Intermittent Digital Glitching in 24 bit Sessions

[MikeAP]

I have been working exclusively on 16 bit sessions the last 3 months - the other day I had to do a 24 bit session and had an annoying, untraceable problem crop back up - in the 24 bit session, I get intermittent digital pops. They're never in the same place twice. You cannot see them on the waveform and they occur on all tracks at one time or another. It is not enough glitching to make it unlistenable, but certainly enough to make it unusable.

I did some troubleshooting on this problem a couple of months ago and thought that I had solved the problem, but I think that this problem doesn't occur on 16 bit sessions - and that's what I've been working on lately. I have had this problem in sessions with as few plugins as 1 amp farm on an insert. It doesn't seem to be directly related to DSP Usage. I have had this problem -only- with 24 bit sessions.

To solve this problem, I grabbed another G3/300 Mixplus/Atto (exactly the same as mine except it does not use the gport - and it is a 300 instead of a 350) I hooked this other computer up to all of my hardware - cables, interface, hard drives, and it worked flawlessly.

My Setup: G3 350, OS9.0, ProTools 5, 192MB Ram, 4 9 gig Barracudas in a rorke enclosure, Mix Plus, Atto Scsi Card, 888/24 I/O, griffin gport. I don't think that it is a plugin related problem as the computer that I used to temporarily replace mine has virtually the same plugins - including Waves 2.8 on both machines.

Troubleshooting attempted: Reinstall DAE/DSI, Reinstall ProTools, Reseat cards, check TDM ribbon cable, ditch all prefs/digisetup, all system settings are within digi recommended standards (disk cached, mem allocation, all other system settings), scsi card is stepped down to correct speed, tried the new updated DAE/DSI.

This is manifesting itself as a clocking problem - these are the exact same pops that you hear when there is poor clocking going on. But I swapped out computers, and the problem went away - I am trying to determine if I have problematic cards or what...

Mike Purcell

[Arno Peeters]

Thanks for posting this Mike ! I had the *exact* same problem, only on a G4AGP/350 and a 001. At first I thought I had overlooked something, since it was my first 24 bit experience, so I wasn't going to make a post of it.
Now you're describing the exact same kind of glitches: you think they would be visible, but sometimes you simply can't spot 'em. I had them also 're-appear': once I though I'd lost them (re-doing crossfades, adjusting vol's etc.).
Since we're on two entirely different sets, could it be DAE-related ?

Any other users having this ? Digi, any clue here ?



------------------
Arno Peeters
www.tapetv.nl

[shawnosimpson]

Hey Mike,
If you have the time to do this, it might help narrow down the possibilities of problematic cards. Try taking all your cards (both systems, scsi and dig cards) out of their respective computers. Be careful you note which cards were where so that you can easily get your other system back in working order. If your computer has an internal scsi (beige G3's did, and i think the b&w did too), just hook your drive chasis to it for the time being. Now install only the core card into the computer you know was working properly. try recording a test tone @ 24bits for a minute or so and see if it pops or clicks. You may want to record it several times as several different files so that you max out your playback track count (on your internal scsi buss, that will be greatly reduced). A sine wave works well for this because you can easily set a reasonable zoom setting and just scroll by to see if there are any inconsistency in the waveform. Also, a digital pop in the audio will come leaping from that normally smooth tone. If all is well so far, try installing your scsi card and doing it again. If all is not well however, you'll want to swap to the other mix core and try it again. If this still doesn't work, then you'll want to examine your drives and chasis (maybe borrow a standalone drive from your dealer). Hopefully you'll be able to progressively add in each card with no problems in the first computer. Then move on to the other computer and repeat the process from the beginning. If the cards and chasis from the first computer that you know is working will not work in the second computer, you'll need to make sure all your extensions are set the same witht he same driver versions and memory allocation. If they are, you might want to try switching the ram from the good machine to the notsogood machine. I once battled ticking and popping for over a month just to find out that both machines had a stock 64 meg chip and 2 more oddball 32meg sticks in them. I simply moved all the 32megs to one machine and the 2 64 megs to the other and the problem went away. that was a couple of years ago when D24 was brand new, so i don't know if that problem would still apply. I hope this helps and please email me if there's anything I can do. Good luck.
--Shawn

[pverna]

Guys,
I don't have an answer to your problem, but for whatever it's worth, I've had a similar scenario. I haven't had a chance to troubleshoot it yet, but it seems to only happen in Direct Outputs mode. When I use the Pro Tools 24-bit mixer, I don't get any clicks and pops, but when I send discrete outputs out of my three 888/24s into my 02R, I get phantom clicks that don't appear on the waveforms.
Like you, I found that these glitches sound like clocking problems, but my clocking configuration is identical whether I'm in Direct Outputs or Stereo Outputs mode.
If I get any answers, I'll pass them on.
Best,
Paul Verna

[DigiTechSupt]

Mike have you run digitest yet?
Are you positive that the pop is NOT written to disk? If it's written to disk then you should check for disk related problems -
- Formatting software, fragmentation ect.

If it's not written to disk then look closer at your clock source. Is there a master sync box in your setup? MTP/AV or a USD ect..

You swapped computers and it all worked fine. How about ram, I've seen bad ram cause quite a few problems.

With the troubleshooting that you've already done it sounds to me like you have a bad mix core card. Try digitest to see if it passes.


B Carter
Digidesign

[MikeAP]

I forgot to mention that digitest shows everything as passing... One of the very first things that I did. Additionally, another problem that I had forgotten about is the fact that the computer will not shutdown after running digi apps - I can't remember if it shuts down after peak/tdm vers to determine whether it is a DAE thing or something related to ProTools/DAE.

As far as ext clocking goes, I eliminated all ext. clock sources. A single 888/24 running on its' internal clock. Although it should not be a problem with the 888 or the cable, as these went over to the other computer and they exhibited no problems.

Rail- Thanks for the troubleshooting tips, but I think it's gonna be a long time before I have both ProTools systems free at the same time in order to sit down for a day and swap components out. And on top of that, it's an intermittent problem. I did a 24bit mastering thing today with no digital glitches at all (and using waves plugins - some TSR requests earlier had come back suggesting that Waves may be problematic. Guess not.)

I guess that I will try and dig up another 128meg of ram and swap it out - although I had suspected ram a while back and ran it through all of the ram tests with tech tool pro and it passed with flying colors.

Any other insights would be appreciated, but I think that I am at an impasse until I can do further troubleshooting on this problem. Would take it to my dealer, but they are no longer a dealer.

After re-reading my initial post, a further clarification re: 16 bit sessions - I always use the 24bit optimized mixer (which has been re-installed several times with no change). The audio files are all 16bit, and I select 16 bit when I create the sessions. If that makes any difference at all.

[BDWillis]

Hey Mike — I had a similar problem a while back that cropped up in a 16-bit session that I subsequently opened on a 9600 with 24-bit hardware, though you may have already ruled out the difficulty I encountered. I was getting the same kind of intermittent digital snaps as you describe, but the one constant was that they were happening around crossfades. As you describe, there were no signs of them in the waveform — nothing traceable visually, and a sonic moving target. I found that whether the glitch would happen at the start or end of a given crossfade would change depending on where I started playback from.

The workaround I discovered (with the producer just across the room, not too nervous…yet!!) was to use the trimmer tool to JUST BARELY change the length of each crossfade; not enough to change its function or sound, but just enough to re-write it to disk. It's been some time since this happened, but I believe it had to do with the PT preferences being set at 24-bit, even though it was a 16-bit session (I don't recall whether "use dither" was checked or not…). Hope this sheds some light! BTW-I found another Castle sticker INSIDE my rack!! You were BUSY!!

Warmest Regards,
Brian

[MikeAP]

Brian-
They're not necessarily around crossfades... It can happen on tracks that are top to bottom no edits... Pretty frustrating, and all I really need to do is isolate the component that's causing the problem. BTW - It was Mike J who was maliciously slapping stickers everywhere.... I may have come up with the idea for inside the rack, but he was the perpetrator.

Mike P

[BDWillis]

Hey Mike —

I've been stewing on this one for days… clocking problems really TORMENT me. A few questions: 1) do the glitches occur in some, all, or just THIS 24-bit session?
2) how were the audio files in this session created - analog in? digital in? If dig. in, what was the source format/machine & did you get there via UFC? 3) were the audio files created on MULTIPLE passes/transfers from the source machine? 4) is the original source data still available to you, so you could load it in via the G3/300, and then take it to your G3/350 to see if the same problem occurs? You've done such a good job eliminating the possibilities, I'm trying to look for ANY logical connection. As we know from working with 48 & especially 32Track machines, digital ISN'T just digital, like they'd have us believe…

This is REALLY fishing in the dark, but have you tried doing a track transfer to create a new session document?! (fast & easy solution to attempt) Also, I really wonder about the RAM thing, per the DigiTech Support post — I've had two Macs with intermittent problems that were only cured when I replaced RAM modules. Good luck with this one… please let me know if you find it, or call & we'll combine brains.

Warmest Regards,
Brian

[trombino]

I'll second or third the idea of it being a RAM related problem. I had similar problems that I made absolutely no sense to me and that all went away when I started troubleshooting my RAM. I had one bad DIMM that caused me problems ONLY with Digi stuff. I'ld take a few hours and start removing your modules to see if the problem goes away.

Mark