780
Mob: +44 (0)7732 278 299

From: Lasse Steen Bohnstedt <[hidden email]>
To: devel Audacity <[hidden email]>
Sent: Saturday, October 5, 2013 10:50 PM
Subject: Re: [Audacity-devel] possible recording bug in 2.0.5

Am I correct in reading this as you having only problems with recording from youtube?
I used 2.0.3 to record from youtube some time ago and found that everything had skips. It seems they changed their buffering some months ago in an attempt to boost performance and it still isn't working as well as it did before that change.
// Lasse

"There is a crack in everything, that's how the light gets in." - L. Cohen


On Sat, Oct 5, 2013 at 9:33 PM, Leland <[hidden email]> wrote:
On 10/5/2013 10:53 AM, Peter Sampson wrote:
Ok with audacity-win-r12584-2.0.5-alpha-22-sep-13.zip
I get good clean recordings with MME, WDS and WDM-KS

Is there an r12591 and an r12592 available?  If so, could you try them as well?

Thanks much,

Leland


------------------------------------------------------------------------------
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application performance.
Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
the latest Intel processors and coprocessors. See abstracts and register >
http://pubads.g.doubleclick.net/gampad/clk?id=60134791&iu=/4140/ostg.clktrk
_______________________________________________
audacity-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-devel



------------------------------------------------------------------------------
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application performance.
Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
the latest Intel processors and coprocessors. See abstracts and register >
http://pubads.g.doubleclick.net/gampad/clk?id=60134791&iu=/4140/ostg.clktrk

_______________________________________________
audacity-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-devel



------------------------------------------------------------------------------
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application performance.
Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
the latest Intel processors and coprocessors. See abstracts and register >
http://pubads.g.doubleclick.net/gampad/clk?id=60134071&iu=/4140/ostg.clktrk
_______________________________________________
audacity-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-devel
Reply | Threaded
Open this post in threaded view
|

Re: possible recording bug in 2.0.5

Peter Sampson
In reply to this post by Gale
1) Tested on r12591 using iTunes player for input.
 
I still get clicks - and they seem to occur consistently at
just around 5.5 second intervals.
 
-------------------------------------------------------------------
 
 
2) Also testing with FMradio=>Edirol UA-1EX=>USB
(to minimize loading on the PC)
 
Still get clicks here too - intervals here are:
11, 16.5, 5.5 seconds
i.e. all suspiciously multiples of 5.5 seconds.
 
Cheers,
Peter.
 
Peter Sampson
Tel: +44 (0)1625 524 780
Mob: +44 (0)7732 278 299

From: Gale (Audacity Team) <[hidden email]>
To: [hidden email]
Sent: Sunday, October 6, 2013 8:12 AM
Subject: Re: [Audacity-devel] possible recording bug in 2.0.5

Leland wrote:

On 10/5/2013 10:53 AM, Peter Sampson wrote:
> > Ok with audacity-win-r12584-2.0.5-alpha-22-sep-13.zip
> > I get good clean recordings with MME, WDS and WDM-KS
>
> Is there an r12591 and an r12592 available?  If so, could you try
> them as well?

r12591 should not have any relevant change in comparison to 12584
unless you think http://code.google.com/p/audacity/source/detail?r=12591
could be implicated?

I've made a r12591:
http://gaclrecords.org.uk/test-builds/audacity-win-r12591.zip.

I won't be able to do r12592 until this afternoon, but feel free if
someone else can do so. 



Gale   



--
View this message in context: http://audacity.238276.n2.nabble.com/possible-recording-bug-in-2-0-5-tp7559811p7559818.html
Sent from the audacity-devel mailing list archive at Nabble.com.

------------------------------------------------------------------------------
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application performance.
Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
the latest Intel processors and coprocessors. See abstracts and register >
http://pubads.g.doubleclick.net/gampad/clk?id=60134791&iu=/4140/ostg.clktrk
_______________________________________________
audacity-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-devel




------------------------------------------------------------------------------
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application performance.
Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
the latest Intel processors and coprocessors. See abstracts and register >
http://pubads.g.doubleclick.net/gampad/clk?id=60134071&iu=/4140/ostg.clktrk
_______________________________________________
audacity-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-devel
Reply | Threaded
Open this post in threaded view
|

Re: possible recording bug in 2.0.5

Gale
Administrator
In reply to this post by Leland (Audacity Team)
Here's r12592 (that is, 12591 plus updated to latest portaudio):
http://gaclrecords.org.uk/test-builds/audacity-win-r12592.zip .

Please test. I have not found a problem recording in HEAD myself.
Peter, were you getting clicky recordings with the built-in sound card,
or your USB interfaces, or both?  


Gale
Reply | Threaded
Open this post in threaded view
|

Re: possible recording bug in 2.0.5

Peter Sampson
Hi Gale,
 
I still get the skips/clicks/bumps with r12592
both with iTunes feeding my internal soundacrd
and with FM tuner feeding my Edirol USB 1-EX.
The interrupts appear at almost 5.5 second intervals.
 
Very similar happened with r12591:
with iTunes/onboard card 5.5 second interval clicks,
with FM/Edirol - clicks at multiples of 5.5 seconds
but not every 5.5 seconds.
 
I did send a response email on r12591 testing earlier
today via -devel, but that never appeared in my inbox.
I will resend (for completeness).
 
Leland has prepared a couple of special builds to try to
isolate the problem (off-list) but I can't download from
his links - I am waiting to hear back from him.
 
Peter
 
 
Peter Sampson
Tel: +44 (0)1625 524 780
Mob: +44 (0)7732 278 299

From: Gale (Audacity Team) <[hidden email]>
To: [hidden email]
Sent: Sunday, October 6, 2013 4:21 PM
Subject: Re: [Audacity-devel] possible recording bug in 2.0.5

Here's r12592 (that is, 12591 plus updated to latest portaudio):
http://gaclrecords.org.uk/test-builds/audacity-win-r12592.zip.

Please test. I have not found a problem recording in HEAD myself.
Peter, were you getting clicky recordings with the built-in sound card,
or your USB interfaces, or both? 


Gale



--
View this message in context: http://audacity.238276.n2.nabble.com/possible-recording-bug-in-2-0-5-tp7559811p7559822.html
Sent from the audacity-devel mailing list archive at Nabble.com.


------------------------------------------------------------------------------
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application performance.
Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
the latest Intel processors and coprocessors. See abstracts and register >
http://pubads.g.doubleclick.net/gampad/clk?id=60134791&iu=/4140/ostg.clktrk
_______________________________________________
audacity-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-devel



------------------------------------------------------------------------------
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application performance.
Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
the latest Intel processors and coprocessors. See abstracts and register >
http://pubads.g.doubleclick.net/gampad/clk?id=60134791&iu=/4140/ostg.clktrk
_______________________________________________
audacity-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-devel
Reply | Threaded
Open this post in threaded view
|

Fw: possible recording bug in 2.0.5

Peter Sampson
In reply to this post by Peter Sampson
Resend - as this appeared not to get delivered.
 
Peter.
 
Peter Sampson
Tel: +44 (0)1625 524 780
Mob: +44 (0)7732 278 299

----- Forwarded Message -----
From: Peter Sampson <[hidden email]>
To: "[hidden email]" <[hidden email]>
Sent: Sunday, October 6, 2013 10:42 AM
Subject: Re: [Audacity-devel] possible recording bug in 2.0.5

1) Tested on r12591 using iTunes player for input.
 
I still get clicks - and they seem to occur consistently at
just around 5.5 second intervals.
 
-------------------------------------------------------------------
 
 
2) Also testing with FMradio=>Edirol UA-1EX=>USB
(to minimize loading on the PC)
 
Still get clicks here too - intervals here are:
11, 16.5, 5.5 seconds
i.e. all suspiciously multiples of 5.5 seconds.
 
Cheers,
Peter.
 
Peter Sampson
Tel: +44 (0)1625 524 780
Mob: +44 (0)7732 278 299

From: Gale (Audacity Team) <[hidden email]>
To: [hidden email]
Sent: Sunday, October 6, 2013 8:12 AM
Subject: Re: [Audacity-devel] possible recording bug in 2.0.5

Leland wrote:

On 10/5/2013 10:53 AM, Peter Sampson wrote:
> > Ok with audacity-win-r12584-2.0.5-alpha-22-sep-13.zip
> > I get good clean recordings with MME, WDS and WDM-KS
>
> Is there an r12591 and an r12592 available?  If so, could you try
> them as well?

r12591 should not have any relevant change in comparison to 12584
unless you think http://code.google.com/p/audacity/source/detail?r=12591
could be implicated?

I've made a r12591:
http://gaclrecords.org.uk/test-builds/audacity-win-r12591.zip.

I won't be able to do r12592 until this afternoon, but feel free if
someone else can do so. 



Gale   



--
View this message in context: http://audacity.238276.n2.nabble.com/possible-recording-bug-in-2-0-5-tp7559811p7559818.html
Sent from the audacity-devel mailing list archive at Nabble.com.

------------------------------------------------------------------------------
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application performance.
Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
the latest Intel processors and coprocessors. See abstracts and register >
http://pubads.g.doubleclick.net/gampad/clk?id=60134791&iu=/4140/ostg.clktrk
_______________________________________________
audacity-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-devel






------------------------------------------------------------------------------
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application performance.
Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
the latest Intel processors and coprocessors. See abstracts and register >
http://pubads.g.doubleclick.net/gampad/clk?id=60134791&iu=/4140/ostg.clktrk
_______________________________________________
audacity-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-devel
Reply | Threaded
Open this post in threaded view
|

Re: possible recording bug in 2.0.5

MartynShaw
In reply to this post by Peter Sampson
Hi

I take it that this is the exact same bug that Leland reported on
http://audacity.238276.n2.nabble.com/Interesting-critter-when-recording-td7559588.html
on 23rd Sept.

I also see that Leland has backed out r12591 from around that date.

I never did see the problem.  Is it fixed again?

Martyn

On 06/10/2013 16:57, Peter Sampson wrote:

> Hi Gale,
> I still get the skips/clicks/bumps with r12592
> both with iTunes feeding my internal soundacrd
> and with FM tuner feeding my Edirol USB 1-EX.
> The interrupts appear at almost 5.5 second intervals.
> Very similar happened with r12591:
> with iTunes/onboard card 5.5 second interval clicks,
> with FM/Edirol - clicks at multiples of 5.5 seconds
> but not every 5.5 seconds.
> I did send a response email on r12591 testing earlier
> today via -devel, but that never appeared in my inbox.
> I will resend (for completeness).
> Leland has prepared a couple of special builds to try to
> isolate the problem (off-list) but I can't download from
> his links - I am waiting to hear back from him.
> Peter
> Peter Sampson
> Tel: +44 (0)1625 524 780
> Mob: +44 (0)7732 278 299
>
> *From:* Gale (Audacity Team) <[hidden email]>
> *To:* [hidden email]
> *Sent:* Sunday, October 6, 2013 4:21 PM
> *Subject:* Re: [Audacity-devel] possible recording bug in 2.0.5
>
> Here's r12592 (that is, 12591 plus updated to latest portaudio):
> http://gaclrecords.org.uk/test-builds/audacity-win-r12592.zip.
>
> Please test. I have not found a problem recording in HEAD myself.
> Peter, were you getting clicky recordings with the built-in sound card,
> or your USB interfaces, or both?
>
>
> Gale
>
>
>
> --
> View this message in context:
> http://audacity.238276.n2.nabble.com/possible-recording-bug-in-2-0-5-tp7559811p7559822.html
> Sent from the audacity-devel mailing list archive at Nabble.com.
>
>
> ------------------------------------------------------------------------------
> October Webinars: Code for Performance
> Free Intel webinars can help you accelerate application performance.
> Explore tips for MPI, OpenMP, advanced profiling, and more. Get the
> most from
> the latest Intel processors and coprocessors. See abstracts and register >
> http://pubads.g.doubleclick.net/gampad/clk?id=60134791&iu=/4140/ostg.clktrk
> _______________________________________________
> audacity-devel mailing list
> [hidden email]
> <mailto:[hidden email]>
> https://lists.sourceforge.net/lists/listinfo/audacity-devel
>
>
>
>
> ------------------------------------------------------------------------------
> October Webinars: Code for Performance
> Free Intel webinars can help you accelerate application performance.
> Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
> the latest Intel processors and coprocessors. See abstracts and register >
> http://pubads.g.doubleclick.net/gampad/clk?id=60134791&iu=/4140/ostg.clktrk
>
>
>
> _______________________________________________
> audacity-devel mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/audacity-devel
>

------------------------------------------------------------------------------
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application performance.
Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
the latest Intel processors and coprocessors. See abstracts and register >
http://pubads.g.doubleclick.net/gampad/clk?id=60134071&iu=/4140/ostg.clktrk
_______________________________________________
audacity-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-devel
Reply | Threaded
Open this post in threaded view
|

Re: possible recording bug in 2.0.5

Leland (Audacity Team)
On 10/7/2013 5:51 PM, Martyn Shaw wrote:
> Hi
>
> I take it that this is the exact same bug that Leland reported on
> http://audacity.238276.n2.nabble.com/Interesting-critter-when-recording-td7559588.html
> on 23rd Sept.

Not really sure about that, but very possible.

>
> I also see that Leland has backed out r12591 from around that date.
>
> I never did see the problem.  Is it fixed again?
>

Yes, it is fixed.  The change needed a bit more work and I thought it
safer just to remove it.

Leland


------------------------------------------------------------------------------
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application performance.
Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
the latest Intel processors and coprocessors. See abstracts and register >
http://pubads.g.doubleclick.net/gampad/clk?id=60134071&iu=/4140/ostg.clktrk
_______________________________________________
audacity-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-devel
Reply | Threaded
Open this post in threaded view
|

Re: possible recording bug in 2.0.5

Peter Sampson
In reply to this post by MartynShaw
Martyn,
 
I'm not convinced that this is "exactly" the same bug that Leland
reported.  From looking at the nabble thread the interrupts there
are massive and easily visible - I could only find mine with careful
listening and then using the spectrogram view on the approximate
location - otherwise they were fairly invisible on the waveform
display until zoomed right in.
 
And for me it occurs not just on WASAPI but also on MME & WDS.
 
I will be testing again on the next nightly that contains Leland's
backing out of r12591 (presumably tonight's)
 
Peter.
 
Peter Sampson
Tel: +44 (0)1625 524 780
Mob: +44 (0)7732 278 299

From: Martyn Shaw <[hidden email]>
To: [hidden email]
Cc: Peter Sampson <[hidden email]>; gale andrews <[hidden email]>
Sent: Monday, October 7, 2013 11:51 PM
Subject: Re: [Audacity-devel] possible recording bug in 2.0.5

Hi

I take it that this is the exact same bug that Leland reported on
http://audacity.238276.n2.nabble.com/Interesting-critter-when-recording-td7559588.html
on 23rd Sept.

I also see that Leland has backed out r12591 from around that date.

I never did see the problem.  Is it fixed again?

Martyn

On 06/10/2013 16:57, Peter Sampson wrote:

> Hi Gale,
> I still get the skips/clicks/bumps with r12592
> both with iTunes feeding my internal soundacrd
> and with FM tuner feeding my Edirol USB 1-EX.
> The interrupts appear at almost 5.5 second intervals.
> Very similar happened with r12591:
> with iTunes/onboard card 5.5 second interval clicks,
> with FM/Edirol - clicks at multiples of 5.5 seconds
> but not every 5.5 seconds.
> I did send a response email on r12591 testing earlier
> today via -devel, but that never appeared in my inbox.
> I will resend (for completeness).
> Leland has prepared a couple of special builds to try to
> isolate the problem (off-list) but I can't download from
> his links - I am waiting to hear back from him.
> Peter
> Peter Sampson
> Tel: +44 (0)1625 524 780
> Mob: +44 (0)7732 278 299
>
> *From:* Gale (Audacity Team) <[hidden email]>
> *To:* [hidden email]
> *Sent:* Sunday, October 6, 2013 4:21 PM
> *Subject:* Re: [Audacity-devel] possible recording bug in 2.0.5
>
> Here's r12592 (that is, 12591 plus updated to latest portaudio):
> http://gaclrecords.org.uk/test-builds/audacity-win-r12592.zip.
>
> Please test. I have not found a problem recording in HEAD myself.
> Peter, were you getting clicky recordings with the built-in sound card,
> or your USB interfaces, or both?
>
>
> Gale
>
>
>
> --
> View this message in context:
> http://audacity.238276.n2.nabble.com/possible-recording-bug-in-2-0-5-tp7559811p7559822.html
> Sent from the audacity-devel mailing list archive at Nabble.com.
>
>
> ------------------------------------------------------------------------------
> October Webinars: Code for Performance
> Free Intel webinars can help you accelerate application performance.
> Explore tips for MPI, OpenMP, advanced profiling, and more. Get the
> most from
> the latest Intel processors and coprocessors. See abstracts and register >
> http://pubads.g.doubleclick.net/gampad/clk?id=60134791&iu=/4140/ostg.clktrk
> _______________________________________________
> audacity-devel mailing list
> [hidden email]
> <mailto:[hidden email]>

> https://lists.sourceforge.net/lists/listinfo/audacity-devel
>
>
>
>
> ------------------------------------------------------------------------------
> October Webinars: Code for Performance
> Free Intel webinars can help you accelerate application performance.
> Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
> the latest Intel processors and coprocessors. See abstracts and register >
> http://pubads.g.doubleclick.net/gampad/clk?id=60134791&iu=/4140/ostg.clktrk
>
>
>
> _______________________________________________
> audacity-devel mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/audacity-devel
>



------------------------------------------------------------------------------
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application performance.
Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
the latest Intel processors and coprocessors. See abstracts and register >
http://pubads.g.doubleclick.net/gampad/clk?id=60134071&iu=/4140/ostg.clktrk
_______________________________________________
audacity-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-devel
Reply | Threaded
Open this post in threaded view
|

Re: possible recording bug in 2.0.5

MartynShaw
In reply to this post by Leland (Audacity Team)
Thanks for the confirmation of the fix!

On 08/10/2013 00:25, Leland wrote:

> On 10/7/2013 5:51 PM, Martyn Shaw wrote:
>> Hi
>>
>> I take it that this is the exact same bug that Leland reported on
>> http://audacity.238276.n2.nabble.com/Interesting-critter-when-recording-td7559588.html
>> on 23rd Sept.
>
> Not really sure about that, but very possible.
>
>>
>> I also see that Leland has backed out r12591 from around that date.
>>
>> I never did see the problem.  Is it fixed again?
>>
>
> Yes, it is fixed.  The change needed a bit more work and I thought it
> safer just to remove it.

That stuff is 'blood and guts' to Audacity, it certainly needs some
careful thought, stuff which I can't find much time for at the moment.
  I thought you were certainly on the right track there though with
stride!

TTFN
Martyn

> Leland
>
>
> ------------------------------------------------------------------------------
> October Webinars: Code for Performance
> Free Intel webinars can help you accelerate application performance.
> Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
> the latest Intel processors and coprocessors. See abstracts and register >
> http://pubads.g.doubleclick.net/gampad/clk?id=60134071&iu=/4140/ostg.clktrk
> _______________________________________________
> audacity-devel mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/audacity-devel
>

------------------------------------------------------------------------------
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application performance.
Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
the latest Intel processors and coprocessors. See abstracts and register >
http://pubads.g.doubleclick.net/gampad/clk?id=60134071&iu=/4140/ostg.clktrk
_______________________________________________
audacity-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-devel
Reply | Threaded
Open this post in threaded view
|

Re: possible recording bug in 2.0.5

Vaughan Johnson
Administrator
On 10/7/2013 4:31 PM, Martyn Shaw wrote:
> Thanks for the confirmation of the fix!

+1.


>
> On 08/10/2013 00:25, Leland wrote:
>> On 10/7/2013 5:51 PM, Martyn Shaw wrote:
>>> Hi
>>>
>>> I take it that this is the exact same bug that Leland reported on
>>> http://audacity.238276.n2.nabble.com/Interesting-critter-when-recording-td7559588.html
>>> on 23rd Sept.
>>
>> Not really sure about that, but very possible.
>>
>>>
>>> I also see that Leland has backed out r12591 from around that date.
>>>
>>> I never did see the problem.  Is it fixed again?
>>>
>>
>> Yes, it is fixed.  The change needed a bit more work and I thought it
>> safer just to remove it.
>
> That stuff is 'blood and guts' to Audacity, it certainly needs some
> careful thought, stuff which I can't find much time for at the moment.

Oh yes, amen to that. Good to remove it for 2.0.5!


>    I thought you were certainly on the right track there though with
> stride!

Sounds right to me. Gaps => stride.

- V

------------------------------------------------------------------------------
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application performance.
Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
the latest Intel processors and coprocessors. See abstracts and register >
http://pubads.g.doubleclick.net/gampad/clk?id=60134071&iu=/4140/ostg.clktrk
_______________________________________________
audacity-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-devel
Reply | Threaded
Open this post in threaded view
|

Re: possible recording bug in 2.0.5

Peter Sampson
In reply to this post by Peter Sampson
Testing on alpha win r12658 8Oct13
with Leland's r12561 reverting r12591
on Windows-7 Tosh laptop
 
I can confirm that this records cleanly now without the
5.5 seconds apart clicks/pops/thumps.
 
Tested:
WASAPI loopback source iTunes (onboard soundcard)
MME stereo mix source iTunes (onboard soundcard)
WDS stereo mix source iTunes (onboard soundcard)
WASAPI loopback source YouTube (onboard soundcard)
WDS FM radio => Edirol UA-1EX (external USB soundcard)
MME FM radio => Edirol UA-1EX (external USB soundcard)
 
all yielded nice clean recordings.
 
Peter
 
Peter Sampson
Tel: +44 (0)1625 524 780
Mob: +44 (0)7732 278 299

From: Peter Sampson <[hidden email]>
To: Martyn Shaw <[hidden email]>; "[hidden email]" <[hidden email]>
Sent: Tuesday, October 8, 2013 12:28 AM
Subject: Re: [Audacity-devel] possible recording bug in 2.0.5

Martyn,
 
I'm not convinced that this is "exactly" the same bug that Leland
reported.  From looking at the nabble thread the interrupts there
are massive and easily visible - I could only find mine with careful
listening and then using the spectrogram view on the approximate
location - otherwise they were fairly invisible on the waveform
display until zoomed right in.
 
And for me it occurs not just on WASAPI but also on MME & WDS.
 
I will be testing again on the next nightly that contains Leland's
backing out of r12591 (presumably tonight's)
 
Peter.
 
Peter Sampson
Tel: +44 (0)1625 524 780
Mob: +44 (0)7732 278 299

From: Martyn Shaw <[hidden email]>
To: [hidden email]
Cc: Peter Sampson <[hidden email]>; gale andrews <[hidden email]>
Sent: Monday, October 7, 2013 11:51 PM
Subject: Re: [Audacity-devel] possible recording bug in 2.0.5

Hi

I take it that this is the exact same bug that Leland reported on
http://audacity.238276.n2.nabble.com/Interesting-critter-when-recording-td7559588.html
on 23rd Sept.

I also see that Leland has backed out r12591 from around that date.

I never did see the problem.  Is it fixed again?

Martyn

On 06/10/2013 16:57, Peter Sampson wrote:

> Hi Gale,
> I still get the skips/clicks/bumps with r12592
> both with iTunes feeding my internal soundacrd
> and with FM tuner feeding my Edirol USB 1-EX.
> The interrupts appear at almost 5.5 second intervals.
> Very similar happened with r12591:
> with iTunes/onboard card 5.5 second interval clicks,
> with FM/Edirol - clicks at multiples of 5.5 seconds
> but not every 5.5 seconds.
> I did send a response email on r12591 testing earlier
> today via -devel, but that never appeared in my inbox.
> I will resend (for completeness).
> Leland has prepared a couple of special builds to try to
> isolate the problem (off-list) but I can't download from
> his links - I am waiting to hear back from him.
> Peter
> Peter Sampson
> Tel: +44 (0)1625 524 780
> Mob: +44 (0)7732 278 299
>
> *From:* Gale (Audacity Team) <[hidden email]>
> *To:* [hidden email]
> *Sent:* Sunday, October 6, 2013 4:21 PM
> *Subject:* Re: [Audacity-devel] possible recording bug in 2.0.5
>
> Here's r12592 (that is, 12591 plus updated to latest portaudio):
> http://gaclrecords.org.uk/test-builds/audacity-win-r12592.zip.
>
> Please test. I have not found a problem recording in HEAD myself.
> Peter, were you getting clicky recordings with the built-in sound card,
> or your USB interfaces, or both?
>
>
> Gale
>
>
>
> --
> View this message in context:
> http://audacity.238276.n2.nabble.com/possible-recording-bug-in-2-0-5-tp7559811p7559822.html
> Sent from the audacity-devel mailing list archive at Nabble.com.
>
>
> ------------------------------------------------------------------------------
> October Webinars: Code for Performance
> Free Intel webinars can help you accelerate application performance.
> Explore tips for MPI, OpenMP, advanced profiling, and more. Get the
> most from
> the latest Intel processors and coprocessors. See abstracts and register >
> http://pubads.g.doubleclick.net/gampad/clk?id=60134791&iu=/4140/ostg.clktrk
> _______________________________________________
> audacity-devel mailing list
> [hidden email]
> <mailto:[hidden email]>

> https://lists.sourceforge.net/lists/listinfo/audacity-devel
>
>
>
>
> ------------------------------------------------------------------------------
> October Webinars: Code for Performance
> Free Intel webinars can help you accelerate application performance.
> Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
> the latest Intel processors and coprocessors. See abstracts and register >
> http://pubads.g.doubleclick.net/gampad/clk?id=60134791&iu=/4140/ostg.clktrk
>
>
>
> _______________________________________________
> audacity-devel mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/audacity-devel
>



------------------------------------------------------------------------------
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application performance.
Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
the latest Intel processors and coprocessors. See abstracts and register >
http://pubads.g.doubleclick.net/gampad/clk?id=60134071&iu=/4140/ostg.clktrk

_______________________________________________
audacity-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-devel



------------------------------------------------------------------------------
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application performance.
Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
the latest Intel processors and coprocessors. See abstracts and register >
http://pubads.g.doubleclick.net/gampad/clk?id=60134071&iu=/4140/ostg.clktrk
_______________________________________________
audacity-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-devel
Reply | Threaded
Open this post in threaded view
|

Re: possible recording bug in 2.0.5

Gale
Administrator
Peter wrote:
> Testing on alpha win r12658 8Oct13 with Leland's r12561
> reverting r12591 on Windows-7 Tosh laptop
>
> I can confirm that this records cleanly now without the
> 5.5 seconds apart clicks/pops/thumps.
Thanks for testing.

And apart from the fact that you got the clicks under MME as
well, do I assume your project format was 32-bit float?

If so, what happens if you set Default Sample Format to 16-bit,
project rate to 44100 Hz and record with r12658 08Oct13
under WASAPI?

I listened to two three-minute WASAPI loopback recordings I made
prior to revert of r12591. These recordings were 44100 Hz and
happened to be made at 16-bit sample format because I had not
gone back to 32-bit after some other test I had been doing. They
were exported as 16-bit WAV.

I heard no light ticks every 5.5 seconds.

But I did get just one loud click in each recording that was not in
the original source. Zoomed in, those look like the glitches in
Leland's images, but nothing like as wide as that.    



Gale
Reply | Threaded
Open this post in threaded view
|

Re: possible recording bug in 2.0.5

Peter Sampson
Gale wrote:
>And apart from the fact that you got the clicks under MME as
>well, do I assume your project format was 32-bit float?
Yup 44.1kHz 32-bit float - with r12658 08Oct13
 
And NO clicks with WASAPI, MME or WDS with onbrad card
NOR when using the Edirol with MME or WDS
 
i.e.the latest nightly records clean for me at default quality settings.
 
 
Gale wrote:
>If so, what happens if you set Default Sample Format to 16-bit,
>project rate to 44100 Hz and record with r12658 08Oct13
>under WASAPI?
 
With 44.1kHz 16-bit under WASAPI loopback I still get NO clicks
but I would kind of expect that as it is writing a lot less data.
 
 
So clean WASAPI recordings at both 16-bit and 32-bit
with r12658 08Oct13
 
Peter
 
Peter Sampson
Tel: +44 (0)1625 524 780
Mob: +44 (0)7732 278 299

From: Gale (Audacity Team) <[hidden email]>
To: [hidden email]
Sent: Tuesday, October 8, 2013 5:14 PM
Subject: Re: [Audacity-devel] possible recording bug in 2.0.5

Peter wrote:
> Testing on alpha win r12658 8Oct13 with Leland's r12561
> reverting r12591 on Windows-7 Tosh laptop
>
> I can confirm that this records cleanly now without the
> 5.5 seconds apart clicks/pops/thumps.
Thanks for testing.

And apart from the fact that you got the clicks under MME as
well, do I assume your project format was 32-bit float?

If so, what happens if you set Default Sample Format to 16-bit,
project rate to 44100 Hz and record with r12658 08Oct13
under WASAPI?

I listened to two three-minute WASAPI loopback recordings I made
prior to revert of r12591. These recordings were 44100 Hz and
happened to be made at 16-bit sample format because I had not
gone back to 32-bit after some other test I had been doing. They
were exported as 16-bit WAV.

I heard no light ticks every 5.5 seconds.

But I did get just one loud click in each recording that was not in
the original source. Zoomed in, those look like the glitches in
Leland's images, but nothing like as wide as that.   



Gale



--
View this message in context: http://audacity.238276.n2.nabble.com/possible-recording-bug-in-2-0-5-tp7559811p7559853.html

Sent from the audacity-devel mailing list archive at Nabble.com.

------------------------------------------------------------------------------
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application performance.
Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
the latest Intel processors and coprocessors. See abstracts and register >
http://pubads.g.doubleclick.net/gampad/clk?id=60134071&iu=/4140/ostg.clktrk
_______________________________________________
audacity-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-devel



------------------------------------------------------------------------------
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application performance.
Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
the latest Intel processors and coprocessors. See abstracts and register >
http://pubads.g.doubleclick.net/gampad/clk?id=60134071&iu=/4140/ostg.clktrk
_______________________________________________
audacity-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-devel