Nice!

[Beck]

SMPTE is not a sycnhronization reference....it's time reference. It took me awhile to wrap my head around that when I was researching the whole sync thing....but then I got it and now understand why a sync reference is needed plus a timing reference for 100% sync every time.

Something has to act as a sync reference for both DAW and tape deck for them to be 100% locked tip to tail, sample/frame accurate, and something has to control the slave and make it chase to keep on time with the SMPTE timing reference.
SMPTE alone can't do both when the tape is master and when mechanics cause the speed/playback of the tape to fluctuate at minute levels....and so the DAW will chase/resample as slave or if not, it will drift at sample/frame level.
I run Samplitude (not a shabby DAW by any standard). It gives me the option to let the DAW chase/resample...or not.
There's a reason the options are there.

Again, do 4-5 dumps of the same click track on a 4-5 minute song from tape to DAW....and then zoom in across all the cick tracks and see what you got. If your DAW doesn't give you the choice to chase/resample or not....then it's automaticlaly set for that, and why you may not see any drift....but then your DAW is resampling on-the-fly.

It's not about any bad implementation" or anything like that....it's the nature of the mechanical transport system and physical tape.
Now, that all might be "good enough" and that's a choice everyone can make.

Oh...
I went online to check out how sync is done in Adobe Audition/CEP...and I came across this on their help forums:

Adobe Community: Do I need one midi-cable or two, in order to sync Audition with an external hardware midi-sequencer?
"What I encountered was the exact problem that you described - there was no way to get MTC or SMPTE to accurately sync the two."

From reading that....people are saying you can't do proper SMPTE/MTC sync with external hardware and AA...????

SMPTE is nothing without a device to interpret it, if that's what you mean, but SMPTE & MTC is absolutely a synchronization reference... The World Standard Synchronization reference for MIDI, in fact. There is a lot of misconception about synchronization on the web, but I've been using SMPTE since 1978, beginning in Television syncing audio to video, without issue. As SMPTE/MTC emerged with the MIDI spec I started using it with drum machines and hardware sequencers, and then DAWs... all without issue except where there was a problem with hardware or software implementation.

The link on the Adobe site is just another of thousands of conversations on the web of people trying to figure out sync that don't really understand it. Thankfully I've been using it and mastered sync before there was a web and all the confusion.

The critical software layer that controls SMPTE/MTC implementation on Echo products is in the Control Panel. The success or failure of any software like Audition or anything else depends on the Layla hardware/software implementation.

Both SMPTE and MTC have certain specifications that must be adhered to in the design of a product. If it doesn't work in a given device it is not functioning properly. In other words its broken... so when evaluating a piece of hardware or software based on that criteria you would have to give it a thumbs down for that feature... and look for something better.

My Echo Laylas would be really surprised to learn they haven't been working all these years I've been syncing then to an ATR. I mean the shock could kill them... so shhh!

 

[Chilljam]

I have very limited knowledge with sync (My current setup involves syncing two ATRs but that's as far as I've ever delved into it) but I don't see the point of syncing the DAW to the tape deck since the tape deck is not synced to any reference and hence will not play consistently every time the OP dumps the tape tracks to DAW. SteveM, you are saying that it is working fine, does that mean you are not having to 'nudge' parts of the tracks from different dumps around to get it all to line up perfectly? If yes, then is this just because the song is short and it will eventually be noticeably out of time or is there something here that is controlling the tape deck to play the same speed each time that I have missed?

Wouldn't it be better to record 'through' the Ampex (Record on all 4 tracks and monitor off repro head and feed that signal to the DAW then each additional parts you add you play a long and listen to the other tracks from the DAW and record through the tape deck again and just slide the entire take back a few milliseconds to make up for the delay between recording and playback on the tape deck)

This way, you will get 100% sync on each take (assuming the DAW is stable) and you get to record on all 4 tape tracks.

 

[miroslav]

SMPTE is nothing without a device to interpret it, if that's what you mean, but SMPTE & MTC is absolutely a synchronization reference...

OK...we can refer to it as that....but it's still not the same as a sync reference that all devices resolve to...it's a timing reference "where and how fast".
I'm not in any way trying to lecture you...I know you have a lot of experience...I'm just relaying the information that I have gained in my research of synchronization practices, and specifically why a synchronizer like the MicroLynx that controls ALL devices that resolve to it, makes the accuracy of the sync between them much more stable and accurate than when you just "free run" a tape deck, and let other devices chase it.

This is right out of the MicroLynx manual and SMPTE guide:
"SMPTE does this; it is an absolute timing reference that indicates both the speed and position of a tape as it travels across a tape machine transport."
IOW....you still need something generating a sync signal that ALL devices resolve to and are controlled by so that they are always locked to and in sync with.

Also from the MicroLynx manual:
"For a Digital/Audio Workstation or digital tape machine to stay in sync with other time code based equipment, a “common” synchronizing speed reference is required. Normally a fixed rate reference like video sync or sample rate clock (word clock) is used."
SMPTE or MTC is not mentioned for this task.

And more:
"Keep in mind that the time code is simply a modulated midband frequency recorded on the tape. This means that it is subject to all the same problems that other audio signals encounter when recorded to magnetic tape including tape dropouts, level shifting, frequency response variances, and distortion caused by dirty, worn, or out of spec heads, transport, or electronics."
This further suggests that SMPTE on tape is not going to be 100% stable/accurate compared to DAW timing...so again, the DAW will need to adjust to overcome that when the tape fluctuates.

In the scenario of the tape deck being the master, and sending SMPTE to the DAW...what is the tape deck resolving to?
It's own internal clock source.
What is the DAW resolving to?
It's own internal clock source.
So...you have two devices passing SMPTE code....BUT...each device resolves to a different clock, or "sync reference".
Sure, the SMPTE is on the tape...but what guarantees that the tape will always play back at the *exact* speed with *zero* mechanical fluctuations...time after time after time?
There is nothing other than it's own internal clock to do that... which can NOT make the tape overcome all/any mechanical transport fluctuation 100%.
That is why decks have wow & flutter and why we need to always adjust the mechanics with age/time.

The DAW can only react to what comes off the tape deck. There is NO direct sync between the clock of the DAW and the clock of the tape deck. If the DAW has to make up for any fluctuations, it will re-sample on the fly.
With a synchronizer like the MicroLynx....you get both the single sync/clock reference and control.
I think in most any studio or video post house where SMPTE is used....there is a single house sync reference that everything resolves to, and synchronizers that keep the machines locked and under one control.

But all that said....I still agree that the drift can certainly be "good enough for Rock & Roll"...so how accurate and precise you want your system to be, it's up to each person to decide, and how much drift will be acceptable.

Oh...I also use Layla24 converters, so it's got nothing to do with them or the DAW software which is pretty robust. When I ran my Fostex G-16 (which even has it's own, on-board sync card) as the Master....there was still small timing differences with multiple passes. Mind you, I never used the DAW in the chase/re-sample mode, which if I had, I would not have the drift...BUT then there would be the re-sampling on-the-fly.
Now I run DAW as master, but since I also now use the MicroLynx, I actually could run either DAW or tape as master without fluctuations between dumps....becuase there is a single sync/clock reference and central control of all devices.

Anyway...drop a click signal on one track of the tape deck.......then do 4-5 dumps to DAW with the deck as master and NO MicroLynx or other synchronizer between them...then compare the 4-5 click tracks you have in the DAW. (The OP should try this with his 440 and see how the track line up at sample level).
Also....don't you actually use a MicroLynx in your setup, Beck...?

 

[miroslav]

...I don't see the point of syncing the DAW to the tape deck since the tape deck is not synced to any reference and hence will not play consistently every time the OP dumps the tape tracks to DAW.

This is the point I've been trying to make.

SteveM, you are saying that it is working fine, does that mean you are not having to 'nudge' parts of the tracks from different dumps around to get it all to line up perfectly?

I would guess that he's just going by his ears....and really, as I've said, the drift with songs in the 3 minute range would be negligible if you just *listened* to the playback.
You have to compare multiple passes of the same track to see what drift there is coming from your tape deck in free run, Master mode. Looking at the drum track against the guitar track is not going to tell you anything, because even when we play, we are never in *perfect* sync...there's always that "human feel", and again, it could sound good enough to you....but if you want to make an accurate assessment of any drift, dump the same track to your DAW 4-5 times and then zoom in to sample level (you may not even need to go that far)....the drift will be there with tape deck as master and no synchronizer controlling both tape and DAW from one reference clock.

Wouldn't it be better to record 'through' the Ampex (Record on all 4 tracks and monitor off repro head and feed that signal to the DAW then each additional parts you add you play a long and listen to the other tracks from the DAW and record through the tape deck again and just slide the entire take back a few milliseconds to make up for the delay between recording and playback on the tape deck)

This way, you will get 100% sync on each take (assuming the DAW is stable) and you get to record on all 4 tape tracks.

Yes...this is the "poor man's" CLASP-like method...and the only sync issue is moving all the tracks the same amount each pass. With a 2-track or 4-track deck where it's impossible to archive ALL your tape tracks since you are reusing the tape tracks....that's an option that might appeal to some.
Personally I like tracking "to tape" and during my tracking sessions only work with the tape deck and console...sans DAW.
Then I will transfer the tracks to the DAW after I have them all on the tape.
I will make dumps of portions of the tape tracks between sessions...like dumping the drums/bass/rhythm guitar when those tracks are done. Then I can run them off the DAW, to rehearse other tracks if needed, saving tape ware...and when I'm ready, I can go back to tracking on the tape.
Even if I end up needing more than 24 (23 actually) tracks, and I maybe dump a few into the DAW....like when I'm grabbing 4-5 lead vocal passes or guitar leads, etc, and I'll dump them to DAW and just save one of each on the tape....
...when I'm done tracking, the tape is full on all tracks, and I have relatively complete archive of my tracking sessions on tape. With any CLASP-like method...there are no tape tracks archived for any future purpose....which may not be an issue for some folks.

 

[Beck]

...but I don't see the point of syncing the DAW to the tape deck since the tape deck is not synced to any reference and hence will not play consistently every time the OP dumps the tape tracks to DAW.

This is the point I've been trying to make.

This is not correct. What you're describing here is how a tape deck would behave by simply starting it at the same time as another device (another tape deck or DAW) WITHOUT synchronization. Speed deviation will prevent two tape decks from agreeing for very long… perhaps a minute or perhaps several minutes, depending on a number of things that impact deviation. But no two devices of any kind will stay in sync indefinitely without one following the other. So we invented sync to keep from happening what you are saying would happen.

What SMPTE is and how it works is what it is, and always has been in the motion picture and music recording industry… before we were using DAWs or even ADAT or even open-reel digital decks. I see there is a fundamental and persistant missunderstanding about what synchronization is (how it works) on this forum and many other forums. It deserves a separate thread, which I will commit some time to after the weekend (hopefully a sticky). I suspect the core problem is the same root as many misunderstandings on these forums… and that is people are understandably looking at things through digital colored glasses because that's where most people come from… but you won't get there that way. The clocks in a DAC and the clocks in a controller/converter for SMPTE/MTC and other synchronization methods are not at all related. But if you come from a perspective where the first exposure to the concept of sync involves word clock then yes it can be very difficult to unlearn that and see time sync as it relates to locking devices together.

When you use SMPTE with a DAW, or what we've always called "SMPTE-Over-MTC" since the late 80's, you have an absolute time reference that is in the same place on the tape at every pass, so there is no drift due to tape speed fluctuations. The sync device follows those fluctuations and thus so does the DAW or sequencer on the other side of the SMPTE-to-MTC converter. Basically you have a SMPTE generator put down a SMPTE stripe to tape. You then need a converter (sync box) to translate SMPTE to MTC (MIDI Time Code) so hardware sequencers and DAWs can make sense of it. Tempo maps and such can get very complicated or be very simple depending on the device used. Once you have that down you need well-designed hardware and software to make things tight. And if you slave the DAW to the ATR (Audio Tape Machine) you can do it all with an inexpensive SMPTE to MTC converter for around 50 bucks or maybe less.

One other note… synchronization is not perfect. There are right and wrong ways to use it. For example if you're laying down a stereo mix that has each instrument placed where you want it in the stereo panorama then both the tracks used to record it should be on the same recorder. That is, if you put the left track on one machine and the right track on another machine synced to the first you will screw up the stereo placement. So while syncing two devices, analog or digital can be very accurate it is not that accurate and that is regardless which device is master or what kind of sync you use.

Well anyway, I need to sit down and mull over exactly how to present this topic to once and for all end the sync mystery. It's all very easy for me because I've been using sync with audio since before MIDI was invented. But there in lies the problem. I have to figure out how to take something that is such a familiar and simple standard for me for many decades now and try to explain it to a new audience that doesn't really understand it at all. Sh!t where's the Tylenol!

 

[miroslav]

Well anyway, I need to sit down and mull over exactly how to present this topic to once and for all end the sync mystery.

Well...it's not much of a "mystery" to me.......I understand the MicroLynx manual very well, and all the SMPTE info I've read. They are very specific, and I think I quoted the MicroLynx manual several times in this thread showing that specific info.

This discussion really needs to focus on open reel deck and DAW synchronization. I doubt when doing film/video/audio in the past it was a common need to do multiple "dumps" from one device to another like is done these days when you have a short track count on a tape deck, but want to record a lot of tracks to tape and then dump them to DAW. Just running two reels with SMPTE, or a VTR and ATR combination, and/or recording more live tracks to an ATR while synchronized with a VTR....is not the same as dumping multiple passes from one device to another. That is really a key piece in this discussion, since that's where/when the drift is most revealed.
Discussing film/video use of SMPTE, while good for reference and overall info, doesn't cover anything about a DAW having to resample in order to keep the audio steady and on time when recording and getting SMPTE from a tape deck that has inherent mechanical fluctuations, wow & flutter, and that is running "free" with no common sync reference that is controlling the speed of the tape deck and the DAW to make sure they are moving consistently, *relative to each other*.

Just having SMPTE on a tape track does NOT control the speed of the tape deck relative to the DAW or the DAW's speed and clock. They are both running “free” with their own internal speed reference, and that doesn’t overcome any mechanical tape fluctuations that also cause the SMPTE *on the tape* to fluctuate.
You seem to imply that mechanical tape fluctuations have no bearing on the actual transmission of the SMPTE that is on the tape.....?

Also...in basic film/video/audio reels....you were not ever able to actually look at the audio tracks like in a DAW...and see that there is drift. Just playing back tracks that have sample-level drift will not reveal that they have drifted any more than 5 guys playing live together.
How important that much drift is to anyone, is something each person can decide.
If it just sounds "good enough" to you...feel free say it's "in sync to me”.

If you can explain why it's not important to have that common sync reference and speed control between devices....that would be a good read.
I mean...why the need for things like the MicroLynx...?...if all that was/is ever needed was just a basic SMPTE/MIDI conversion box between any two devices..?
Heck...I would be happy to strip down my system to nothing more than the tape deck, DAW and a small SMPTE/MTC converter....but that only gets you to “good enough”.

Funny, you now say that synchronization is "not perfect"...well really, I think that's all I've been saying in this thread, and that there are different levels of sync and ways to achieve it!!!
I think I've raised very specific points, and provided technical info right out of sync manuals to back it...so while I know that you have past experience in audio and video...if you have different views/answers to the questions/issues I've raised than I've gotten from a lot of other sources....just claiming those sources are all wrong, doesn't really say anything.



.

 

[SteveM]

SteveM, you are saying that it is working fine, does that mean you are not having to 'nudge' parts of the tracks from different dumps around to get it all to line up perfectly? If yes, then is this just because the song is short and it will eventually be noticeably out of time or is there something here that is controlling the tape deck to play the same speed each time that I have missed?

Wouldn't it be better to record 'through' the Ampex (Record on all 4 tracks and monitor off repro head and feed that signal to the DAW then each additional parts you add you play a long and listen to the other tracks from the DAW and record through the tape deck again and just slide the entire take back a few milliseconds to make up for the delay between recording and playback on the tape deck)

This way, you will get 100% sync on each take (assuming the DAW is stable) and you get to record on all 4 tape tracks.

No, no nudging. I've tried that method and found it's a drag. Most of my songs are short so it's not an issue for me to change anything right now. But I don't understand how there'd be drift anyway because the time code is locked with the playback of the DAW, so if the tape slows down or speeds up at all so do all the tracks on the DAW, at the same moment. Unless your talking milliseconds. The syncman was designed for it and never even occurred to me that it wouldn't work.

 

[miroslav]

But I don't understand how there'd be drift anyway because the time code is locked with the playback of the DAW, so if the tape slows down or speeds up at all so do all the tracks on the DAW, at the same moment.

Not sure how much of the posts you've read.......but I covered that, and described as option #2 a few posts back.
Basically...your DAW is resampling on-the-fly, and if that's OK with you, then you are good to go.
Re-sampling on the fly is kinda' like having variable bit rate streaming of MP3/video files...they are constantly adjusting the delivery rate to account for changing bandwidth....aka speed at which the file can be delivered.
In most cases, your DAW will cover the resampling and you will not notice much....but it's a point of choice, do you want that constant resample happening in order to maitain sync....or would you rather have sync without the resampling.
My choice is the latter.

If you also had the option to NOT let the DAW chase/resample (like my DAW has)...then you will either have the drift (I use to), or you would ***need a sync reference (house sync, etc) and a synchronizer*** (which I now have) ***to control the tape deck's speed*** and to keep it steady/same every pass, in order for the DAW not to have to resample on-the-fly and also not have any drift.

 

[SteveM]

delivery rate to account for changing bandwidth....aka speed at which the file can be delivered.
In most cases, your DAW will cover the resampling and you will not notice much....but it's a point of choice, do you want that constant resample happening in order to maitain sync....or would you rather have sync without the resampling.
My choice is the latter.

Actually, I'd prefer a tape deck with more tracks ! .. but I have to make do for now. Also do not like the tape off the playback head and then shifting trick. It was a drag. At least I can hear this back in real time and if it sounds good then it worked. But what exactly is being re-sampled? All the tracks on the DAW each time they're played back or the stripe?

 

[Beck]

Not sure how much of the posts you've read.......but I covered that, and described as option #2 a few posts back.
Basically...your DAW is resampling on-the-fly, and if that's OK with you, then you are good to go.
Re-sampling on the fly is kinda' like having variable bit rate streaming of MP3/video files...they are constantly adjusting the delivery rate to account for changing bandwidth....aka speed at which the file can be delivered.
In most cases, your DAW will cover the resampling and you will not notice much....but it's a point of choice, do you want that constant resample happening in order to maitain sync....or would you rather have sync without the resampling.
My choice is the latter.

If you also had the option to NOT let the DAW chase/resample (like my DAW has)...then you will either have the drift (I use to), or you would ***need a sync reference (house sync, etc) and a synchronizer*** (which I now have) ***to control the tape deck's speed*** and to keep it steady/same every pass, in order for the DAW not to have to resample on-the-fly and also not have any drift.

This is the misunderstanding of how MTC works I've been trying to clear up over many threads for a while now... that's why I want to give a more once for all exhaustive treatment of sync in general. MTC is a software/MIDI level process that does not effect the sample rate of converters. Word clock is a different story and totally different process, which does control the sampling frequency at the converter level. Again, MIDI Time Code (MTC) works after the signal is converted. There is no "Resampling" per se. This all happens very quickly (microseconds) in the first software layer. MTC works the same way with software sequencers when there is no digital audio to sample. There are three stages in a DAW where speed can be controlled. The first is the ADC, but that requires control of the ADCs clock rate. What you are saying above is mostly correct if speaking of using a master clock rather than the internal clock of the DAC. But the MIDI process involving SMPTE/MTC does not interact with a converter's clock. It's not written that way and has no way to do so. What you have with MTC is a software level synchronization patterned after LTC SMPTE with hours, minutes, seconds and frames per second. It is not as accurate as word clock, but does not have to be for its purpose. It's more than accurate for the task, which means it lacks nothing for linking two audio devices in time.

For people to get their head around it you really have to take digital sampling out of the equation, which is how it was when we were using SMPTE/MTC in pre-DAW times with hardware and software sequencers. I know that is hard to do for people who think of sync in terms of external clocks in the digital realm. SMPTE has been with us since 1969, and with me since 1978. SMPTE over MTC has been with us as an international standard for controlling hardware and computer based systems since about 1987. It is how it was done and if it didn't function properly you took the defective device that couldn't accurately follow MTC back to the store and bought something that could. Yeah, you may have to fiddle with it like you do buffer size and what not, but when all the smoke clears it should work... or it is not complaint with the specifications it claims to support. Slaving the tape deck is one option, but it opens a whole new can of worms... and for most decks it is not even an option. Any tape deck can act as master. Relatively few can be slave. A good DAW with properly implemented MIDI functionality should perform as slave with flying colors. Those that list MTC as a feature, but can't follow it accurately are broken!

 

[miroslav]

But what exactly is being re-sampled? All the tracks on the DAW each time they're played back or the stripe?

The tracks as they are being recorded into the DAW.

Here's an Audio Engineering Society paper desribing a process that corrects the typical Wow & Flutter problem during analog tape transfers.
While the paper is focsued on a new corrective process....the rea point is that there ARE issues with analog transfer that have to be overcome by digital medium doing the sampling. THe DAW is gong to do it's best to "fix" any speed fluctiations during the tape-to-DAW transfer...and it has to resample on the fly to do that.

ftp://ftp.bestweb.net/aes117.pdf

Implicit in the nature of most analog-to-digital transfers is the assumption that the actual recording and playback mechanism is operating at a constant rate; however, due to a variety of mechanical considerations (as well as potentially other causes), this assumption is only approximately valid in practice.

IOW...the audio tracks have "digital adjustments" that were made during recording to keep them at the desired rate. I'm sure there's a better way to say that....but that's what you have when the DAW slows/speeds to keep pace with tape deck fluctuations.



Look at post #2: slave Daw to tape to remove wow/flutter? - Gearslutz.com

analog tape machines have wow and flutter, period, there will ALWAYS be wow and flutter. what people complain about is the jitter. that can happen to the DAW while it's trying to chase the tape deck. The concept with black burst and a machine synchronizer is to make the transport as stable as possible, also some word clock boxes like the aardvark Time Sync II have really great clocks that make jitter be much less of an issue

And post #4:

The main problem with this whole scenario is that wow and flutter won't happen in the same place twice... it's not locked to the tape, but more a function of the motor operation and the mechanics of the deck you're using. It will always happen randomly, relative to the tape playback position... not only is it printed in the original recording, but will also be induced at a different location on playback.

So your DAW will have those tape playback irregularities transfered to it...and in trying to overcome any of those tape speed fluctuations (at small levels, not necessarily very audible/obvious speed shifts)...the DAW still has to do something to maintain your chosen rate relative to what the SMPTE is doing (fluctuating).

 

[miroslav]

This is the misunderstanding of how MTC works....

Beck...

I'm talking about mechanical Wow & Flutter affecting the SMPTE on the tape track of a deck that has NO synchronizer controlling it and NO reference sync...and the DAW having to make up for the speed drift by resampling.

That anomaly is clearly defined in the links above I provided...to include the AES paper.
That's what I'm talking about it.
Even SteveM has witnessed that "when the tape speed goes up/down, so those the DAW speed"....the audio must be re-sampled for that to work.

 

[Chilljam]

When you use SMPTE with a DAW, or what we've always called "SMPTE-Over-MTC" since the late 80's, you have an absolute time reference that is in the same place on the tape at every pass, so there is no drift due to tape speed fluctuations.

This is the part that gets me because I don't see how this applies to SteveM's scenario where he has an Ampex that cannot follow an absolute time reference since it's transport cannot be controlled by an external device - the Ampex can only be followed and we surely we all understand this. The sync reference (SMPTE) on the Ampex is at the same place on the tape at every pass but the DAW is not controlling the tape transport and since it is 'reading' the timecode from the tape that is controlled by the Ampex transport, the timecode is fluctuating DIFFERENTLY each time the tape is played (each subsequent dump) and the best the DAW can do is follow these fluctuations which will not be the same on every dump.

The sync device follows those fluctuations and thus so does the DAW or sequencer on the other side of the SMPTE-to-MTC converter.

This is all fine, I have no issue with understanding that a sync device can be used and the DAW can follow the timecode that is striped on the tape. I can accept that SMPTE is an absolute time reference but an SMPTE timecode printed on a tape that is run by a transport that cannot be controlled externally and is running completely independently of any reference will not play the same every time.

 

[Beck]

This is the part that gets me because I don't see how this applies to SteveM's scenario where he has an Ampex that cannot follow an absolute time reference since it's transport cannot be controlled by an external device - the Ampex can only be followed and we surely we all understand this. The sync reference (SMPTE) on the Ampex is at the same place on the tape at every pass but the DAW is not controlling the tape transport and since it is 'reading' the timecode from the tape that is controlled by the Ampex transport, the timecode is fluctuating DIFFERENTLY each time the tape is played (each subsequent dump) and the best the DAW can do is follow these fluctuations which will not be the same on every dump.



This is all fine, I have no issue with understanding that a sync device can be used and the DAW can follow the timecode that is striped on the tape. I can accept that SMPTE is an absolute time reference but an SMPTE timecode printed on a tape that is run by a transport that cannot be controlled externally and is running completely independently of any reference will not play the same every time.

I'm getting that syncing feeling.

Correct... it will not play the same every time, and this is why sync was invented. The important things to remember are that 1) The SMPTE track is in the same place relative to the other tracks on the tape just as the other audio tracks are to each other and once printed it will never change. 2) A DAW is potentially far better than a tape machine at following another device in time. This is due to the forces present in a mechanical device, such as inertia in a rotating flywheel. As good as it can be it will never approach the speed of correction an electronic device, such as a hardware sequencer or DAW is capable of. But I say "Potentially" because as I've previously mentioned a particular DAW (Computer/audio interface/software combination) may fall short due to flaws in design... just as some analog decks that are designed to follow sync don't perform as well as others. This is all fair game when reviewing a product. Does it do sync well? If the answer is no then look for one that does.

Before digital it was common to sync two or more mechanical devices together, especially in audio/video production. One always has to be the master and time keeper. Good tape decks are precisely tuned machines with negligible speed deviation in synchronization terms... even cassette decks are plenty steady. A tape deck will run within it's wow & flutter specs and speed deviation specs when acting as master. A well-designed electronic device will easily follow the these minute changes.

When you make multiple dumps to digital you always have a scratch track or metronome to tell you where you are, so you aren't blindly recording tracks each time. I mean, everything starts in time and it stays in time as you build and transfer tracks just like multitracking if you were to use the tape deck only. You could do the same thing with two analog decks... as we always have before digital... one master and one slave. It's the same process, but the DAW if performing properly will be more precise in following the time code. The SMPTE or other time code track stays the same as you build tracks, then erase and record new tracks. It's all basic multitracking, but with the addition of a DAW.

 

[miroslav]

I can accept that SMPTE is an absolute time reference but an SMPTE timecode printed on a tape that is run by a transport that cannot be controlled externally and is running completely independently of any reference will not play the same every time.

Right....but I will go so far as to say that just SMPTE on tape is *NOT* an *absolute* time reference on its own, when it's fluctuating due to mechanical speed/Wow & Flutter issues with most tape decks on the planet. Even a brand new, top of the line Studer had some W&F right off the assembly line (though most likely much, much less than lower quality decks).
Unless you have a *sync reference* that drives and holds the speed as steady as possible of two devices when transferring audio...the drift and timing issues between than can vary, and can even become audible, something most easily heard when you do 4-5 transfers of the exact same audio track....then listen for the drift/phasing.

Yes, as everyone seem to agree, a tape deck will not play 100% the same every time......so while it is true that the SMPTE on the tape and the tracks on the tape are always fluctuating in sync relative each other...when you drop 3 tracks to DAW, rewind, drop 3 more tracks, rewind, drop 3 more tracks (which is what the OP is doing)....
...those individual dumps are not 100% in sync relative to each other any more, since there will be a different set of fluctuations at different places, per dump.

So...if the master is fluctuating differently every time, then the slave has to somehow adjust its speed to all those fluctuations.

The DAW reads the incoming SMPTE/MTC and internally translates/ties it to its internal Word Clock, which then speeds up and slows down the DAW recording to help match it to that incoming, differently fluctuating SMPTE from the tape deck.
The DAW has to "varispeed" itself and re-sample on-the-fly, since the Word Clock is changing minutely at different spots to keep pace with tape deck SMPTE fluctuations, and then it will internally spit out a processed audio tracks of a constant sampling rate/speed. That re-sampling on-the-fly and processing can induce jitter/artifacts in the audio which are then permanently a part of the digital audio tracks.
That can be inaudible, for sure...and like I've said a few times, it can be "good enough for R&R"...but it is there, and if the DAW has to work harder (due to old, worn out tape deck, and/or less than flawless capstan/transport design sending SMPTE), there could be more serious digital audio glitches.
To avoid the resamling/varispeed....you can have the DAW not chase and try to maintain sample accurate sync, and again, you may end up with a "good enough for R&R" result with all your tape dumps, but then you will have some drift that you might have to nudge into place along sections of a track(s).

The best case for avoiding the most drift and any re-sampling on the fly when doing multiple dumps, is to have the DAW run as Master and also have a synchronizer/control in-between it and the tape deck, with the tape deck's capstan under full control of the synchronizer (which is actually under the control of the DAW/Master via Word Clock in most cases.

The point here is not that you can't run a tape deck as Master without any synchronizer....you can...but that IS the worst-case sync scenario (it will give you the worst sync)....and just about everyone that has ever done any ATR/VTR/DAW sync will agree with that.....maybe.
If there were not any sync issues between analog decks....then synchronizers would never have been invented.
The same issue exists between an analog deck and DAW.

Here's another guy that explains it maybe in a better way than what I've been saying....read the section on "Analog to Digital Sync" about half-way down.

Articles

If I need to put up more reference links and quotes to support my points....I can.
Even my own DAW describes almost the same scenarios and considerations of tape/DAW sync, and options to have a DAW chase when slave, but with the understanding that then there will be *re-sampling* on-the-fly.

 

[Beck]

The DAW reads the incoming SMPTE/MTC and internally translates/ties it to its internal Word Clock, which then speeds up and slows down the DAW recording to help match it to that incoming, differently fluctuating SMPTE from the tape deck.
The DAW has to "varispeed" itself and re-sample on-the-fly, since the Word Clock is changing minutely at different spots to keep pace with tape deck SMPTE fluctuations, and then it will internally spit out a processed audio tracks of a constant sampling rate/speed. That re-sampling on-the-fly and processing can induce jitter/artifacts in the audio which are then permanently a part of the digital audio tracks.

That is not happening with MTC! SMPTE over MTC is a software MIDI process in a DAW and does not control word clock whatsoever. A little more research on the internal workings of a DAW is probably in order here. That would help everyone to visualize what's going on in the box and at what level.

The best case for avoiding the most drift and any re-sampling on the fly when doing multiple dumps, is to have the DAW run as Master and also have a synchronizer/control in-between it and the tape deck, with the tape deck's capstan under full control of the synchronizer (which is actually under the control of the DAW/Master via Word Clock in most cases.

The point here is not that you can't run a tape deck as Master without any synchronizer....you can...but that IS the worst-case sync scenario (it will give you the worst sync)....and just about everyone that has ever done any ATR/VTR/DAW sync will agree with that.....maybe.

Actually no. Those of us who worked in professional environments where syncing two or more devices was regularly done always ran the ATR as master unless we were working with video, and then the VTR was master. Once using Virtual tracks through MIDI sequencers was common place and later when MIDI sequencing was integrated into DAWs the practice of running the ATR as master continued. This was the standard before amateur recording came to dominate this industry and tried to reinvent the wheel. Much like analog in general, there's an entire history of what was happening in professional studios that is unknown to most recordists today. You really had to be there...

A synchronizer for running the ATR as master with a DAW as slave is as simple as the least expensive sync box ever made , which was the JL Cooper PPS-1 (Circa 1987). I bought mine new in 1989 for $89.00 in Roseville Illinois. You can get them for $20.00 used on eBay now. Before that SMPTE/MTC synchornizers could cost thousands of dollars, and though I used them in two of the studios I worked in, I could not afford anything that did SMPTE for myself until the PPS-1 came out. The PPS-1 did/does sync with both MIDI Clock w/Song Position Pointer and SMPTE/MTC. The former is known as "Smart FSK" because it kept track of where you were in a song. Before Smart FSK you had to rewind the song all the way to the beginning to establish sync again. Smart FSK w/Song Position Pointer was more SMPTE-like in the way it worked. So yes you need a synchronizer to convert SMPTE to MTC before its fed into the DAWs MIDI-In port, but it's very simple.

Here's another guy that explains it maybe in a better way than what I've been saying....read the section on "Analog to Digital Sync" about half-way down.

Articles

Thanks for the links here and in previous posts. However none of them address the details of SMPTE/MTC through the MIDI software layer of a DAW, but rather word clock... and this last one is a discussion about how to remove wow & flutter while transferring to digital, not about synchronization. The OP is looking for a way to fix a bad recording on a deck with audible wow & flutter by using a plug-in or other software. The OP on the Gear Slutz thread also does not really understand what the options are for what he's trying to do, so like in many threads all over the web he doesn't know how to ask the question. And Phillips is specifically addressing sync by control of a device that converts to word clock, though that is not the only way to do it, and certainly not the preferred way to do it. As I mentioned earlier, so-called sample-accurate sync is overkill and not necessary to keep two devices in sync for any type of audio recording.

When a tape deck is the slave it still has wow & flutter... and in fact a tape deck under control of an external synchronizer has much worse wow & flutter than it does otherwise. This goes back to the mechanical compensation I was talking about with slaving the ATR. A tape machine cannot recover as quickly as a sequencer or DAW and consequently the wow & flutter can easily enter the audible as it tries to keep time with the master. This was always a known compromise when slaving ATR to VTR, but slaving VTR to ATR was even worse for the video, so was not done. When sequencers and later DAWs came along this problem was solved with the much tighter sync capability of the computer staying in time with the ATR. Today since the focus of synchronization is on the digital world and master clocks to sync only digital devices much of the information about hybrid analog/digital systems was lost. Many people here are trying to figure it out for the first time, but I've never stopped doing it the way we always did it, so I'm not rediscovering it. My studio is and has always been this way. At any point in time (if you could go back and have a looksie) you'll find the standard practice was to slave the digital machine to the analog machine for a couple reasons. 1) Many analog decks cannot be slaved at all. 2) digital decks/DAWs make time adjustments much faster than analog decks. We even synced Alesis ADAT to analog using the BRC, which had built in capability for just such a purpose. But that was early 90's when more studios had analog and digital working together.

The bottom line is still this: Some people are saying if your DAW cannot stay in time over several passes when slaved to the ATR then you should switch the other way around. I'm saying, no, and that your setup, procedure or equipment is defective and you should get that right. Could be user error, such as time lag from overloading the MIDI stream. Could be bad hardware. Could be bad software. Could be a lot of things, but if you jump to slaving the ATR you only cheat yourself out of a better synchronized digital/analog system and invite the hits even the best ATR will take when controlled by an external device.

 

[miroslav]

When a DAW is recovering to speed fluctuations from tape SMPTE, what is controlling the speed of the DAW....?

It's internal word clock....which looks at the incoming SMPTE/MTC, and then "vari-speeds" the DAW to keep pace.
You seem to suggest that 1.) when the SMPTE/MTC fluctuates from the tape, 2.) even though the DAW chases it, there is no direct connection between the SMPTE/MTC and the internal digital clock of the DAW...???
Sorry...that makes no sense.

Also....when I'm talking specifically about a DAW and digital clocking tied to analog time output....you then talking about what use to be in the analog ATR/VTR days isn't really directly relevant to this conversation other than from a historic perspective and general SMPTE info.
With a DAW, converters, and digital clocks there's another layer that goes beyond pure analog SMPTE. The digital clocks have to interact with the analog SMPTE signals off the tape, which are converted to MTC on most DAW apps (though some better ones can run with true SMPTE time), and then the digital clock drives/vari-speeds the DAW based on that info.....there IS a direct connection, even though you don't set anything, or see anything AFA as the DAW's internal clock ticking/working, and all you see is the SMPTE counter rolling.
Plus....in the ATR/VTR world....it ain't just "SMPTE" driving the car. The speed (clock) is tied to house sync in the form of video black-burst fed to all devices, in most cases and in most A/V situations, where even the smallest sync issue can wreak havoc on the final video product.
Pure analog tape deck to analog tape deck is by far the most forgiving hookup AFA sync is concerned.
But...you already know all that.

Oh...
Not to mention, MTC is even less accurate than pure SMPTE, so just 'cuz it's converted from SMPTE to MTC doesn't in ANY WAY improve the timing, which you seem to imply as I'm reading it....?

 

[sweetbeats]

My Tascam BR-20T slaves to Cubase via my MicroLynx...never more than +/-1 subframe error for an entire 2500' reel at 15ips, and the subframe errors are only occasional and just flickers at that...otherwise the two are lllllllocked.

Its not fair to say wow and flutter increase on the atr when slaving...it is machine and synchronizer dependent. I do agree that lots of machines won't lock up like my BR-20T does, but boy-howdy of one of the main reasons to set the atr as the master is to avoid DC servo cogging and audible artifacts that result, and to slave the DAW because it can make speedier corrections, well then I guess I'm okay...test tones don't sound any different when the BR-20T is freewheeling as compared to slaving and that's where I'd hear it...this is even true when varispeeding the DAW on the fly.

Tim I really do TOTALLY get what you are saying. I really do. But I guess I just don't have the same issues you cite as far as the atr goes when slaving and as a result all my equipment that depends on wordclock for sync gets to follow that precise reference as master rather than be reliant on the atr transport as the sync reference.

I don't see this as a one size fits all deal. I think about Pianodano's setup and it is relatively complex and I don't think it would *work* if his MS-16 was master.

And I also believe I'd be tearing my hair out if I had an Ampex MM-1200 and was trying to slave it to the DAW. It'd work...I just know I'd be seeing a lot more error frequency and margin on the MicroLynx, and I hate to think what it would be like trying to chase-lock on-the-fly. The BR-20T is a thing of beauty to watch...The MicroLynx barrels it in on the sync reference like a predator and once it hits it resolves within a second...maybe 2 at times. I suspect this is the exception and not the rule, but for me it works.

When I read in my Cubase manual that Steinberg acknowledges there are errors when slaving the DAW, and that their solution (in order to maintain the integrity of the digital audio) was to just let it be off, I thought "oh cool...so my digital audio will maintain integrity...and my sync relationship will...wait a minute...be off?" Yep. Yes that is NOT how everybody does it. Not trying to make anything like a blanket statement because that would be just wrong, but I love my DAW interface and Cubase...I've been using it for years and it is like an old shirt and I don't want to change, so I have a tape deck and synchronizer that aptly handle an atr slave scenario.

My 2 pence.

I'm doing the total wussy thing and saying "I'm not saying what's right or wrong, just what's right for me."

Back to you...

 

[miroslav]

My Tascam BR-20T slaves to Cubase via my MicroLynx...never more than +/-1 subframe error for an entire 2500' reel at 15ips, and the subframe errors are only occasional and just flickers at that...otherwise the two are lllllllocked.

Its not fair to say wow and flutter increase on the atr when slaving...it is machine and synchronizer dependent.

Yeah...me too with the MicroLynx and my DAW (Samplitude)....I get the triple L-L-L on my ML....all three Locked.

Did you read the section in the ML manual where it describes the specific SMPTE tape striping process so that your deck actually ends up virtually W&F free?
It is true that if a tape deck has real W&F issues and worn/bad transport design, having a synchronizer push/pull to keep the speed steady can be at the expense of the W&F coming into more focus. The MicroLynx tape striping process works to eliminate it rather than add to it...but you need to go with video black-burst as your sync reference both during the tape striping process and again during the actual deck synchronization...it's all described in the manual.
You can pick up a standalone video BB generator for under $50. I have a Sigma standalone that I used initially, but then I git the Aardvark Aardsync II, with on-board video sync generator...so now I feed that to the MicroLynx.

I often check the counter on my MX-80, and I have set it to show speed instead of time. The speed is shown in XX.XX ips.
When the ML has everything locked, the speed is rock-solid at 15.00 ips.

 

[Beck]

When a DAW is recovering to speed fluctuations from tape SMPTE, what is controlling the speed of the DAW....?

It's internal word clock....which looks at the incoming SMPTE/MTC, and then "vari-speeds" the DAW to keep pace.
You seem to suggest that 1.) when the SMPTE/MTC fluctuates from the tape, 2.) even though the DAW chases it, there is no direct connection between the SMPTE/MTC and the internal digital clock of the DAW...???
Sorry...that makes no sense.

As I explained earlier, SMPTE/MTC is a software level control option and does not effect the clock of the ADC. It probably makes no sense to you because you got a bum steer about how this all works from the start and you're having a hard time reconciling it... which is what you have to do, as many people do. The interweb is full of misinformation, as well-meaning as it may be. Look, I understand this is an advanced concept to most people today. It's not to me because I've been involved with synchronization for so long, while even most of my contemporaries back in the day did not use it enough to completly understand it either. I hope everyone is willing to keep an open mind or that "Aha" light bulb won't have a chance to go on when you finally get it.

Its not fair to say wow and flutter increase on the atr when slaving...it is machine and synchronizer dependent.

It's not a question of fairness, but of simple physics. The mechanical nature of an ATR makes it impossible for it to accomplish even subtle speed variations to maintain sync with another device while under external control of said device without increasing wow & flutter. But you are correct that how much an impact it has while under external control will vary from one machine to the next. Before we had drum machines, sequencers and DAWS to more precisely follow tape than tape could follow them we did (and still do) sync ATRs and that's all we had. But it was always lacking until electronic devices began to emerge with greater reaction speed in slave mode than an ATR could ever hope to accomplish.

What the amateur industry is understanding the least today is the concept of standards, which remained the same from one studio to another so everyone could cooperate. SMPTE-EBU/MTC is one of those universal/international standards for synchronization of various devices, and striping a tape to control all the other devices in a studio is how we did it and still do it in studios where analog and digital are still used side-by-side. When understood and done correctly it works as it should. If it is not working, something is wrong with methodology or product.