brief audio dropouts

[James Lo]

I'm getting <0.5 sec, randomly distributed audio dropouts, max 3 in a 35 min show, sometimes none.  The dropouts sound "smooth", i.e. without any pops or glitches, as if I grabbed the fader on the mixer and dipped it briefly.  I'm running OS X 10.8.4 on a mid 2012 MacBook Pro, QLab 3.0.6 beta, and have an AudioFire4 with the latest console software (which admittedly is not so recent in the scheme of things).  In Console I can see QLab reporting dropped frames.  Here is the first 3 mins of my dress rehearsal, during which I heard one dropout, filenames deleted:

6/18/13 8:06:06.377 PM QLab[218]: Warning: dropped 478 frames (0.01 seconds of audio) : 

6/18/13 8:06:07.387 PM QLab[218]: Warning: dropped 7168 frames (0.16 seconds of audio) :

6/18/13 8:06:49.428 PM QLab[218]: Warning: dropped 512 frames (0.01 seconds of audio) :

6/18/13 8:06:50.427 PM QLab[218]: Warning: dropped 17408 frames (0.39 seconds of audio) : 

6/18/13 8:06:50.438 PM QLab[218]: Warning: dropped 20480 frames (0.46 seconds of audio) : 

6/18/13 8:08:35.086 PM QLab[218]: Warning: dropped 16384 frames (0.37 seconds of audio) :

I've disabled the screensaver, computer sleep, display sleep, and hard disk sleeping.  I was going to try turning off wifi but I thought I'd check with this forum first.  I'm running 3.0.6 beta because the current release was aborting fairly regularly on a pair of auto follow relative fades--the beta version hasn't failed at that point yet.  Any advice on how to investigate or address this issue would be appreciated.

[James Lo]

Disabling wifi and bluetooth seems to have gotten rid of the dropped frame warnings, I wonder if that means I'm out of the woods....

[James Lo]

Man, I'm still getting dropped frames!  What else could it be?

[Jeremy Lee]

Wow. They haven't updated their drivers for about a year and a half.  There's no guarantee that it will work on 10.8.  Do you have another audio interface available?

Also, make sure that you have your drives set to not go to sleep.  If you're using external drives, the energy saver prefs might not take care of it.

-- 

Jeremy Lee

    Sound Designer, NYC - USA 829

        http://www.jjlee.com

[Chris Ashworth]

I am currently trying to track down a few sporadic reports of audio glitches like this. This case could be a driver issue, or it could be something on the QLab side.

I'm about to head to the airport but will update the list when I have more information, suggestions, or questions.

C

[James Lo]

Yeah, I'm worried about that too, but I get dropped frame log entries even when I use the built-in output alone.  And yes, my drive is set to never sleep and I'm only using the internal drive.  I just found that iTunesHelper was loading at startup and disabled that, but I'm still getting dropped frames.  Is it reasonable to expect to never get dropped frame log entries?

[Chris Eschweiler]

Have you tried running the media from an external FW800 or Thunderbolt
drive?

I know that it's better to run video from a fast external drive in some
situations...perhaps this applies here, as well?

[Jeremy Lee]

It is expected to not have any dropped frames...

--
--
Change your preferences or unsubscribe here:
http://groups.google.com/group/qlab
 
Follow Figure 53 on Twitter: http://twitter.com/Figure53
 
---
You received this message because you are subscribed to the Google Groups "QLab" group.
To unsubscribe from this group and stop receiving emails from it, send an email to qlab+uns...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

[James Lo]

Alas, I don't own one.  I have a really large but slow USB drive, and while I can't see how that could be better than the internal drive, I will definitely try it when I become desperate enough :)

[Christopher Ashworth]

James,

Would you be able to send me the full console logs via?  You can do so from inside QLab by going to "Help" --> "Contact Support" and checking the box to include console logs.

I'd like to review them to see if any more ideas pop up. It is sometimes important to see the dropout logs in the context of other things that are going on at the same time.

Thanks very much, and apologies again for the trouble.

Best,

Chris

[James Lo]

Thanks Chris, logs are on their way

[James Lo]

Here's some more data on this issue:

I rigged up a MIDI controller to be my GO button so I could watch the Console message window as I stepped through my show.  It seems that when the dropped frames are logged, they appear a second after an audio cue starts, no matter whether the cue is started by GO or by an auto-follow/continue.  Not all audio cues produce a dropped frame log entry, and of those that do, some seem more likely than others.

And then, if I exit QLab and restart this test again several times, at some point all of the dropped frame messages stop for no apparent reason and I can't get them to start again until I reboot my machine.  Obviously I can't prove that I would never get a dropped frame message sometime in the future, but I've reproduced it 3 or 4 times already.  So maybe this is a temporary workaround?

Finally, if I didn't make it clear before, just because there is a dropped frame log entry doesn't mean that I heard it.  I'm not even sure that there is a clear relationship between the ones that I hear and the reported duration of the drop.

[Nigel Hogg]

If I may join in this discussion, I'm getting a 'glitch' occasionally at the start of a cue. I fire the cue using a midi note on message, and then the first few milliseconds of the cue seem to 'stumble'. Is this a problem with midi triggering, or should I pre load the cue just before the end of the previous one. The problem is random, and doesn't happen to the same cues each time.

I'm also getting occasional dropouts, but will try the latest beta tomorrow to see if that helps. Does anyone have any advice, please?

I'm using an 01v96i over usb as my sound card.

[James Lo]

My "workaround" procedure worked for me tonight!  After line check I just muted the channels and ran the show 6 or 7 times (stopping each time when I saw the first dropped frame message) until I didn't see anymore dropped frame messages, then restarted QLab one last time and let it sit idle until curtain.  I didn't hear any dropouts, all the cues started instantly, and afterward the Console log was clean.  (i've been using 3.0.6 b2)

[Chris Ashworth]

James,

I suspect this may have been luck rather than a fix, but I don't know for sure.

One common thread I'm beginning to notice is the use of MIDI to trigger. 

Does the same problem occur if you use mouse or keyboard?

Could be a red herring, but trying to find a thread to pull on... 

Best,

C

(mobile)

[James Lo]

It doesn't seem to matter how it's triggered (that said, I never use the mouse, only spacebar or MIDI)

[James Lo]

I dunno Chris, my workaround worked again tonight.  Not a single dropped frame.  Did I get lucky twice?

Another observation:  as I kept running the show repeatedly as a part of my workaround, I noticed that each run would get farther and farther along before seeing the first dropped frame.  

And finally, a realization:  I have a few cue files that are large, e.g. 25 mins of 44.1 stereo (but I only play the first couple of minutes).  Could this be the cause?

[Nigel Hogg]

I don't know if it helps, but I had a dropout free show last night. I've adopted a technique of pressing the go button firmly and holding for half  a second rather than a quick jab - that seems to prevent the glitch on triggering. It was slightly annoying to have a clean show, because I've nearly finished building the show in v2 to roll back until the problem is sorted. 

If it helps at all I had an unnerving rehearsal last week when I pressed the midi go button to fire a cue and qlab quit completely! I couldn't send the logs, unfortunately, as that machine isn't on the internet, but it does concentrate the mind when running a show with a full house.............

Thanks for your rapid response to problems, Chris, it really is appreciated!

On Wednesday, 19 June 2013 04:44:13 UTC+1, James Lo wrote:

[Andy Leviss]

On Sun, Jun 23, 2013 at 3:56 AM, Nigel Hogg <nh...@me.com> wrote:
> I don't know if it helps, but I had a dropout free show last night. I've
> adopted a technique of pressing the go button firmly and holding for half a
> second rather than a quick jab - that seems to prevent the glitch on
> triggering. It was slightly annoying to have a clean show, because I've
> nearly finished building the show in v2 to roll back until the problem is
> sorted.

Are you talking about a MIDI go button here? What message is the
button sending on release, if any? And what is your trigger set to in
QLab? My first thought was, "That's crazy, holding down the button
couldn't possibly do that," until I realized that there is a MIDI
message sent on release of many remotes, including ours here at Duck's
Echo, and now I'm wondering if something could be glitching there...

-Andy

[Christopher Ashworth]

On Jun 23, 2013, at 12:53 AM, James Lo <revers...@gmail.com> wrote:

I dunno Chris, my workaround worked again tonight.  Not a single dropped frame.  Did I get lucky twice?

I don't know yet, as I am stumped on this so far.

Another observation:  as I kept running the show repeatedly as a part of my workaround, I noticed that each run would get farther and farther along before seeing the first dropped frame.  

And finally, a realization:  I have a few cue files that are large, e.g. 25 mins of 44.1 stereo (but I only play the first couple of minutes).  Could this be the cause?

Hm. It does lead to a question:

Are you running in edit mode or show mode?

In edit mode, if the inspector is showing the waveform view, that could result in some particularly heavy disk access for long files.

-C

[Alex Lewer]

I'm seeing the same, happening in both edit and show modes. Seems to be on almost every audio cue, and I can consistently reproduce.

24/06/2013 11:28:34.242 QLab[1753]: Warning: audioDeviceDidOverload: Built-in Output

24/06/2013 11:28:35.127 QLab[1753]: Warning: dropped 1024 frames (0.01 seconds of audio) : /Users/Alex/Dropbox/FWIS/FWIS QLab/Score/1m2 QLab Masters/1m2 a v2.wav

24/06/2013 11:28:53.894 QLab[1753]: Warning: audioDeviceDidOverload: Built-in Output

24/06/2013 11:28:54.834 QLab[1753]: Warning: dropped 512 frames (0.01 seconds of audio) : /Users/Alex/Dropbox/FWIS/FWIS QLab/Score/1m4 QLab Masters/1m4 a.wav

24/06/2013 11:29:34.140 QLab[1753]: Warning: audioDeviceDidOverload: Built-in Output

24/06/2013 11:30:08.277 QLab[1753]: Warning: dropped 1536 frames (0.02 seconds of audio) : /Users/Alex/Dropbox/FWIS/FWIS QLab/Score/1m4 QLab Masters/1m4 c v6.wav

24/06/2013 11:30:37.988 QLab[1753]: Warning: dropped 1024 frames (0.01 seconds of audio) : /Users/Alex/Dropbox/FWIS/FWIS QLab/Score/1m4 QLab Masters/1m4 d v3.wav

24/06/2013 11:30:58.839 QLab[1753]: Warning: dropped 1024 frames (0.01 seconds of audio) : /Users/Alex/Dropbox/FWIS/FWIS QLab/Score/1m4 QLab Masters/1m4 e v2.wav

24/06/2013 11:31:48.472 QLab[1753]: Warning: audioDeviceDidOverload: Built-in Output

24/06/2013 11:31:49.449 QLab[1753]: Warning: dropped 512 frames (0.01 seconds of audio) : /Users/Alex/Dropbox/FWIS/FWIS QLab/Score/1m4 QLab Masters/1m4 g.wav

24/06/2013 11:31:58.646 QLab[1753]: Warning: audioDeviceDidOverload: Built-in Output

24/06/2013 11:31:59.591 QLab[1753]: Warning: dropped 1024 frames (0.01 seconds of audio) : /Users/Alex/Dropbox/FWIS/FWIS QLab/Score/1m4 QLab Masters/1m4 h.wav

[Christopher Ashworth]

Hi Alex,

On Jun 24, 2013, at 6:55 AM, Alex Lewer <al...@hitwave.co.uk> wrote:

> I'm seeing the same, happening in both edit and show modes. Seems to be on almost every audio cue, and I can consistently reproduce.

Thanks. Questions:

1) What speed storage medium? (5400 / 7200 / SSD)
2) How are you triggering the workspace?
3) Does it change anything if you manually preload cues?
4) Do you hear a corresponding glitch, or just see the logged frames?

This is quite frustrating. (At least, it is for me, I imagine it is for everyone else too.) This is my top priority to solve right now.

-C

[Alex Lewer]

Hi Chris,

In answer to your questions:

1) What speed storage medium? (5400 / 7200 / SSD)

7200

2) How are you triggering the workspace?

Clicking 'Go' or space bar

3) Does it change anything if you manually preload cues?

Even if I use a LOAD cue to load all the audio files in the sequence, the dropout still occur (and still appear in Console)

4) Do you hear a corresponding glitch, or just see the logged frames? 

Yes, sometimes there is a corresponding glitch in the audio, but not consistently.

--
--
Change your preferences or unsubscribe here:
http://groups.google.com/group/qlab

Follow Figure 53 on Twitter: http://twitter.com/Figure53

---

You received this message because you are subscribed to a topic in the Google Groups "QLab" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/qlab/gsIpvWZU9Ok/unsubscribe.
To unsubscribe from this group and all its topics, send an email to qlab+uns...@googlegroups.com.

[Nigel Hogg]

I'm using a midi remote I made using a midi generator from Highly Liquid - it sends note on messages ( and  note off messages on release). I've set the 'musical midi' section of Qlab to respond to the corresponding messages (using capture). I can only assume that if Qlab gets a note off message too quickly, it glitches.

While this may explain the glitch at the start of the cue, it can surely have no bearing on dropouts later on.

Nigel

[Christopher Ashworth]

An update on this:

I've found and fixed one source of dropped frames. I'll be posting a link to a new beta build for people to try.

One important note about interpreting log messages: the logs are throttled so that QLab isn't spending too much time writing logs to disk if there are a lot of dropped frames. (Which could aggregate the situation and make it worse.) Thus, the moment the log actually appears may not the same time the drop actually occurs; it may show up somewhat later.

I'll provide a little more detail when I put up the link to the new build.

-C

[James Lo]

I've done two quick but incomplete tests this morning with 3.0.6-b3 and it appears to address my issue, thanks.