Topic: Syncing word clock question

[Brian Skalinder]

I've never synced word clocks before, and am uncertain if it provides the functionality I need.  I'm hoping syncing the word clock across two different ADCs will allow me to record 2 stereo WAVs, each from a different ADC onto a different recorder, but maintain the ability to easily sync the WAVs in post.  Basically, the same ease of use for 4-channel recording, a la the R-4 or 744, but using two different ADCs and recorders.  (I have no desire to align WAVs at the start and adjusting every 2-3 minutes due to drift, etc....been there, done that.)

So let's say I have two ADCs w/ recorders.  ADC 1 has a word clock output, ADC 2 has a word clock input.  If I sync ADC 2's word clock to ADC 1's word clock, will I be able to accomplish what I want?  Based on what I've read, I think so, but am seeking confirmation from those with direct experience.

[heath]

i believe that what you are trying to do will be accomplished.  the only problem syncing will be the start time difference base on the how long it takes to start one recorder, then the other.  but that is easily overcome by lining up the beginning in a daw.  from that point on, you shouldn't experience drift.

[Brian Skalinder]

i believe that what you are trying to do will be accomplished.  the only problem syncing will be the start time difference base on the how long it takes to start one recorder, then the other.  but that is easily overcome by lining up the beginning in a daw.  from that point on, you shouldn't experience drift.

Yup, easy enough to sync once at the start of the file, I just don't want to have to do it 25-30 times per recording.  Excellent.  Thanks, Heath!

[BayTaynt3d]

Bri,

I'm curious about something. Whenever I've had to sync two things, be it two audio sources or a video and audio source, etc. (do this all the time in video prod), I usually just sync the head (start), then go all the way to the end, and stretch/compress the timeline of one of the tracks until I have tail sync too. Then, I'm done, with only two steps. This can be made much easier by making sure to have a head and a tail "slate" to sync from (loud distinct hand clap into both sources at start and end work great for this, but you can do it from the music alone too if you have a good NLE). Now, some folks might cringe at the thought of stretching, but the reality here is you rarely need to stretch more than 1 second, but probably more like milliseconds. Let's say it's a full second over a two hour set, that's only stretching one second across 7200 seconds, and as far as I can tell, with a good algorithm, won't change pitch and won't be noticeable at all. Once the head and tail are synced, then render out and track and flac. This is easy in Vegas, but I've got to think it would be in other mixer apps too? Am I missing something here? Is this just a totally sacreligious approach? Coming from video editing, this was just a no-brainer approach for me (as long as the drift is minimal, which it usually is for me, less than 1/2 second or less nearly all of the time). How do you even make an "adjustment" every few minutes anyway?

[Brian Skalinder]

I usually just sync the head (start), then go all the way to the end, and stretch/compress the timeline of one of the tracks until I have tail sync too. Then, I'm done, with only two steps.

Tried that before using Audacity, Audition, and Wavelab, and have not succeeded.

How do you even make an "adjustment" every few minutes anyway?

Chop up one source into 3min blocks.  Then align each block individually with the "complete" source.  This is what I've had to do in the past in order to avoid audible effects from misalignment.  Maybe my previous ADCs clocks had excessive drift. 

[BayTaynt3d]

I usually just sync the head (start), then go all the way to the end, and stretch/compress the timeline of one of the tracks until I have tail sync too. Then, I'm done, with only two steps.

Tried that before using Audacity, Audition, and Wavelab, and have not succeeded.

What didn't work?

By the way, (WARNING: going for the Vegas fluffing), one thing that is AMAZING about Vegas is it's ability to be playing and looping a region of the timeline WHILE YOU EDIT. This is virtually unheard of in video editors, including the most pro ones out there like PPro and FCP. A lot of folks kind of chuckle when they hear I use Vegas for Video, but man when I show them that feature their freakin' jaw usually drops. Anyway, the reason why I mention this in the context of this conversation, this feature in combination with Vegas' intuitive stretching UI is that it makes it syncing the tail ridiculously easy IMHO.

First, I sync the head. That's easy, done. Then I usually go all the way to the end of the set, and hopefully there is some talking like the band saying thanks and whatnot. What I do is then zoom in pretty tight to a 15 second stretch or so, select the region, and hit play such that it'll just keep looping. Then, I'll zoom in even further, way in. I'll first eyeball it if possible with waveforms, and in vegas you just CTRL-drag the tail to stretch (instead of trim). Then, once I'm close, I'll just hold the CTRL-drag handle and move it shorter or longer, ALL WHILE IT IS LOOPING. So, I get immediate feedback as I'm dragging such that I can hear the echo come and go. Once satisfied, I let go of the dragging. And boom, perfect sync across the entire set. It takes a while to get the hang of it in Vegas (and there's some subtle things I glossed over in the steps, but that's basically it). I am somewhat of a power user for several years (doing video in Vegas), and now I can sync like that in maybe 5 minutes or less tops.

[Javier Cinakowski]

Damn, I never thought of streaching one of the sources.  Great idea.  You could measure how far they are apart and streach by that much.  Im gonna give this a shot in CEP


Thanks.....

[Brian Skalinder]

What didn't work?

Time stretching or compressing one of the WAVs to match the other.  No matter what setting I used, they did not sync up over the course of the complete recording, i.e. I still had to re-align every few minutes (or maybe a bit less frequently, if memory serves).  It's been a while...maybe I did something wrong... 

[BayTaynt3d]

What didn't work?

Time stretching or compressing one of the WAVs to match the other.  No matter what setting I used, they did not sync up over the course of the complete recording, i.e. I still had to re-align every few minutes (or maybe a bit less frequently, if memory serves).  It's been a while...maybe I did something wrong... 

Interesting. Well, one thing is that the assumption is that the drift is consistent throughout the whole set. That may or may not be true in some instances, but I guess I'm thinking of analog tape more than anything else where speed of the tape can fluctuate (faster at the beginning and slower by the end or vise versa), but I wouldn't think that type of thing would happen with a digital recorder (even DAT). Seems like it'd be a consistent drift based on the different clock speeds in the different units. I wonder if it might be possible by jitter? Strange.

[F.O.Bean]

hmmm, wishin the v3 i have had both optical output AND word clock now, tho i would usually NEVER do the syncing real-time, always done for me in post

[Steve J]

What didn't work?

Time stretching or compressing one of the WAVs to match the other.  No matter what setting I used, they did not sync up over the course of the complete recording, i.e. I still had to re-align every few minutes (or maybe a bit less frequently, if memory serves).  It's been a while...maybe I did something wrong... 

Interesting. Well, one thing is that the assumption is that the drift is consistent throughout the whole set. That may or may not be true in some instances, but I guess I'm thinking of analog tape more than anything else where speed of the tape can fluctuate (faster at the beginning and slower by the end or vise versa), but I wouldn't think that type of thing would happen with a digital recorder (even DAT). Seems like it'd be a consistent drift based on the different clock speeds in the different units. I wonder if it might be possible by jitter? Strange.

Would it not be because there is a fractional difference between the two digital sources (i.e., each source is fractionally different from the other, whether by accuracy of the internal clock or by the physical mechanism, as in the case of a DAT deck); and in order to match the audio as one piece, you have to stretch *both* souces to match the fractional time difference, much like finding the lowest common denominator in a fraction math problem?

Just thinking...

[BayTaynt3d]

What didn't work?

Time stretching or compressing one of the WAVs to match the other.  No matter what setting I used, they did not sync up over the course of the complete recording, i.e. I still had to re-align every few minutes (or maybe a bit less frequently, if memory serves).  It's been a while...maybe I did something wrong... 

Interesting. Well, one thing is that the assumption is that the drift is consistent throughout the whole set. That may or may not be true in some instances, but I guess I'm thinking of analog tape more than anything else where speed of the tape can fluctuate (faster at the beginning and slower by the end or vise versa), but I wouldn't think that type of thing would happen with a digital recorder (even DAT). Seems like it'd be a consistent drift based on the different clock speeds in the different units. I wonder if it might be possible by jitter? Strange.

Would it not be because there is a fractional difference between the two digital sources (i.e., each source is fractionally different from the other, whether by accuracy of the internal clock or by the physical mechanism, as in the case of a DAT deck); and in order to match the audio as one piece, you have to stretch *both* souces to match the fractional time difference, much like finding the lowest common denominator in a fraction math problem?

Just thinking...

Yes, but I was hypothisizing (sp?, LOL) why Brian could get the head AND tail in sync, but still have stuff out-of-sync in the middle. The only theoretical way that is possible is that the drift wasn't consistent. I've personally never had that problem, I just sync the head, stretch the tail to sync, and the entire set is synced after those two steps -- do it all of the time.

[Steve J]

I think I've only had it work out once. 

I imagine that, in my own case, I was trying to balance the physical factors (where the mics were placed in relation to the source and the delay caused there), the mechanical factors (one source being from a DAT and the others from a laptop or other hard disc/CF recorder) and the human factor (one SBD source I had the FOH engineer dumped extra reverb into one of the vocal channels). Trying to synch all that up was a chore and a half.

My current has made things much easier synch-wise; but there are still times when I have to have a remote recorder set up in the room. I'm really interested to hear about Brian's results in his testing.

[TNJazz]

hmmm, wishin the v3 i have had both optical output AND word clock now, tho i would usually NEVER do the syncing real-time, always done for me in post

I think the word clock signal could be grabbed from one of the AES outputs with a custom cable.

It can, and devices can be clocked via SPDIF (coax or optical) as well.

[ethan]

I've never synced word clocks before, and am uncertain if it provides the functionality I need.  I'm hoping syncing the word clock across two different ADCs will allow me to record 2 stereo WAVs, each from a different ADC onto a different recorder, but maintain the ability to easily sync the WAVs in post.  Basically, the same ease of use for 4-channel recording, a la the R-4 or 744, but using two different ADCs and recorders.  (I have no desire to align WAVs at the start and adjusting every 2-3 minutes due to drift, etc....been there, done that.)

So let's say I have two ADCs w/ recorders.  ADC 1 has a word clock output, ADC 2 has a word clock input.  If I sync ADC 2's word clock to ADC 1's word clock, will I be able to accomplish what I want?  Based on what I've read, I think so, but am seeking confirmation from those with direct experience.

This is exactly what I'm planning on doing with my two AD1K's...I'm just waiting on the BNC cable I ordered from B&H.... damn holidays