Digital Syncronization

[Hornplayer]

Can anyone tell me the best way to sync my Fostex D-160 to my MOTU 2408 mkll, then into my computer running Sonar? I'm trying to copy tracks into Sonar for mix down, and getting a bunch of noise, which is not on the Fostex.

I'm using the optical (light pipe) out of the d160 into the motu optical in, using the motu to slave to the d160. (ie using the clock input off the Fostex) When I transfer the tracks this way to Sonar, bad noise! I can only conclude that the sync is off. Do I need an external sync device, or can I do this all internally?

I've checked and played with latency, to no avail, and checked the output of the recording. That's ok too. It's just the transfer to the computer into Sonar that's bad.

Anyone out there know exactly how to do this?
Thanks for the help.
John

 

[Track Rat]

You don't need any other devices. You should be able to make the transfer with what you have. What are you setting the MOTU to when making the transfer? It should be set to ADAT optical, not ADAT sync. Also make sure the bit depth and sample rates match.

 

[Hornplayer]

Trackrat:

Thanks for the reply. On the PCI console of the MOTU I have it set to 2408 Bank A clock. (where the d160 is inout). Shold that be set to PCI 324 ADAT?

Bit rate and sample depth match (16 bit, 44100 sample rate) on both units.

Maybe a latency problem?

John

 

[Hornplayer]

Well the units sync up now, thanks for the advice trackrat, but the noise is still there. Now it sounds like a buzz. Grrrrrrr.

I've tested the recording. Played it back through the mixer into amp and speakers, so I know the recording is ok. It has to be in the transfer from the d160 through the motu to Sonar. Tweaked every latency parameter I know about.
Danged if I can find it.

Yikes!!!!!

John

 

[skippy]

If the D160 has word clock out like my D1624, I would suggest configuring its word clock output to drive the word clock input on the MOTU- and then configure the MOTU for external word clock sync, rather than trying to sync to the ADAT stream. Deriving clock from an ADAT signal certainly does work, but I don't think that it is the best way to go: you get more jitter than you want, and in the degenerate case (marginal signla-to-noise on the optical connection), you can get sync losses and dropouts.

You don't need to go to an elaborate house clock setup, but I believe that you can get benefits by running word clock separately from the ADAT stream. At least, it works better for me with my 1624 and a Hammerfall.

This is all predicated on having word clock connections available on both units, though. Given that, I'm 99% certain that all you need is a BNC-BNC cable and reconfiguring the clock setups on both ends to solve your problem.

One other item: the MOTU should be able to lock very solidly to either the ADAT stream, or word clock. If going to word clock solves the problem, then I'd recommend checking out the health of the TOSLINK hardware in your rig: are the cables too long/kinked/damaged, are the connectors dirty/dusty, are the send and recieve ports clean? One dust-bunny in the wrong place can wreak havoc with an ADAT optical setup.

If you can't get a good lock on the ADAT stream, perhaps there is some situation that is killing the noise margin on that optical connection. If that's the case, you'll get bit errors in your data even after you solve the clocking problem, and that's no good either!

Best of luck- let us know how it goes...

 

[Hornplayer]

Skippy. Thanks for the reply.

BTW Skippy. No word clock out on the d160. (no bnc connection on the recorder, and nothing in the firm ware on the recorder indicating word clock support. There is MMC support however)

Here's what has happened so far in the saga of the noise.

changed out the toslink cable with a shorter brand new one.
Still noise
Checked all three light pipe connections on the motu for dust/dirt.
Checked light pipe connection on d160 for dust/dirt. Tested all three motu light pipe connections
Still noise
changed the output on the recorder to analog out, and hooked up to the motu through analog out to motu analog in.
Still noise
(this one I didn't expect to still have noise)

So I've concluded that either:
A. something wrong with motu
B. Electricity related
C. Something in Sonar set up wrong.(software related)

Going to test the motu in stand alone mode, disconnected from the computer

Change electrical circuits

Record directly to Sonar through different sound card.

If anyone has any other suggestions I'm all ears.

Stay tuned.

If none of the above works

pray to the recording gods for forgiveness for what ever I have done.

John

 

[skippy]

Thanks for the update- I didn't know if the D160 had word clock or not. Now I know!

If you have noise when you were going *analog out* on the 160 to analog in on the MOTU, then you have a very desperate clocking problem- and it is with the D160, for sure.

Sounds to me like both units are trying to slave to the other unit's ADAT output: a clock resolver loop. If the D160 was configured for internal clocking, it would *not* have given you noise on the analog outs. As it is, it sounds like the two PLLs are talking to each other, and drifting until they get to their limits- and then one or the other snaps back, and the cycle repeats- but with s couple of missed/misshapen clocks during each excursion.

The Fostex does a dodgy job of slaving to external ADAT clocks, so I'd use it as the clock master, _not_ as a slave. Make *sure* that it is set to internal clock sync, and not to ADAT clock. That should give you clean analog playback: nobody else's clock will be interfering, and the D160's clock is certainly stable enough for our purposes.

The most critical thing is to get clean analog playback, so that you know that the D160 has a stable clock environment. Until you have that, nothing else matters.

At that point, the MOTU should be able to sync to the now-clean ADAT stream coming from the D160, set up as sync slave with its clock source set to one of the input ports from the D160.

Keep the faith- this is a solvable problem. Getting clocking schemes set up in always an interesting trial! Hope that helps...

 

[Hornplayer]

Well, I'm beginning to know a great deal about what doesn't work.

Here's the latest.

I ran the d160 recorder via lightpipe out to the digital mixer in. Played the recording back. Sounded just fine. Then took the master analog out of the mixer to the motu analog in, then selected the analog out puts on the motu within Sonar. Same noise.

I monitored the signal out of the mixer, it was good.

I changed the mixer clock from internal to solo, (slave) mode and left the motu in internal. no change on the noise.

Changed the Sonar clock from audio to internal, and still no joy.

I'm wondering if the chain from the mixer, to the motu to Sonar is nothing more than a digital daisy chain and should be slaved together and mastered by one source. I'm not sure. Mixer and motu are word clock (bnc connections) equipped.

At this point I'm at a loss to explain anything else.

It seems to me the problem is between the motu and sonar. I know the signal is good at the input of the motu. After that I don't know where it gets corrupted.

The only avenue I haven't tried is an external clock, syncing all units to it.

Any suggestions on a reasonably priced word clock unit SkippY?

Still keeping the faith but losing ground.

John

 

[JR#97]

Have you tried slaving the D160 to Sonar via Midi? You might not get rock solid sync like you would if you had word clock, but it should work. I don't know about the noise problem.

I transfer from my FD8 to Sonar and vice versa all the time without wordclock. I don't get any sync problems whatsover.

 

[skippy]

Well, since the Fostex doesn't have word clock in, getting a master house clock generator doesn't buy you as much as you'd like. So it's my opinion that the Fostex has to act as the master clock generator, and everybody else has to slave to it.

Since you know that you can get solid sync from the Fostex to the mixer, you're halfway there. Can you set the mixer up to send word clock out? If so, then you could use the mixer to derive the master word clock from the D160 ADAT outputs, and then send it from the mixer's word clock out to the MOTU's word clock in. Set the MOTU for external word clock sync: that should have the MOTU and the mixer locked up solidly. Then do whatever you need to do to get Sonar to derive_its_ clock from the MOTU's external word clock input, and it should work.

I agree that the basic problem is between the MOTU and Sonar. I've never used Sonar, so I don't know how its timebase is configured. But it has to be possible to somehow tell it to use the MOTU's external clocking rather than some internal timebase. That ought to clear it up- but then, I've been wrong before...

MIDI sync is okay for positioning- I use MTC to slave all my gear together for that. But it isn't applicable to this sample-level sync problem.

 

[Hornplayer]

Thanks again for the replies guys.

What I'm going to do is concentrate on the motu to sonar interface first. I agree skippy that, that is where the problem lies. Trouble is I've had this working before with a Yammi md8 recorder on win 98se). Mystery why the Fostex is acting up. (I'm still optimistic that the d160 is an upgrade from the yammi md8 <smile>) I upgraded operating systems on the computer to win 2000 pro, and didn't pay attention to the sync settings between sonar and motu. (shame on me!)

Going to test a straight recording from the motu to sonar in various configurations of clock sync, and get them talking. At least I'll know if those two are sync'd..

Skippy, sonar allows a number of time piece sync configurations so I agree with you that it has to be possible to sync up motu and sonar. I'll hit the sonar and motu manuals hard and see if I can turn something up out of those.

If and when I get that straightened out I'll work on the mixer to motu to sonar via word clock. The mixer does have word out, and that may be a solution as well. That might be the best way to set the studio up anyway.

Going to take the weekend to beat this up. Lock myself in the studio until I have a solution. If I don't emerge by Monday morning with a solution, I'll make sure my wife sends in a search party for me. <smile>

I'm optimistic that I'm close, but I know close only counts in horseshoes and hand grenades.

Stay tuned. I'll share all I learn this weekend.

Thanks again to all for the suggestions. I hope others on the forum benefit from this discussion. It's been a good one.

John

 

[Hornplayer]

Well the saga of the noise has been solved.

Skippy, you were right in that it was a sync problem, but not in the way you'd expect.
All my digital equipment was sync'ing just fine, and there in lies the problem.

I realize this is a Fostex forum, but I'll post this anyway, and if anyone knows anyone having trouble with digital syncing, with Sonar, pass this along.

Sonar in a sense is it's own worst enemy when it comes to digital syncing. It is so powerful, it can create sync problems. Let me explain.

In Sonar you can select any sound card in your system for record master timing and a different one for playback master timing. My particular set up has 3 sound cards. One cheapie that is on the motherboard, a sound blaster, and the motu 2408 mkll. In sonar I selected the motu for record master timing, and the soundblaster for playback master timing. You guys are ahead of me here. <smile>

Sure enough when you record on one master clock, and play back on another master clock, you've set up sync problems, because each sound card has it's own clock, and there are minor differences between them. At play back time the clock that was recorded on does not sync to the sound card clock being used for play back. Instant no sync. Sounds like no sync, acts like no sync and is no sync.

I changed sonar to use the motu for both recording timing master and playback timing master. NOISE GONE! Works in all regimens.
Lightpipe and analog. Perfect transfer. I tested 8 tracks analog and digital, and it worked fine.

All sonar users must read and understand chapter 17 (Audio Performance) in the manual WORD for WORD. Here is where I found the problem, and a direct quote from the manual. "Every sound card's clock crystal is slightly different, which causes minor differences in the actual playback rate on each card. These differences may lead to slight synchronization problems if you use one card for recording and a different one for playback...." Page 17-5. Pretty discreet sentence to potentially cause such calamity.

Lesson here is using different sound cards for recording sync and playback sync can give you fits. Just because sonar lets you do it does NOT make it a good idea.

Skippy, thanks much for your advice. You kept me on the no sync sluthing path. You were dead on right.

BTW. The Fostex d160 has NO INTERNAL CLOCK. It must be slaved to digital devices at all times. It cannot be used as a master. That is straight from Fostex support line. Just a tid bit to keep for future reference. Interestingly it has the software loaded to support word out, but no hardware on the box.

Thanks again to all who replied.
Hope this helps someone down the road.
John

 

[skippy]

Glad I could help. The debug for noise problems in digital rigs is always to start with the clocks, so it was easy to offer up that advice!

I think Fostex support may be a little off base there- I find that hard to believe, because a standalone HDR with no internal clock source *at all* would be completely useless outside the confines of a complete studio rig. The thing certainly must have an internal clock: I got the idea for my remote rig from seeing other folks use a D160 with a small mixer to do location recording, where the D160 simply records the signals present on its analog inputs, sampling with its own internal clock source.

I just checked out the old reviews on the unit- it does have the same sort of internal clock source that my 1624 has, with varispeed and all. So the Fostex guys did mislead you on that: the setup menu should have an entry for clock setup, and one state of it should indeed be "internal". http://www.studio-sound.com/archive/dec97/r_fostexd160.html . You just have to love the Fostex support folks, sometimes! It's too bad that they force you to buy the 8345 for word clock I/O: that damned thing is _expensive_, and most of that cost would be wasted unless you needed to sync to SMPTE.

I do this type of completely-standalone remote work with my D1624s, so that I can simply pull the drive carrier out of the road rack unit, and plug it into the studio unit after a day of remote tracking. In that case, I have the road unit set up for internal clock, and the studio unit set up for word clock in...

 

[ausrock]

Hornplayer and Skippy,

I've been watching this thread with interest even though most of it is "above" me as I have had no reason to date to deal with synching of any sort....................although in the future this may change somewhat. Just thought I'd warn you Skippy.....lol. You may be sorry I have your email addy.

I just had a quick read through my D160 manual again and it does show numerous configurations where the D160 is "Master" although these are almost always with other Fostex gear and always using MTC.

OK, I'll crawl back into my comfortably ignorant cave now...........lmao.