I know that ADAT isn't popular here...

S

smtcharlie

emotion over sonics
But I just want to see if anyone has run into this problem as well. I have one ADAT and a lightpipe card to transfer to PC. Lately, I need a few more than 8 tracks on songs, so I thought that I would be a bit clever and bounce tracks/mixes from the computer and back to free up tracks. Basically, I copy tracks to computer, make a stereo mix, then copy that back to 2 tracks of ADAT at a different point in the tape. Then I have 6 open tracks for more instruments.

I realize that when I copy the second group of tracks to computer, I'll have to cut the wave files to get them lined up with the original bunch, that seems to go fine. But I am having a problem with the tracks - they are in time at the beginning of the song, but not in time at the end of the song. It's a pretty significant difference. All the new tracks sync with each other just fine, so it's not some odd problem with one of the tracks (or one performance that's crappy that I didn't remember being so crappy at the time!).

Does anyone know why 2 sets of tracks recorded on ADAT would be different speeds after being digitized on computer? ADAT Xt, Wave Center PCI card, Dell desktop, into Saw Pro. And I use Sound Forge to cut the wave files such that they have the same start point.

Thanks for the advice in advance...
 
A(int) DAT Peculier

First of all, Charlie, what's a guy from Indy doing up at 1am EST the night/morning before the big race??? ;)

Second of all, I still own an ADAT XT myself, so you're not completly among enemies here. :)

Third, your actual problem. My guess is its related to the fact that it sounds like you have no common timecode synchronization going on (correct me if I'm wrong about that.) If you are just moving from timebase to timebase to timebase (the PC/SAW clock to the ADAT clock and back again) with no master clock playing traffic cop, there is indeed a fair chance that your clips will wind up out of sync.

While it may seem logical that digital time is digital time and a rose is a rose, and that you can move around digital timebases freely, for reasons which it is frankly just to late at night for me to get into right now, that is not really the case. What's really needed in that situation is the selection of a "master timekeeper". This can be in the form of selecting one of your digital devices as the master (the other being a slave), or having an external, impartial 3rd device serving as the master timecode generator for the whole studio to sync to (this is often referred to in A/V studios as "house black.")

Being that it's after midnight here and an hour later there, I'm going to leave a season-ending cliff-hanger here so I can get to bed and get up in time to see Danica Patrick becone the first female to win the 500 tomorrow. :) In the meantime, you have some starter information re timecode and synchronization and a hint as to how you might be able to avoid the problem but cutting down on the amount of timebase shifting you have to do (hint hint) ;)

Good night, I'll check in on you after the race :)

G.
 
Stick a click or some sort of common reference sound before the tracks. When you bring the second set of files back into the software you should be able to zoom in and line up the files (from the inserted click) manually with no problems. I still use adats and have just recently started mixing in software. I don't have a Adat light pipe card so I sometimes dump 16 or more tracks into software individually and line them up manually. In other words, it can be done and should be alot easier considering your situation.
 
Also- if you only have one Adat, why not just record the next 6 tracks or whatever using only the adats converters. So lets say your first 8 tracks were drums and you dumped them into the software. Why not just overdub directly into the software via the adat/light pipe card, rather than doing 5-6 tracks at a later point on the tape?
 
Ah, yes, that mysterious time code concept (which I don't understand). Thanks for the quick answers, this does indeed set me on the right track. There's plenty of reading around here on this concept...it's funny how I only learn things when I have a problem....

Wireneck, yes, I could record straight to computer after dumping the first round of tracks into my computer. I'm just funny about computers...I work for a high-end math software company, I'm no stranger to computers....but the feel of recording audio on them just doesn't sit right with me. It's a funny thing. I take some sort of comfort in having an actual tape rolling (even thought ADAT is fairly digital...).

And the race will be on every major news platform in town for week, it's safe to say that I won't miss any details!
 
Are you using the ADAT sync port on your ADAT? That should help. You should also run a clock either from your ADAT or from your CPU. Even a click wont help since the out of sync data may actually need to be stretched or shrunk. This is a common issue with syncing stuff. Clocking is extremely important.
 
previous posts are correct, I am using no central time clock, and I am not using the sync port in the ADAT. I suppose that I assumed that what you played was what you get...which, after reading your posts, I now realize was a bit silly since I'm shuffling data in digital form via lightpipie.

thanks again for all the help, you have given me a couple ideas for future recordings...the current batch will need some work. And my computer has gone totally crazy in this process (applications are failing to launch....)....so this has turned into a total mess!
 
smtcharlie said:
Does anyone know why 2 sets of tracks recorded on ADAT would be different speeds after being digitized on computer? ADAT Xt, Wave Center PCI card, Dell desktop, into Saw Pro. And I use Sound Forge to cut the wave files such that they have the same start point.

Thanks for the advice in advance...
Click to expand...

Are you leaving the sample rates the same?
 
UB802 said:
This don't make no sense at all. SOMETHING is wrong SOMEWHERE.

You SHOULD be able to do what you are trying to do just fine. If they start out "in time" as you claim, they should STAY in time throughout, as the ADAT machine IS NOT like analog tape where the speed can fluctuate. The timing of digitally recorded tracks to tape is fully buffered, and the tape speed is only adjusted to keep the storage of the digital info going at a proper speed.

You can in fact vary to speed of digital tape and have pitch issues!

So, something is VERY weird here. I really can't tell you what though! The ADAT should be recording at the same speed ALL THE TIME.
Click to expand...
Alright. Southside Glen nailed it. He should as he owns an ADAT. It is most definitely sync related and can probably be remedied by running syc from the ADAT to your comp. Of course that requires you to have the proper connectors on your computer. Usually you can only find this on higher end cards like a Delta 1010(not LT). In which case you might as well record straight to disk but that's another story.
 
UB802 said:
This don't make no sense at all. SOMETHING is wrong SOMEWHERE.

You SHOULD be able to do what you are trying to do just fine. If they start out "in time" as you claim, they should STAY in time throughout, as the ADAT machine IS NOT like analog tape where the speed can fluctuate. The timing of digitally recorded tracks to tape is fully buffered, and the tape speed is only adjusted to keep the storage of the digital info going at a proper speed.

You can in fact vary to speed of digital tape and have pitch issues!

So, something is VERY weird here. I really can't tell you what though! The ADAT should be recording at the same speed ALL THE TIME.
Click to expand...
If the adat and the computer are not looking at the same clock, they will fall out of sync. Contrary to popular belief, digital isn't perfect. The variation isn't as much as with analog but over a five minute song, it can get quite noticable. The problem here is that he is going back and forth several times. Every time the transfer is made the timing problems get added together. If he is loosing 1/8 sec each transfer (for example), by the time he gets the stuff onto his computer-> makes the sub mix-> transfers it to the tape-> performs more tracks-> transfers the new tracks to the computer... he has lost almost 1/2 second accross the song.

Smtcharlie, I used to think as you do. I used to record to DA-88's then transfer it to the computer. It was a waste of time. It is so much cooler to be able to punch something in without loosing the take before. Tape machines give off heat, suck electricity, and are just one more thing to break down. I know you only have the one, so it isn't as bad as my 5, but still. Just try it without the tape.
 
UB802 said:
You SHOULD be able to do what you are trying to do just fine. If they start out "in time" as you claim, they should STAY in time throughout, as the ADAT machine IS NOT like analog tape where the speed can fluctuate.
Click to expand...
Ah, but the data can and does fluctuate when they are all following their own digital clocks. Sample rates, frames per second, drop code vs. non-drop code timing, mathematical conversions and clock frequency inacuracies when moving from one clock to another, starting clocks between samples, and even digital jitter can all throw digital signals out of sync if they are not all coordinated and following the same clock.

The problem here stems greatly from the fact that he is going into the DAW and bouncing down to a 2-channel WAV for the 1st pass, and then making a second pass of 6 more channels into the ADAT and then sending it all back into the DAW again. There are several timbase conversions happining in that total signal path, and those timebase conversions will cause timing problems at the end of the chains.

The irony here is that if he passed the signals between the ADAT and the DAW via the analog path, then you'd be right, he could do it in such a way where "what you hear is what you get" and the sync problems would disappear. I'm not recommending that he do all that A/D conversion, just illustrating the point that when you stay in the digital domain but move from machine to machine, you have to have a common timebase and "traffic cop" to keep all those 1s and 0s ducks in a row.

G.
 
Go Danica!

Hahaha, UB, we're not ganging up on you. Check the timecodes (no pun intended) on the last three messages...looks like we all just happened to be typing basicly the same answer at abou the same time. I wonder if we all got through watching the race at the same time??? :D

G.
 
yes, I'm leaving sample rates the same. It's a much, much smaller difference than changing sample rate. I don't even notice any problems within the first 1-2 minutes of the song...but by minute 7, things are plain not together. So a slight drift in time code seems to make sense to me.

but I agree that there could be an element of something funky going on here - after noticing the problem, applications now don't start, they refuse to run after a re-install (in fact, one refused to uninstall for a while), Windows doesn't boot half the time, and there are recurring errors that state that I'm missing .dll files. Not good...I can't think of anything to do except wipe the drive and start with a fresh Windows install...

So unfortunately, I've got much bigger problems that I originally thought.
 
Be careful, you might be losing format on the hard drive. I'd back up all importatnt data that you have ASAP (if you can) and run a Scandisk or Norton Disk Doctor (or similar) on it. Hopefully you're not losing the drive or controller altogether. :(

Good luck!

G.
 
One technique that you may want to try is to stripe SMPTE to one of the ADAT tracks and slave the DAW to SMPTE during transfers. Granted you will lose one additional track, but it should allow you to sync the bounces more accurately. I've done this with ADATs and Cakewalk by using a box that converts SMPTE to MIDI time code for Cakewalk. If your DAW can read SMPTE and slave directly from an analog input it's worth a shot.

The best way to sync ADATs with another machine or DAW is by using a BRC or possibly the universal slave driver if using Pro Tools, but there are other boxes that do a good job as well. Also if using a BRC you also will not need to give up one of the tracks to SMTPE since it generates this for you.
 
Using the 9 pin control on ADAT's is pretty much necessary for sample accurate transfers. Then make sure the only one device in your chain is sending the master signal. Either clock the ADAT's to your soundcard, or vice versa.
 
Be careful, you might be losing format on the hard drive. I'd back up all importatnt data that you have ASAP (if you can) and run a Scandisk or Norton Disk Doctor (or similar) on it. Hopefully you're not losing the drive or controller altogether.
************

Yeah, my CD writer software was one that crapped out, but that I could re-install...and everything is now backed up.

Thanks again for all the guidance, one post above in particular gave me an idea. I think that I did actually copy the "reduction mix" from computer to ADAT by analog. I wasn't really that concerned about the quality since that were just reference tracks. So I may be able to salvage this project by copying the newer tracks onto my computer analog as well to eliminate these clock issues (and save me editing that I'm not so sure that I can handle)....

But, of course, my computer has to decide to work before trying this!
 
UB is right for the most part this time. However, not having a master is not a guarantee of clicks and pops. When you have misclocked digital equipment you can have clicks, pops, dropouts, sync loss, pitch and time issues etc... You could get all of those, and occasionally even none of those. Basically be not clocking properly you are asking for a large amount of problems that can range from extremely subtle to absolutely terrible and anywhere in between.
 
Hmmmm...yes, this is interesting. I think that I set up the WaveCenter card such that it is always the master clock. There was some thinking or reason for it when I did this ~2 years ago originally, but I can't remember what that was (someone I know probably recommended that I do this).

In the past, yes, I do remember hearing clicks and pops when copying via lightpipe from the computer back out to the ADAT, which supports the above.

But for this current batch of stuff, I used a USB convertor (Edirol) to go Computer -> USB converter to analog -> patchbay -> analog in on ADAT. Again, the USB hook-up is in the front which is easier than the lightpipe jack in the back of the computer (again, I didn't really care about the quality of the stereo reference mix). So I still want to try copying the new tracks analog to see what that does....

I also have only one optical cord, which I use for both directions. Sounds like I'll need to get another one.

Very interesting, and great advice. I really, really appreciate the guidance. Maybe everyone has gone through this before (I hope that you haven't), but I just completed a fairly large project when I ran into this problem. I was really upset!
 
******************************
The irony here is that if he passed the signals between the ADAT and the DAW via the analog path, then you'd be right, he could do it in such a way where "what you hear is what you get" and the sync problems would disappear. I'm not recommending that he do all that A/D conversion, just illustrating the point that when you stay in the digital domain but move from machine to machine, you have to have a common timebase and "traffic cop" to keep all those 1s and 0s ducks in a row.
******************************

Just to close the thread, I did get my computer working again (sort of) after a clean install of Win XP. It's still acting fairly flakey, but I hope that it will hold together just to get this project done, at least.

Thanks again for the advice, it was very helpful. And now that I understand what I was doing (and the pitfalls), the advice above turned out to be the trick that I needed. Since I was going analog from computer to ADAT, I re-copied the files analog one by one to the computer and edited the start point so they all sync'ed up. This did the trick!
 
You must log in or register to reply here.
Back
Top