Hi everybody,
I'm unsuccessfully trying to transfer some audio from an ADAT LX to my G4 using the 828 mkII. I'm using the factory 9pin cable but no signal is coming to the 828. I thought it was a Cubase problem, but it's the same even using the CueMix console. Is my machine dead? Do I have to reset it? Or maybe is there something wrong in my ADAT? Any help would be apprecciated.
Thanks
No signal from ADAT 9pin
Moderator: James Steele
Forum rules
Discussion related to installation, configuration and use of MOTU hardware such as MIDI interfaces, audio interfaces, etc. for Mac OSX
Discussion related to installation, configuration and use of MOTU hardware such as MIDI interfaces, audio interfaces, etc. for Mac OSX
No signal from ADAT 9pin
-------------------
MacG4 1.2 Ghz, OS 10.3.9, MOTU 828 MkII, Cubase SL 3.0
MacG4 1.2 Ghz, OS 10.3.9, MOTU 828 MkII, Cubase SL 3.0
The 9-pin is actually for sync and transport stuff. If you're only dumping eight tracks, I'm not even sure you'd need to use it. You really just need the optical (lightpipe) output of the LX to go to the input on the 828, and then put the 828 into 'external' sync from the ADAT optical input. The optical lines carry a wordclock sync signal of their own, so you shouldn't have a problem, unless there's a connection issue.
After that, you just need to designate the inputs in your recording app (and/or the MOTU panel), so you've got access to the 8 ADAT lines you just connected, and can record them.
Hope that helps,
George
After that, you just need to designate the inputs in your recording app (and/or the MOTU panel), so you've got access to the 8 ADAT lines you just connected, and can record them.
Hope that helps,
George
Hi George,
thanks so much for your suggestions. My problem is *also* syncing the ADAT with Cubase, but I gave up on it cause I didn't found a way, not even on this board. So I tried to found out *at least* how to dump the tracks, adjusting them by hand (and ear) on the Cubase song.
ADAT 9 pin cable is also for audio transfer and you should see the LEDs lighting on the front panel of the 828 if any signal is coming. It doesn't happen to me.
As far as optical cable, I tried several times but I get neither sync nor audio. It looks like the 828 completely ignore the ADAT no matter what I do (I tried SMPTE, optical, 9 pin and it's like no ADAT is present).
A last information can maybe be useful: if I use AudioDesk i get syncing (mac as a slave) but not audio...
Thanks
thanks so much for your suggestions. My problem is *also* syncing the ADAT with Cubase, but I gave up on it cause I didn't found a way, not even on this board. So I tried to found out *at least* how to dump the tracks, adjusting them by hand (and ear) on the Cubase song.
ADAT 9 pin cable is also for audio transfer and you should see the LEDs lighting on the front panel of the 828 if any signal is coming. It doesn't happen to me.
As far as optical cable, I tried several times but I get neither sync nor audio. It looks like the 828 completely ignore the ADAT no matter what I do (I tried SMPTE, optical, 9 pin and it's like no ADAT is present).
A last information can maybe be useful: if I use AudioDesk i get syncing (mac as a slave) but not audio...
Thanks
-------------------
MacG4 1.2 Ghz, OS 10.3.9, MOTU 828 MkII, Cubase SL 3.0
MacG4 1.2 Ghz, OS 10.3.9, MOTU 828 MkII, Cubase SL 3.0
MrT,
Hey again. I'm 99% positive that ADAT thing has no form of audio travelling on it, but I haven't had much luck getting the exact details on what it specifically does send. A while back I was looking at that, and the regular 9-pin sync protocol, to see if I could grab some form of "output only" transport/track arm commands for something I have, without dumping more data on the MIDI streams. I never got any good details, but I may go fishing again soon. The non-ADAT one will probably be easier to find.
The 'ADAT' 9-pin sync line is supposed to carry the wordclock for the sample sync, and the addressing info to keep the overall session, frames, or whatever locked together. The WC, which can come from a number of sources, has no real way of specifying where it is within a song,etc. It's sort of like if you had two gears with the same "tooth per inch" spacing, and locked them together. If you pulled them apart for a minute, and let the drive gear keep running, and then slammed them back together, they'd be sync'd again, but there's no telling "where" the second wheel would be rotated to, when it started back up. (sorry if you knew that already)
The ADAT one obviously carries some transport crap too, because the first device can initiate the play,record,etc. It didn't look like it carries track arm stuff, but I could be wrong.
It also looked like when the ADAT "recorder" devices pick up an incoming signal on that 9-pin input jack, they automatically kick into external sync/slave mode, so they may be screwed, if your DAW or master isn't sending the information. As the recorders usually have a pair for i/o, and have to be connected in a "chain", it's possible that the ADAT sync capable audio hardware needs to be at the top and act as master. The confusing part is that Nuendo/Cubase and all, have the ability to slave to non-optical "ADAT sync", so maybe the single 9-pin ADAT hookup on the cards and interface boxes is different somehow, or maybe it's "user selectable", and doesn't jump into sync like the input on the ADAT's. (???)
Here's the weird crap that may be beyond our control:
The Steinberg (or maybe *all* ASIO) apps will have that ADAT option as a sync source with ADAT hardware, but an older Nuendo2 manual here says it's tied to the ASIO Positioning Protocol, and that "as of this writing", it won't be available with non-ASIO OSX drivers or something. This may explain why AudioDesk worked for you, but not Cubase. The other part is, that MOTU was infamous for *not* being too eager to comply with the (industry standard) Steinberg protocols, which most apps were successfully using, so even if you had the ability, there's no telling what sort of anomalies may be buried in there somewhere. Post production people tend to find that sort of crap, but I'm not sure we'd notice any. I'm just glad that the "regular" external clock sync finally got fixed with my 424. For a while, many of us thought it never would.
I still think the safe route is to 'disconnect' any 9-pin lines you've got happening between the DAW/828, and just use the optical. With the out from the LX, going to the 828, you should be able to slave to "ADAT optical", and just initiate play from the ADAT, while Cubase is already running (in record). I think if you have a bi-directional optical loop happening, with both lightpipe cables, and have the ADAT in "digital" clock mode (with the 828 on internal), the ADAT may be in some sort of conversion mode where you wouldn't get the playback signals you wanted (maybe even a feedback loop from the DAW). I really have no idea on that, but it's probably in the manual.
I think a proper "full sync" config, with that 9-pin and all, is the only way to get "sample accurate" bi-directional transfers between the two. Fortunately, that should really only be a problem if you're trying to move individual tracks to the DAW for editing, and then send them back to the *exact* same location on a tape. I worked at a place where I had to do that between a Korg 1212, a BRC, three ADAT XT's, an MTPAV, and Bias Deck on a Beige G3, and it wasn't worth a crap. I always ended up having to nudge things around and calculate offsets, unless the tracks were totally isolated so you couldn't hear the phase shifts anyway. If you're just running 8 lines, you should be able to dump them all in one pass, work on them, and then send the whole stream back to a tape (if you even really need to). As someone else pointed out recently here, there's so much more you can do with audio tracks in a DAW, that you may as well do it all in there (maybe use the ADAT for remote, or "previously archived" old stuff).
Good luck with it!
George
PS- As I'm sure you know: The 9-pin you'll see mentioned in sync prefs,etc., is actually an unrelated protocol for transport/sync junk. That's some type of Sony format which just uses the same type of D-sub connector (may have originated on video decks). That's usually linked to DAW's with the RS-422 (serial) ports on the computer.
*Also, (I know you mentioned using a "factory" cable) remember that many of the 9-pin's you find for 422,etc. may have weird numbers of lines in them, or different pinnings from end to end. I bought one to rip one end off for something a while back, and it only had about 4 or 5 actual connected leads inside it, and they weren't even going pin-to-pin. The ADAT one may not be "1to1" straight through either.
More confusing crap: Apparently, some of the more elaborate ADAT stuff has that 422 port too (in addition to the ADAT sync ports). For instance, here are some pinouts, which of course, are listed as *ADAT Sync* RS-422 (two different things), but the PDF link is for their M20, which describes it as the Sony protocol.
http://pinouts.ru/data/adat_sync_pinout.shtml
If you see any good PDF or anything on the actual 'ADAT' sync connector format, I'd be interested in it.
Hey again. I'm 99% positive that ADAT thing has no form of audio travelling on it, but I haven't had much luck getting the exact details on what it specifically does send. A while back I was looking at that, and the regular 9-pin sync protocol, to see if I could grab some form of "output only" transport/track arm commands for something I have, without dumping more data on the MIDI streams. I never got any good details, but I may go fishing again soon. The non-ADAT one will probably be easier to find.
The 'ADAT' 9-pin sync line is supposed to carry the wordclock for the sample sync, and the addressing info to keep the overall session, frames, or whatever locked together. The WC, which can come from a number of sources, has no real way of specifying where it is within a song,etc. It's sort of like if you had two gears with the same "tooth per inch" spacing, and locked them together. If you pulled them apart for a minute, and let the drive gear keep running, and then slammed them back together, they'd be sync'd again, but there's no telling "where" the second wheel would be rotated to, when it started back up. (sorry if you knew that already)
The ADAT one obviously carries some transport crap too, because the first device can initiate the play,record,etc. It didn't look like it carries track arm stuff, but I could be wrong.
It also looked like when the ADAT "recorder" devices pick up an incoming signal on that 9-pin input jack, they automatically kick into external sync/slave mode, so they may be screwed, if your DAW or master isn't sending the information. As the recorders usually have a pair for i/o, and have to be connected in a "chain", it's possible that the ADAT sync capable audio hardware needs to be at the top and act as master. The confusing part is that Nuendo/Cubase and all, have the ability to slave to non-optical "ADAT sync", so maybe the single 9-pin ADAT hookup on the cards and interface boxes is different somehow, or maybe it's "user selectable", and doesn't jump into sync like the input on the ADAT's. (???)
Here's the weird crap that may be beyond our control:
The Steinberg (or maybe *all* ASIO) apps will have that ADAT option as a sync source with ADAT hardware, but an older Nuendo2 manual here says it's tied to the ASIO Positioning Protocol, and that "as of this writing", it won't be available with non-ASIO OSX drivers or something. This may explain why AudioDesk worked for you, but not Cubase. The other part is, that MOTU was infamous for *not* being too eager to comply with the (industry standard) Steinberg protocols, which most apps were successfully using, so even if you had the ability, there's no telling what sort of anomalies may be buried in there somewhere. Post production people tend to find that sort of crap, but I'm not sure we'd notice any. I'm just glad that the "regular" external clock sync finally got fixed with my 424. For a while, many of us thought it never would.
I still think the safe route is to 'disconnect' any 9-pin lines you've got happening between the DAW/828, and just use the optical. With the out from the LX, going to the 828, you should be able to slave to "ADAT optical", and just initiate play from the ADAT, while Cubase is already running (in record). I think if you have a bi-directional optical loop happening, with both lightpipe cables, and have the ADAT in "digital" clock mode (with the 828 on internal), the ADAT may be in some sort of conversion mode where you wouldn't get the playback signals you wanted (maybe even a feedback loop from the DAW). I really have no idea on that, but it's probably in the manual.
I think a proper "full sync" config, with that 9-pin and all, is the only way to get "sample accurate" bi-directional transfers between the two. Fortunately, that should really only be a problem if you're trying to move individual tracks to the DAW for editing, and then send them back to the *exact* same location on a tape. I worked at a place where I had to do that between a Korg 1212, a BRC, three ADAT XT's, an MTPAV, and Bias Deck on a Beige G3, and it wasn't worth a crap. I always ended up having to nudge things around and calculate offsets, unless the tracks were totally isolated so you couldn't hear the phase shifts anyway. If you're just running 8 lines, you should be able to dump them all in one pass, work on them, and then send the whole stream back to a tape (if you even really need to). As someone else pointed out recently here, there's so much more you can do with audio tracks in a DAW, that you may as well do it all in there (maybe use the ADAT for remote, or "previously archived" old stuff).
Good luck with it!
George
PS- As I'm sure you know: The 9-pin you'll see mentioned in sync prefs,etc., is actually an unrelated protocol for transport/sync junk. That's some type of Sony format which just uses the same type of D-sub connector (may have originated on video decks). That's usually linked to DAW's with the RS-422 (serial) ports on the computer.
*Also, (I know you mentioned using a "factory" cable) remember that many of the 9-pin's you find for 422,etc. may have weird numbers of lines in them, or different pinnings from end to end. I bought one to rip one end off for something a while back, and it only had about 4 or 5 actual connected leads inside it, and they weren't even going pin-to-pin. The ADAT one may not be "1to1" straight through either.
More confusing crap: Apparently, some of the more elaborate ADAT stuff has that 422 port too (in addition to the ADAT sync ports). For instance, here are some pinouts, which of course, are listed as *ADAT Sync* RS-422 (two different things), but the PDF link is for their M20, which describes it as the Sony protocol.
http://pinouts.ru/data/adat_sync_pinout.shtml
If you see any good PDF or anything on the actual 'ADAT' sync connector format, I'd be interested in it.
Hi George,
HUGE thanks for all your suggestions/advices and for the efforts you put in finding a solution to my problems. It doesn't happen everyday that you post a question and a unknown person spends his own time trying to fix it. Again, many thanks.
Now for the good news: I finally found out that my *optical* cable was dead! So I tried with another cable and swithced to your suggestion: *it works* or, better said, almost works. I get the audio signal, with each (or all) tracks selectable on the application (cubase). My problem is: no sync yet, but I believe to know why.
What you say about compatibility (cubase vs MOTU, APP etc.) must be true: In my "sync setup" window, I'm not able to select an ASIO driver as a clock source (dimmed option, I can send you a screenshot). I thought "Ok, let's transfer the audio and let's put it "in sync" manually, on the song window".
It doesn't work cause the tempo is different. In other words, if you put in sync the beginning of the song, you have the song out of sync in the middle. If you put in sync the end, it goes out at the beginning.
I had to slow down the tempo in the master track to achieve a decent result, but it's a tricky workaround. I blamed it on the frame rate, but it doesn't affect the speed, so I don't know what to do. Maybe some kind of driver is required... If I find a solution I'll let you know. Meanwhile any word from you would be greatly apprecciated.
Keep well,
Marco
P.S
Maybe is somehow relevant: I'm using just one optical cable: ADAT out -> 828 in. I don't have another at the moment.
P.S.
Even if I dropped the ADAT way, I'll try to find out if there's some PDF about the actual ADAT sync connector format, I'll let you know.
HUGE thanks for all your suggestions/advices and for the efforts you put in finding a solution to my problems. It doesn't happen everyday that you post a question and a unknown person spends his own time trying to fix it. Again, many thanks.
Now for the good news: I finally found out that my *optical* cable was dead! So I tried with another cable and swithced to your suggestion: *it works* or, better said, almost works. I get the audio signal, with each (or all) tracks selectable on the application (cubase). My problem is: no sync yet, but I believe to know why.
What you say about compatibility (cubase vs MOTU, APP etc.) must be true: In my "sync setup" window, I'm not able to select an ASIO driver as a clock source (dimmed option, I can send you a screenshot). I thought "Ok, let's transfer the audio and let's put it "in sync" manually, on the song window".
It doesn't work cause the tempo is different. In other words, if you put in sync the beginning of the song, you have the song out of sync in the middle. If you put in sync the end, it goes out at the beginning.
I had to slow down the tempo in the master track to achieve a decent result, but it's a tricky workaround. I blamed it on the frame rate, but it doesn't affect the speed, so I don't know what to do. Maybe some kind of driver is required... If I find a solution I'll let you know. Meanwhile any word from you would be greatly apprecciated.
Keep well,
Marco
P.S
Maybe is somehow relevant: I'm using just one optical cable: ADAT out -> 828 in. I don't have another at the moment.
P.S.
Even if I dropped the ADAT way, I'll try to find out if there's some PDF about the actual ADAT sync connector format, I'll let you know.
-------------------
MacG4 1.2 Ghz, OS 10.3.9, MOTU 828 MkII, Cubase SL 3.0
MacG4 1.2 Ghz, OS 10.3.9, MOTU 828 MkII, Cubase SL 3.0
It sounds to me like there is too much confusion going on.
To simply transfer ADAT tracks to your computer, you only need a TOSLINK cable from the ADAT out of the ADAT to the ADAT In of the computer interface. You need to set the ADAT to external clock and the MOTU unit to internal clock. Make sure the MOTU is set for 44.1 or 48 and the ADAT is recorded at 44.1 or 48K. Make sure your recording software is set to record a 44.1 or 48K file, make sure your software is set so track one receives track one, etc.
Hit "play" on the ADAT and "record" on the computer.. bingo.
The ADAT sync is to "time lock" the samples and only clock and transport data is sent..NOT track arm data. And NO audio is sent throught the sync cable.
You do not need to sync your ADAT to your computer if you are simply transferring tracks from the ADAT to the computer. All the tracks will be in perfect alignment after recording into the computer...if they were in perfect alignment on the ADAT.
To simply transfer ADAT tracks to your computer, you only need a TOSLINK cable from the ADAT out of the ADAT to the ADAT In of the computer interface. You need to set the ADAT to external clock and the MOTU unit to internal clock. Make sure the MOTU is set for 44.1 or 48 and the ADAT is recorded at 44.1 or 48K. Make sure your recording software is set to record a 44.1 or 48K file, make sure your software is set so track one receives track one, etc.
Hit "play" on the ADAT and "record" on the computer.. bingo.
The ADAT sync is to "time lock" the samples and only clock and transport data is sent..NOT track arm data. And NO audio is sent throught the sync cable.
You do not need to sync your ADAT to your computer if you are simply transferring tracks from the ADAT to the computer. All the tracks will be in perfect alignment after recording into the computer...if they were in perfect alignment on the ADAT.
Thanks for the clarification on the 9-pin. I wasn't actually sure of what was in it, other than it not having any audio. I was just thinking track arm stuff because, when I had to deal with some a while back, I thought I remembered the BRC being able to handle all the record status over the 3 ADAT XT's we had, and don't remember any other means of getting that status around the chain. I thought it did some sort of "ID assignment" around the loop, for the ADAT's it saw. 
Take Care,
George
Take Care,
George
Many thanks to everybody.
My problem is both transferring and syncing. I solved the first one thanks to George and a fresh optical (toslink) cable, cause the one I used before was broken, which added a lot of confusion in my investigation. Once i replaced it, I was able to transfer my tracks on the computer.
My second problem is syncing: i was *never* able to sync the computer to the adat. The only exception is: wth the 9 pin the application starts after the adat, but I'll never know if this is real *sync*, cause no audio comes from the cable (as everybody knows here).
So it looks like I can have audio but no sync (optical), or sync but no audio (9 pin cable)... Funny, isn't it?
My problem is both transferring and syncing. I solved the first one thanks to George and a fresh optical (toslink) cable, cause the one I used before was broken, which added a lot of confusion in my investigation. Once i replaced it, I was able to transfer my tracks on the computer.
My second problem is syncing: i was *never* able to sync the computer to the adat. The only exception is: wth the 9 pin the application starts after the adat, but I'll never know if this is real *sync*, cause no audio comes from the cable (as everybody knows here).
So it looks like I can have audio but no sync (optical), or sync but no audio (9 pin cable)... Funny, isn't it?
-------------------
MacG4 1.2 Ghz, OS 10.3.9, MOTU 828 MkII, Cubase SL 3.0
MacG4 1.2 Ghz, OS 10.3.9, MOTU 828 MkII, Cubase SL 3.0
MrT,
Hey again, and sorry I hadn't gotten back in from the last reply.
I'll often be in here a bunch, and then get busy with something and disappear for a while.
On the sync:
Keep in mind, there are actually about three different clock formats you're dealing with in digital audio / MIDI sequence sessions. They sort of depend on one another, and you can't really let the devices "guess" at any of them, if the info is unspecified or unavailable.
There's the aforementioned digital audio clock, which lines up the audio samples between your hardware, if there's any sort of audio link (optical) necessary. Then, there's the timecode clock, which specifies where, within the huge chunk of digital audio tape/disc, or session data, the actual minute/second/frame counting begins. And lastly, there's the sequence/tempo clock, which controls your song and beat sync (speed).
As an example: If you had a video, which you knew was going to come on a specific channel at a specific time today (a very specific time, like within a 30th of a second), and you had composed a MIDI sequence for it, and wanted to listen to your new sequence, while you watched it (in real time), you would need several details to be taken care of first. You'd obviously need your clock to be set to the exact same time as the TV station's clock, and for reliable, continual sync, one clock device would need to watch the other the whole time they were expected to lock together, in case they should drift apart due to their own mechanics. With this taken care of, along with an exact "known" starting time for the video, if you were to set your tempo and sequence up to play at 120bpm, and the song was drastically slower (like 90bpm), even starting them at the exact same time, using two perfectly synchronized clocks, would sound horrible within a couple seconds. So the MIDI "tempo clock" info really has no connection to whether or not the projects, sessions, wordclock and tapes are locked together, or set to the right formats. However, if you knew the exact tempo, down to a decimal percentage of a BPM, but you were going by the "second to second" resolution of the clock on your wall, and had no reliable means of keeping it lined up after the start of the song, you may still end up sounding out of sync, depending on where your playback actually started, and how stable all the clocks were.
So basically, you need both for sequencer/MIDI sync, unless you have devices which are capable of sending and receiving the MIDI tempo clock directly (bypassing all the absolute/hours, minutes & seconds crap). That sort of clock would be things like MIDI Clock, Beat Clock, SPP (Song Position Pointer), and the various "frequency shift" junk (FSK) which you can record on tape. They're considered more "primitive" than SMPTE/MTC, and most of the sequencer apps and devices which recognize the real TC formats, will do their own conversions, to calculate the MIDI (beat) clock locations from the incoming/outgoing SMPTE clock time throughout the playback (after you tell them the tempo and start times). The cool part is, most of the DAW/digital audio stuff (possibly with some outside assistance) is already constantly aware of it's own exact location while it plays, from the much higher resolution clocks and sync data which it's already using, so there's no real need for "old timey" time code, but something has to do the conversion to outgoing SMPTE addresses or MIDI beats for you (like the 9-pin standard).
Now for the messy part: If you run enough lo-tech "MIDI clock" devices into smarter software and hardware, which can display more accurate tempos (like a DAW sequencer), you start noticing how far off some of the actual tempos are from unit to unit. None usually land exactly on the BPM tempo you've set them for, which is similar to "guessing" the tempo, even if you knew the start time. To make a long story short there, if you're trying to sync to sequenced tracks you've got on tape, you should probably get a more accurate idea of what the tempo was that you were working with. One way might be to lock the computer to the ADAT (wordclock, ADAT optical, or 9-pin), and dump as much of the song (or a channel from it) into the DAW software and do a beat/tempo calculate between two known measures and beats in the song. Ideally, get two which are as far apart as you can, and obviously make sure you've counted them properly. After that, you *should* be able to set your master tempo in the sequencer to something closer to what the ADAT is putting out. If the tempo is correct, your SMPTE frame rate settings are correct between the two, you know the actual (timecode) start point of the sequence, and you've got a reliable lock between the two devices (via MTC,9-pin,etc), it should stay relatively sync'd throughout the song. Having it get off toward the end, like you describe, sounds more like a "freewheel" scenario, where your playback is just triggered and told to "start", but there is no further checking or adjusting between the two. If it is indeed in the usual "continuous resync" type of lock throughout playback, and the two devices are *reliably* locked together, but it's still drifting over time, I'd suspect one of the previous issues (either the actual tempo needs a minor "recalculation", or the frame rate or prefs are somehow mismatched). I'm really not even sure what all the different devices do, as far as frame rates, when they're in slave mode, but they may even detect the incoming format and set themselves (don't count on that).
The only other catch is the rare event that the original MIDI clock source or machine that tracked it, was "less than perfect". I guess if the tempo was 94 at the start of the song, and 97 near the end, or god-forbid, an actual "human" tempo source, your beat calculation would just give you an "average" for the whole thing, and there's probably no way in hell that you could lock to it, unless you manually drew up a full tempo map with all the changes.
All that crap is probably a lot easier if you just go ahead and dump the 8 channels worth of ADAT into your software and do all the work from then on in the DAW, but if you're anything like myself, you probably would like to see the two machines cooperate as their specs claimed they could, even if you don't need to.
BTW- If you do that "one or more channel" optical dump from the ADAT, and then do a "tempo detect" on it, like I suggested, you should be able to set your master tempo to it, and shift the audio track's start to a good "1 count" somewhere. If the tempo of your ADAT song isn't consistent from start to finish, you should be able to see it across the tempo grid bar, or whatever you've got in your sequencer app., while scrolling across all the waveform (beat) peaks. <-- guessing you've got drum tracks or something in there??
Hope some of that can help (and that you didn't already know most of it),
George
Hey again, and sorry I hadn't gotten back in from the last reply.
I'll often be in here a bunch, and then get busy with something and disappear for a while.
On the sync:
Keep in mind, there are actually about three different clock formats you're dealing with in digital audio / MIDI sequence sessions. They sort of depend on one another, and you can't really let the devices "guess" at any of them, if the info is unspecified or unavailable.
There's the aforementioned digital audio clock, which lines up the audio samples between your hardware, if there's any sort of audio link (optical) necessary. Then, there's the timecode clock, which specifies where, within the huge chunk of digital audio tape/disc, or session data, the actual minute/second/frame counting begins. And lastly, there's the sequence/tempo clock, which controls your song and beat sync (speed).
As an example: If you had a video, which you knew was going to come on a specific channel at a specific time today (a very specific time, like within a 30th of a second), and you had composed a MIDI sequence for it, and wanted to listen to your new sequence, while you watched it (in real time), you would need several details to be taken care of first. You'd obviously need your clock to be set to the exact same time as the TV station's clock, and for reliable, continual sync, one clock device would need to watch the other the whole time they were expected to lock together, in case they should drift apart due to their own mechanics. With this taken care of, along with an exact "known" starting time for the video, if you were to set your tempo and sequence up to play at 120bpm, and the song was drastically slower (like 90bpm), even starting them at the exact same time, using two perfectly synchronized clocks, would sound horrible within a couple seconds. So the MIDI "tempo clock" info really has no connection to whether or not the projects, sessions, wordclock and tapes are locked together, or set to the right formats. However, if you knew the exact tempo, down to a decimal percentage of a BPM, but you were going by the "second to second" resolution of the clock on your wall, and had no reliable means of keeping it lined up after the start of the song, you may still end up sounding out of sync, depending on where your playback actually started, and how stable all the clocks were.
So basically, you need both for sequencer/MIDI sync, unless you have devices which are capable of sending and receiving the MIDI tempo clock directly (bypassing all the absolute/hours, minutes & seconds crap). That sort of clock would be things like MIDI Clock, Beat Clock, SPP (Song Position Pointer), and the various "frequency shift" junk (FSK) which you can record on tape. They're considered more "primitive" than SMPTE/MTC, and most of the sequencer apps and devices which recognize the real TC formats, will do their own conversions, to calculate the MIDI (beat) clock locations from the incoming/outgoing SMPTE clock time throughout the playback (after you tell them the tempo and start times). The cool part is, most of the DAW/digital audio stuff (possibly with some outside assistance) is already constantly aware of it's own exact location while it plays, from the much higher resolution clocks and sync data which it's already using, so there's no real need for "old timey" time code, but something has to do the conversion to outgoing SMPTE addresses or MIDI beats for you (like the 9-pin standard).
Now for the messy part: If you run enough lo-tech "MIDI clock" devices into smarter software and hardware, which can display more accurate tempos (like a DAW sequencer), you start noticing how far off some of the actual tempos are from unit to unit. None usually land exactly on the BPM tempo you've set them for, which is similar to "guessing" the tempo, even if you knew the start time. To make a long story short there, if you're trying to sync to sequenced tracks you've got on tape, you should probably get a more accurate idea of what the tempo was that you were working with. One way might be to lock the computer to the ADAT (wordclock, ADAT optical, or 9-pin), and dump as much of the song (or a channel from it) into the DAW software and do a beat/tempo calculate between two known measures and beats in the song. Ideally, get two which are as far apart as you can, and obviously make sure you've counted them properly. After that, you *should* be able to set your master tempo in the sequencer to something closer to what the ADAT is putting out. If the tempo is correct, your SMPTE frame rate settings are correct between the two, you know the actual (timecode) start point of the sequence, and you've got a reliable lock between the two devices (via MTC,9-pin,etc), it should stay relatively sync'd throughout the song. Having it get off toward the end, like you describe, sounds more like a "freewheel" scenario, where your playback is just triggered and told to "start", but there is no further checking or adjusting between the two. If it is indeed in the usual "continuous resync" type of lock throughout playback, and the two devices are *reliably* locked together, but it's still drifting over time, I'd suspect one of the previous issues (either the actual tempo needs a minor "recalculation", or the frame rate or prefs are somehow mismatched). I'm really not even sure what all the different devices do, as far as frame rates, when they're in slave mode, but they may even detect the incoming format and set themselves (don't count on that).
The only other catch is the rare event that the original MIDI clock source or machine that tracked it, was "less than perfect". I guess if the tempo was 94 at the start of the song, and 97 near the end, or god-forbid, an actual "human" tempo source, your beat calculation would just give you an "average" for the whole thing, and there's probably no way in hell that you could lock to it, unless you manually drew up a full tempo map with all the changes.
All that crap is probably a lot easier if you just go ahead and dump the 8 channels worth of ADAT into your software and do all the work from then on in the DAW, but if you're anything like myself, you probably would like to see the two machines cooperate as their specs claimed they could, even if you don't need to.
BTW- If you do that "one or more channel" optical dump from the ADAT, and then do a "tempo detect" on it, like I suggested, you should be able to set your master tempo to it, and shift the audio track's start to a good "1 count" somewhere. If the tempo of your ADAT song isn't consistent from start to finish, you should be able to see it across the tempo grid bar, or whatever you've got in your sequencer app., while scrolling across all the waveform (beat) peaks. <-- guessing you've got drum tracks or something in there??
Hope some of that can help (and that you didn't already know most of it),
George
HI George,
this time it's me getting back to you late, sorry.
Many thanks for your time and your deep and exhaustive answers (as usual!). Besides all the things coming out in these interesting posts (I know much more about the whole sync matter now!) I think my problem is compatibility between Steinberg software and MOTU hardware. Now I can transfer my stuff via optical cable and that's fine but - as you guessed - tempo is not consistent from start to end so, even if I put everything in sync by hand at the start of the song, I go out at the first chorus... I have to split the parts in groups and sync them one by one, manually.
It's a workaround, but it's better than re-record the whole song.
Speaking about sync, the problem is: Cubase doesn't see the 828 as a sync device at all. If I look at the "transport settings" window, the "ASIO device" select button is dimmed. I *never* had luck with sync using Cubase and MOTU, never. I even tried SMPTE, but nothing happens...
A couple of useful informations: I'm using one single optical cable (adat out -> motu in) and................... if I use a MOTU application (AudioDesk or DP) everything goes ok, guess why?
this time it's me getting back to you late, sorry.
Many thanks for your time and your deep and exhaustive answers (as usual!). Besides all the things coming out in these interesting posts (I know much more about the whole sync matter now!) I think my problem is compatibility between Steinberg software and MOTU hardware. Now I can transfer my stuff via optical cable and that's fine but - as you guessed - tempo is not consistent from start to end so, even if I put everything in sync by hand at the start of the song, I go out at the first chorus... I have to split the parts in groups and sync them one by one, manually.
It's a workaround, but it's better than re-record the whole song.
Speaking about sync, the problem is: Cubase doesn't see the 828 as a sync device at all. If I look at the "transport settings" window, the "ASIO device" select button is dimmed. I *never* had luck with sync using Cubase and MOTU, never. I even tried SMPTE, but nothing happens...
A couple of useful informations: I'm using one single optical cable (adat out -> motu in) and................... if I use a MOTU application (AudioDesk or DP) everything goes ok, guess why?
-------------------
MacG4 1.2 Ghz, OS 10.3.9, MOTU 828 MkII, Cubase SL 3.0
MacG4 1.2 Ghz, OS 10.3.9, MOTU 828 MkII, Cubase SL 3.0