Weird glitch noise due to Digital Timepiece

[themagichoof]

I've got some old Tascam DA-38 machines that I need to transfer some stuff out of into DP. My DA-38 sync out is going to a Digital Timepiece to a 324 card (via ADAT sync cable) and I have a TDIF cable from the DA-38 to a 2408. That's a typical sync setup.

When I hit Play on the DA-38 there's a loud glitch noise shortly after. I suspect it's the DTP locking onto the Address. The DTP converts the Tascam's ABS time into MTC among other things, so I'm thinking that when the DTP finally locks itself into shooting out MTC it gives that glitch. The Address LED comes on as soon as the glitch is heard so that's my suspicion.

The audio sounds completely fine before and after since the Wordclock is rock solid and that LED is always on. Funnily enough I never had this problem when syncing up an ADAT machine at all.

[themagichoof]

Apparently the Digital Timepiece loses Wordclock once it gets to the point that it's converting ABS time into the plethora of other sync signals that it shoots out. It's not just ABS time though. I hooked an ADAT to it as the master and the DTP also takes a second or three to start generating all of that sync info, which makes the DTP stop generating Wordclock at the point of sync. The reason I'm hearing that glitch is that the TDIF connection to the 2408 doesn't like that Wordclock is dropped for brief second.

Also, as a side note, I can't get Performer to sync and start playing/recording with sample-accurate sync when going directly out of the DTP into my 324 sync card. For some odd reason I have to throw an ADAT machine into the chain before it hits the 324 card. Going directly out of the DTP into the 324 won't get sample-accurate sync to start playing. The ADAT in the chain makes it work its magic.

MOTU has said that they can't replicate either of these problems. I bought a nearly brand new DTP thinking it was the culprit but it does the same thing.

[Shooshie]

I don't have a DTP, and I'm certainly no sync guru, but I'm trying to think what might cause such glitch, and what might be subject to variation so that MOTU may not be replicating it. The only thing I can think of is that for some reason BEFORE it syncs there may be some problem in the connection or in which device is being considered the master. Do you have other devices in the chain that could show up in the Hardware list as a master? Could you be using the Mac as a master? It seems like the glitch would happen if it's changing masters in mid-stream.

I remember way back that I went for a few years without understanding the way MOTU implemented the master selection menus. That was a long time ago, and I'm not suggesting that you are mistaken, just that there may be something in your setup that eluded you as it did me for a long time.

Then again, maybe the DTP doesn't work right. But I'd sure go over my settings with a fine-toothed comb.

Shooshie

[themagichoof]

I've been through two Digital Timepieces which MOTU suggested as the problem. Both give the same thing so I can rule that out.

Once the DTP notices that it's receiving either Tascam's or the ADAT's ABS time (after I hit Play on an ADAT/DA-38), the DTP takes those sync signals and converts them to the plethora of its possible sync out signals. The DTP takes anywhere from 100ms-3 seconds to get all of those signals ready for output after I hit Play on an ADAT/Tascam machine, and once it's at the point that it has good, converted signals to output on all of its possible sync out formats, the DTP hiccups on what seems like every single one of them (in this case, ADAT Sync Out, but it hiccups on the regular BNC WC out which I tested too).

At that point, the TDIF jet noise appears. My theory is that the TDIF connection wants a stable, solid WC reference at all times and when there's a brief hiccup in the WC that the TDIF is syncing to... you'll definitely hear it.

I'm gonna have to do more testing to further pinpoint the problem, but I really shouldn't have to. MOTU can't replicate it so they say but I'm going to double check with them.

[themagichoof]

Also, the 2nd issue is a weird one. In short, Digital Performer (or Cubase even) should start playing/recording when you hit Play on your Tascam/ADAT machine sample-accurately. Here's what my testing revealed:

ADAT machine -> 324 (Performer syncs fine)
Tascam machine -> DTP -> ADAT machine -> 324 (Performer syncs fine, with the exception of that damned TDIF jet noise)

However!

Tascam machine -> DTP -> 324, Performer does NOT sync.

I have to put a freakin' ADAT box in the chain and make that hit the 324 instead of the DTP to get Performer to sync.

MOTU can't replicate this either, but again I'm going to double check.

[themagichoof]

Just for kicks I tried something different as far as glitch noises.

ADAT machine Lightpipe in/out -> 2408
ADAT machine Sync Out -> DTP ADAT Sync In
DTP BNC WC Out -> 2408
324 PCI Console set to 48k and 2408 Wordclock Sync In

The DTP hiccups as per usual after hitting Play on the ADAT. When monitoring the Lightpipe signals via Cuemix, there's a glitch noise, then an audio dropout, then it comes back on and is fine. This proves that the entire DTP is dropping all of its Sync Out signals, no matter the type.

It's either the DTP was never made to run well as a slave, all of them drop their signals after receiving any sort of ABS time, and MOTU is mistaken about this, or my 2408 is shot. I'm leaning towards the former at this point.

[Shooshie]

Not surprisingly, I can't make heads or tails of it. I wish I could, but maybe MOTU will get it figured out. I haven't read much about the Digital TimePiece, but it just seems awfully strange that they'd put such a flagship-type product out there that can't sync without causing major gaps and noises.

Will it sync without hitting Play? Seems like maybe it should get synced up before you actually do any transport commands, but then again it may require hitting play for it even to try. The last time I did a lot of syncing was over 10 years ago, and I don't remember the exact behaviors and all. But I used to do a lot of it, and I never had problems like what you're describing. That was when my MIDI TimePiece would sync with SMPTE or MIDI Time Code.

I sure hope you get it figured out. And I hope that MOTU isn't sitting on a few thousand DTP boxes knowing that they all do this.

Shooshie

[mikehalloran]

About this time, we need to know what your hardware and apps are. Include the firmware version of your DTP and your OS.

Put it in your signature -- that's why we do it.

[themagichoof]

I've got multiple systems but it's mainly the hardware that's the issue as far as the jet noise.

DTP - ROM 7.2.5
324 - Rev.3
424 - unsure, but it's a PCI-X card
2408 - (2) vers a 1.1 vers b 1.2 cv 0.10
Digital Performer 2.7.2
Mac OS8.6/9.2.2

The DTP automatically grabs a stable Workdclock from an ADAT/DTRS unit when you initially connect them and it's noted by the illuminated Word LED. When you hit Play on an ADAT/DTRS unit though, the DTP notices that it's got some incoming ABS time. It starts converting that ADAT/DTRS ABS time into MTC and it's noted by the illuminated Address LED. However the Address LED only comes on after 100ms to 3 seconds after hitting Play, and it's at THAT point that the DTP hiccup occurs for a brief second, dropping all of its Wordclock outputs, which is noted by the Word LED going off and coming back on along with the Address LED coming on. The Tascam Sync Out, ADAT Sync Out, and BNC Out all have their Wordclock dropped.

http://www78.zippyshare.com/v/14284010/file.html

^^^ this is the TDIF noise. The ADAT one is similar. It doesn't sound like much, but when you're tracking at -20db, it's LOUD.

I admit that letting the DTP be the master works fine with no drops, but that goes completely against how I needed things to be set up. MOTU haven't responded so I'll try to call them later.

The part about Performer not syncing to a signal directly out of the DTP and needing an ADAT in the chain is another issue altogether.

[Gravity Jim]

I don't have a Digital TimePiece and I don't use my MTP/AV for synch anymore, so I can't be much help, except to say this:

You drew a conclusion about the problem before you started trouble shooting, and most of what you have said about your process indicates that you are still holding that conclusion as correct without actually testing the theory. There is no doubt that MOTU did not create a product that's raison d'être is synchronization but won't synch, so you can rule that out.

I would start over, looking at your connections, without any assumption as to what is happening inside the DTP. I am not a synch expert, although I did synch my rig to SMPTE time code for many years before the video guys stopped using tape, and my intuitive guess is that you either have a bad driver install, a weird set-up in Clockworks, or have time code or synch signals coming into the DTP from more than one source.

Synch was the most frustrating studio tech that I can recall. Good luck with the problem.

[Shooshie]

Synch was the most frustrating studio tech that I can recall.

Indeed it was, but I always read the MOTU manuals, and when I went in, I just did what I had learned, and it always worked. Studio people thought I was a genius and a timecode guru, but the truth was that before I walked into a studio with a MIDI Time Piece for the first time, I had never synced anything in my life. I'd memorized MOTU's instructions, and it just worked. And it always worked.

MOTU is known for making things that work, which is why this sounds very strange. I just feel like there is a mismatch going on somewhere in his setup, like maybe it's slaved to one type of code, but relaying to others in another, with some kind of invalid link in the chain.

I'm clueless, but it's "educated cluelessness." Instinctively I believe it should work. Yet the themagichoof seems like the sort of guy who has done this more than a few times, so I have to give him credit, too. Anyway, he's talking to MOTU, and that's about the best thing he can do. I hope they get it figured out.

Meanwhile, if I were themagichoof, I'd do what Gravity Jim said and replace some drivers, reinstall some software, throw out Preferences, run Safe Boot, and fix permissions. There just may be a problem in there somewhere.

Shooshie

[themagichoof]

You drew a conclusion about the problem before you started trouble shooting,

I've been testing and trouble shooting for almost two months now on and off. That's long enough to draw a conclusion I think, it's the only thing that makes sense at this point.

I tested this without a computer at all because the DTP itself is dropping Wordclock, no need for installing drivers to notice that. However the other issue which is Performer only syncing in certain scenarios is where my computer comes into play, but it's not looking I'm setting up anything wrong.

I called MOTU and I should have a response soon. Hopefully they'll be able to test and confirm. We'll see.

[themagichoof]

MOTU responded. They don't have the necessary hardware to test. They did say that the DTP takes a second to lock up when sending it MIDI, but that isn't the problem.

Hopefully someone will read this in the future that has an ADAT/DTRS and can check.

[mikehalloran]

and fix permissions

Back to what was posted about assumptions. It's been so long since any of the rest of us have worked with OS 8/9, I wonder how many read that and didn't remember that you don't fix permissions before OS X.

You can - and should - rebuild the desktop now and then.

I probably still have old copies of Norton Utilities for rebuilding directories and fixing resource forks etc. in 8/9.

[Shooshie]

and fix permissions

Back to what was posted about assumptions. It's been so long since any of the rest of us have worked with OS 8/9, I wonder how many read that and didn't remember that you don't fix permissions before OS X.

You can - and should - rebuild the desktop now and then.

I probably still have old copies of Norton Utilities for rebuilding directories and fixing resource forks etc. in 8/9.

I totally missed that he was working in Mac OS8 and Mac OS9. Seeing that he didn't tell us that until about 5 posts ago, after we'd been looking at this for days, I'm not surprised that I missed it, since my machines are using OSX.8.x and OSX.9.5, and just glancing over his specs I assumed THAT's what he meant. My only presumption was that he was working on a system from this century!

I have to bow out of this discussion. I can barely troubleshoot current software and hardware. I'm not about to speculate on why a bunch of 15-20 year old equipment and software are not working properly. The original poster might have made that clear in the beginning!!!

Shooshie