DAW chasing Tape (Reaper and MS-16)

M

man-bot

New member
I finally had a real reason to try to get my DAW to chase tape (reaper is the DAW and Tascam MS-16 is my machine).

Had a 15 track digital recording to dump to tape... so I figured I try the tips I was given and pulled out my JL Cooper PPS-1 and thought I would give this a whirl again. (Sad trombone)... PPS-1 was dead.

However I'm pretty jacked right now as I got it working without a go between box!! (other than my digital interface)

Not sure if this is only with Reaper as I don't use any other DAW (and I hardly use reaper).

I striped the tape using Reaper's timecode to track 16 of the MS-16 (at about -3 level) and then ran an out from the MS-16 to input 1 on my interface. Right clicked on the play button in Reaper to enable Reaper to look for timecode before playing and set it to look on input 1 and BAM!! Rock solid... started on time, if I FF or REW it would jump to that spot and lock almost instantaneously... didn't drift over the 3 minute test I did.

This was a big PITA when I tried it before and this was as simple as it could be.

So - if anyone wants their DAW to chase tape, get reaper and try this setup... I had it working in 3 minutes.

(FYI - I need to sync because my digital output is only 8 tracks but I can mix all 15 tracks at the same time if I dump to tape and mix through my board... would be impossible to line up the first pass with the second pass of dumping tracks to tape without this).

Hopefully it will help someone!
 
First off...if you are happy with the results...then by all means don't change anything. :thumbs up:

That said...now I'll tell you why that setup is not really a "sync" setup...and also some potential issues it can create.

A true sync setup implies a single clock that drives everything.
Your setup has two different clocks, and they are not tied in any away. You have the internal DAW clock, which is basically your computer's clock...and the MS-16 has a clock (probably a PLL/crystal deal) that drives the capstan. I'm not sure on that deck if that clock has any impact on the reel motors. IOW...is it just telling the capstan to spin at a set speed, or can it drive the entire transport...?
Regardless...that MS-16 clock, has no direct connection to the SMPTE on the tape.
In a proper SMPTE setup, where the transport was fully microprocessor controlled...the internal clock would be read by the processor AND you would feed the SMPTE off that track BACK to the deck and to that internal processor...that way, the clock and the SMPTE are in sync with each other.

Keep in mind...SMPTE is NOT your reference clock...it's simply a time/position reference.

Next...your DAW with its internal clock is not in any way tied to the deck's internal clock...it's only reading the time/position reference off the SMPTE tape. That may work in general, if there are no fluctuations...but, it may also drift at the sample level, which means that since the DAW is chasing, any sample has to be accounted for internally by either skip/drop or re-sample on the fly. Those sample "skips/drops" may be very rare, and from a "good enough for Rock & Roll" perspective, easily ignored...but, for some situations, any sample skip/drop or forced re-sample is not acceptable...and in some, it may even be audible, which certainly would not be acceptable in any situation.

It is cool that Reaper is able to read SMPTE directly from the audio source as opposed to some MTC/SMPTE conversion...though ideally, you would want the DAW to be master, since that way there is never any chance of sample skip/drop/re-sample....and let the deck slave. However, the deck would then need the same clock source, which there is no way to obtain from the computer.
So...the cleanest, most accurate way is to have a single clock that feeds both deck and DAW..and then the SMPTE is read by both, and either can be master, though again the DAW as master is optimal...but it requires that the deck is able to chase/lock to the DAW, for which you need a deck that has a microprocessor that can do that.

Again...if you are happy with the results...you are good to go. I just wanted to cover some key points about "sync" and "good enough sync". You have the latter....which in your case, with your gear, is all you can do.

Sometimes I wish I didn't know all that, as I have spent a good amount of research/learning/experimenting with a few different DAW/deck sync setups over the years, and I've gotten many a migraine getting it all understood and clear in my head...but I wanted to have a true synchronized system (which I now do)...so I had to fill my head with all that nonsense (and that was the very short version)! :D
 
My parade is feeling wet :)

I'll have to test it to see if/just how much it drifts. While I don't grasp your post 100% miroslav it is clear you have spent a lot of time and thought on how to do this rock solid. Admittedly I'm hoping this works well enough for me. Really though any drift will be an issue.

I'll experiment over the coming weeks and let you know how I fare.

Mike
 
No...it's not "bad news"...just saying it's not an absolute/perfect sync method. :)
It can "work"...how well/acceptable...you just have to try it and see.

If you look at the Reaper sync settings...the stuff about drift/re-sample and skip/drop becomes clear.
Look at the sections that mention drift/re-sync/skip/drop...which shows that Reaper has tolerance limits.
When you have a tape deck just running on its own...outputting SMPTE off a track...with no direct connection to the internal clocks...you have to work within min/max tolerances.

static1.squarespace.com.png
 
Last edited:
I am tempted to try it. Right now I'm happy with my home made poor man's clasp system
 
Keep in mind that samples are very small compared to milliseconds...and we musicians can "ignore" several milliseconds without it screwing up our timing. I think up to 5ms is almost not noticeable. I mean, when we play we drift some.
You get about 44 samples/millisecond at a 44.1 sample rate.

The more common issue maybe the amount of total drift from start to end...especially if you're doing a longer song.
Also...if the DAW has to skip/drop samples to stay in time and you pick up an audible artifact...then that's another thing that would be a problem.

You may find on some songs...it's just fine..on others, the feel is off...on another, you might hear an artifact...so if you have no other options, I say go for it and see what you have.

Also, when you are mixing down...and you keep rewinding and starting over many, many times until you get your mix...each pass may be slightly "off" and different from the previous. How bad, how off...it's going to be minor...and you just have to see what is acceptable/noticeable...or not.

I didn't mean to put you off your method...just wanted you to be aware of all the considerations.
 
:) - it's all good! Based on what you have stated above I wouldn't be any further ahead if my PPS-1 was working as I don't believe it provides a clock.

Ultimately my intentions are 1 pass - dump 8 tracks, then dump another 7... theoretically it should help me get everything lined up at the start perfectly. If it goes all wonky... well it's usefulness is significantly reduced.

I agree it is nowhere as stable as the solution you present - but it also involves no extra pieces of equipment which is pretty cool in my books.

I will likely try a recording method in future where all audio except vocals are on tape and vocals are "poor man synced" on digital for editing purposes (it's easier to comp). I'll see at this point just how valid a tool it will be.

Mike
 
When I was messing with that stuff early on...I could get things lined up decently...but when I would do a couple of sync dumps, and I included the click or the drums in both, as I find those tracks make lining up the easiest....that's when I say the sample drift issues.

At regular view...everything looks good...but when you zoom in deeper at sample level, you start to see the drift. The transients from the two click/drum tracks don't all line up. You can line them up perfectly at the start...and then as you scroll toward the end...the drift comes into play.
What kind odd at first is that at times you see the drift in one direction...like track two will be falling behind track one...then they start to line up again, then track two goes a bit ahead...then back again...etc...and that's caused by the tape not being locked directly to the DAW. IOW...the SMPTE looks good, but the tape transport fluctuations drift in/out in small increments.

Mind you...much of that is "good enough for Rock & Roll"...and only in extreme cases do you actually start to hear the drift between tracks. So it's all about what's acceptable. Like a song where you have big guitar strumming and lots of re verb/delay and everything is kinda "chugging"...a tiny bit of drift is not a concern....but if it's a very percussive/tight song, then it could be an issue.

So these days...even though I have a rather involved DAW/tape sync setup...and I always lock them when I'm transferring tracks...
...ever since I move to the 2" 24 track deck, my transfers are usually single pass. So in reality, I don't even need to worry about sync. I could just dump the track on the fly into the DAW. However, I like to sync/lock everything just in case I decide down the road I want to redor any tracks or add more from the tape.
back when I was running of my 16 track deck...I would often do two dumps from tape to DAW...sometimes three...so under those circumstances, track drift was more of a concern.

I think you will be OK with what you have for the most part...but just keep eye and ear on it, and be aware....that's all I was trying to say here. :)
 
Actually Miroslav you are not correct. What the user is trying to do is do two sets of 8 dumps. Now consider this. If he could dump all 16 at once, he would have a clean dump of all tracks, but the tempo would be slightly offset from the original recording by the fact that the machine was slightly off speed when recording and when replaying. But this is not a problem. Now he has striped a track with accurate SMPTE. He does the first pass with Reaper syncing to the SMPTE. Then he does a second pass with Reaper synchronising to the SMPTE. If Reaper is synchronising correctly the 2 dumps will be synced. They will still have the variations from original speed caused by the tape recorder, but the two dumps should be in sync. If the user recorded SMPTE when he did the original recording, then the synced track would be exactly the same pitch as on the day of recording. If you are using this method, and the tracks dont sync, then it is Reaper that is not synchronising correctly.
 
radardoug said:
Actually Miroslav you are not correct.
Click to expand...

I agree with Miroslav here, the problem is that the DAW is chasing the SMPTE time code which is printed on tape so the SMPTE is controlled by the tapes transport which is not synced to anything.
 
radardoug said:
If Reaper is synchronising correctly the 2 dumps will be synced. They will still have the variations from original speed caused by the tape recorder, but the two dumps should be in sync.
Click to expand...

You're assuming that the variations will always be the same because of the SMPTE track....but, as I said earlier, the SMPTE is not a clock reference, and it has no control of the tape deck or any direct connection to it's tape speed/capstan...the SMPTE is just the timeline/position reference.
The DAW and the tape deck have NO common *clock*...and....the tape deck's clock and the SMPTE tape track have no direct "connection". There would need to be an on-board microprocessor that can both read the SMPTE off the tape track from a loop-back SMPTE signal, and read the speed of the capstan/motors...and then control one based on the other.
Instead...the SMPTE on the deck is just running "wild" at the mercy of the tape's transport steadiness.

On the DAW side, the internal clock and SMPTE do have a connections...which is why it's better to have the DAW be master, and because in doing so, it won't be at the mercy of the tape deck's scenario described above.
As I mentioned in my original post...a single reference clock the both DAW and deck are following is the ideal solution.

If the tape deck transport and internal clock are pretty solid...not having that connection to the SMPTE on the tape track can be less noticeable, but while the tape clock is driving the speed of the tape/capstan...it has no way of knowing what the SMPTE is doing, and therefore adjusting that capstan/tape speed (talking about micro-adjustments) to always keep the tape speed and the SMPTE time, internally synchronized.
You need a deck that can do that with its own microprocessor or an external sync box that can directly control the tape speed and read the SMPTE.

Not to trying to spam...:)...but in my thread here about the selling of my Fostex G16...that tape deck has the microprocessor and the optional sync card to do what I am describing.
My current setup with my 2" deck...I use an external Timeline MicroLynx box that is tied to the deck and has full control of its capstan, while it also reads the SMPTE off the tape track...plus I've got an external clock box that drives both the MicroLynx and also my DAW/interfaces...so that's what keeps it all on the same, sample accurate page.

Again...you can get by without getting that involved...but there is some sacrifice in the sync...which may or may not be acceptable, depending on your needs and your audio material.
 
I'll just want to add this to the above...don't lose your mind over this shit! :D

I'm a bit nuts and obsessive at times with some of this stuff, and for me, a lot of it was an exercise in seeing how far I could get, and as best as I could take my setup...but TBH, back when I was not quite there, the sync was rarely bad enough to make the music suffer.

Heck, if you were just going to sync a couple of tape decks...a few frames drift wouldn't be critical.
It's the DAW thing and knowing how picky digital audio streams can be if they are trying to "adjust" to a fluctuating analog source.
Also...if you are going to sync audio to video (analog or digital)...then you have to be rock solid, then you can't have frame drift...because you could end up with a '70s Kung Fu movie, where first the mouths move, and then you hear the words. :p

I've also done a couple of years in a TV studio environment and some multimedia development work...so I have too much of the "pixel anal" mentality in my head...which can be obsessive, but not always really necessary for doing some Rock & Roll tracks.
 
Yes, in fact it would be better to have the tape recorder slaved to a computer running the timecode. That way you pull the tape recorder to correct speed. However the problem is that Reaper is not syncing correctly. If it was there would be no problem. It would be interesting to see if Protools was any better. I doubt it.
 
radardoug said:
Yes, in fact it would be better to have the tape recorder slaved to a computer running the timecode. That way you pull the tape recorder to correct speed.
Click to expand...

Though again...there is NO WAY that a DAW can "pull the tape recorder to correct speed" if it has no direct control of the tape recorder's transport, which is not a simple thing to achieve. It requires a dedicated synchronizers, and also a single clock reference feeding everything, as I mentioned above...so for the basic home setup, that tends to be out of reach.
 
Actually you can buy Zeta 3 synchronisers on Ebay very cheap. This was one of the best synchronisers and just requires a simple cable to the machine. If you buy one and want the cable details contact me.
 
Thanks, I don't need anything. I'm using the TimeLine MicroLynx synchronizer in my setup, which were the common studio sync solution back in the day when everyone was still rolling tape (and I have a few of them...rack units & keyboards..gotta protect my interests)...so I'm good to go. :)

I made my own com cable from the synchronizer to the deck. I have cable schematics for most of the devices that the MicroLynx supported, in case anyone is looking for that.
 
I'm having a bad night. Thanks for this thread. It's helping to take my mind off of things a re-center a bit.

I know there are kind of two camps of sync ideology but I could never be comfortable with one of them because one of them causes audio information to be dropped. And I'm not okay with that.

One method is like you and your buddy are going for a walk and you want to stay in step with each other, only your buddy pays no attention to you, and spend your time keeping up, sometimes having to do a little skip to change your footing, to either speed up or slow down your step to stay in sync. You're the DAW chasing the tape machine.

Now how about a scenario where all the audio is preserved and things are locked together and referencing a reliable time source, like you and your buddy now doing a three-legged race where your legs are strapped together so you keep in step, and some guy who looks like Gandolf coaching you when to step...that's more like a setup with a common reference clock source for DAW and tape machine, and a device that controls the speed of the tape deck so it is *locked* to the DAW. No music is lost, and the two players never drift.

Sort of a dumb analogy.

TimeLine Micro Lynx is the shizzle. I have a primary unit and a backup unit, and two keyboards...I have the word clock card for it and even the 3rd machine card.

It's the "right" way to do sync, but "right" isn't always realistic or necessary. I just can't stand the concept of chopping up my digital audio.
 
Just as a follow up - this method does seem to work. I've been able to get 15 tracks on to tape in what sounds like perfect time.

On ocassion the computer will lose track, the audio drop out momentarily but comes back in perfect time. So I do have to watch the transfer and note if there are any dropouts and go back and punch them in in sync. Sometimes it doesn't happen at all.

So conclusion - it works. Maybe not perfectly, but for my purposes it works just fine :)

Mike
 
You must log in or register to reply here.
Back
Top