Digital Playback from Mac to Behringer X32 problems

rob aylestone

Well-known member
i was doing a playback show that required no audio problems, so I decided two MacBooks with QLab and a dual box so ever6 cue would fire off both macs and if anything happened, a hot changeover saves the day. Audio via usb out to the X32. If used this dozens of times for recordings so I know it works fine. One run through. A dance show with lots of kids, but no time for a proper dress rehearsal so during the run through, done in stops and starts, and digital distortion starts to creep in. Not bad at first but getting worse during the playback. Pull out usb, re plug. Next song starts fine then starts to gradually sound worse and worse, so swap to the other Mac which on the analogue out is fine. During a break, I swap them. The usb connection does the same thing. This points to the x32 not the macs. Ran analogue for the show and it was fine, but what happened to the digital connection. Recording has always been fine, but playback not good. I’ll try it on another mixer today if I get the chance but is this a known issue?
 

rob aylestone

Well-known member
I wondered this - especially as unplug-replug reset it for a while, but I've never gone to the X-32 before, only from it? IS it just this X32, or all of them? I'll just have to try and test others.
 

Farview

Well-known member
I had the same sort of problem. It was a long time ago, but it would decide to change/lose the sample rate in the middle of the show. Did it a couple times and I just gave up and ran it analog after that.
 

jamesperrett

Active member
I'd be looking to see which device thinks it is the word clock source. Possibly in playback mode the computer wants to be the clock source but the X32 is continuing to work on its own clock.
 

keith.rogers

Well-known member
I assume the cable was checked/replaced. So, it's PCM data over USB, and there's really no "clock" here, as the audio data should have the information necessary for the X32 to convert it to analog. Either the data is corrupted (a lot) and error correction/retries cannot be done in realtime to keep the analog output clean, or the X32 is failing internally. (I assume this is all core audio on the Mac.)

P.s. I just saw in the post about MP3 processing that (guessing) you were using MP3s and that would have added some demand on the Mac to convert to the X32 expectation (PCM sample rate and bit depth). It’s possible that slowed things down or maybe one or the other ends was resource constrained because data was going both ways and needed more resources for conversion than could be gathered. Would be interesting to know if having Wav/Aif files at the correct sample/depth would have exhibited the same problem, assuming I’m not totally off the rails with my jump to the MP3 usage thinking.
 
Last edited:

rob aylestone

Well-known member
Playing back mp3 files doesn't cause any processing overhead does it? The mp3 file hard work was done converting it from wav or something uncompressed down to mp3. The macs are happy playing these back - and the playing back of an mp3 is a lot less stressful than playing video, so the mac is just sending a digital stream - the X32 seems the culprit - but it's happy sending lots of channels as wavs at 48K to the mac, it seems odd it cannot process two files going the other way? James' idea on the clock source might be the issue, and that's perhaps where the data is getting corrupted? I'll do some more tests?
 

keith.rogers

Well-known member
Playing back mp3 files doesn't cause any processing overhead does it? The mp3 file hard work was done converting it from wav or something uncompressed down to mp3. The macs are happy playing these back - and the playing back of an mp3 is a lot less stressful than playing video, so the mac is just sending a digital stream - the X32 seems the culprit - but it's happy sending lots of channels as wavs at 48K to the mac, it seems odd it cannot process two files going the other way? James' idea on the clock source might be the issue, and that's perhaps where the data is getting corrupted? I'll do some more tests?
I'm working on the assumption that the X32 is not doing the MP3 conversion, so then there would be processing overhead at the OS X end. Of course, converting to analog would probably be about the same processing, at least in the CPU end I might assume, but it's just possible that there's some more contention in Core Audio when it's being converted to another digital format and the pushed across the USB. You haven't said what the "digital distortion" is, but slow downs and clicks, pops, whatever usually indicate a resource problem *somewhere*. And, of course, if the data does get corrupted along the way, it would be up to the "driver" design (almost certainly just hardware or firmware in the X32) to decide how to deal with that. I'd guess the X32 is not designed for robustness in the digital-in corner of all the things it does, but that's just a guess.
 

rob aylestone

Well-known member
The X32 won't play mp3s from the in built player, so I assume any conversion would be done in the Mac? The distortion is a gradual roughness then fizzes and a sort of growing sound that you know will eventually fail. Can't describe it and no chance to record it sadly. That weird digital distortion that just sounds worrying? I've still not got to trying it in the store - one of those jobs for next time, I guess.
 
Top