SMPTE/MTC Sync?

[hatbeardglasses]

After trying out the various common DAW's, it was so refreshing to find this program, which most intuitively accomplishes almost everything I've wanted to happen. Things I've banged my head against the wall trying to figure out in other programs have come far easier in Reaper.

One thing I'm really looking forward to is the ability to slave Reaper to SMPTE or MTC, and thus to a tape machine. Is this in development and soon to come?

thanks again for your fantastic work.

[Justin Frankel]

Originally Posted by hatbeardglasses

After trying out the various common DAW's, it was so refreshing to find this program, which most intuitively accomplishes almost everything I've wanted to happen. Things I've banged my head against the wall trying to figure out in other programs have come far easier in Reaper.

One thing I'm really looking forward to is the ability to slave Reaper to SMPTE or MTC, and thus to a tape machine. Is this in development and soon to come?

thanks again for your fantastic work.

This is something we're planning on doing.. though it wouldnt be full sync, it would just be triggering from timecode, and perhaps resyncing if it gets too far off (but that distance would be upward of a second or two).


If your tape machine syncs to SMPTE or MTC, REAPER can send it.. including SMPTE LTC (timecode as audio)..


-Justin

[The Spark]

I have to say if it would ever be possible for Reaper to sync to incoming SMPTE time code, that would be amazing. Like if you could designate a certain input on your interface as a SMPTE input and then have Reaper chase the SMPTE, that would be a great feature!

[Airon]

If you're using tape machines, you're always looking to sync everything to it. The machine control may come from a central unit or the computer, but everything hooks in to the tape unit.

All the transfers I've always done, be it to Umatic, DA-88/98, DAT or digital VCR-like devices(VMOD for example), we've always slaved the audio machine to the most unstable source. And that's always a mechanical device like the tape machines.

The Fairlights varisped their playback in the most precise manner. I know it's hardly possible with a DAW, but syncing to external timecode delivered by LTC, VITC, MTC, SMPTE, Bi-Phase or networked rewire connections will be a very big plus for Reaper in the future.

Btw, Reaper may find a universal use for simple laybacks to Digibeta, HD tape and perhaps even Dolbymod(if it functions as a dubber well enough) if you give it timecode in/out capabilities.

I'd certainly love to see Reaper being a high performance dubber. But for that it needs to be able to sort out machine control input and run from timecode.

[dawhead]

Originally Posted by Airon

The Fairlights varisped their playback in the most precise manner. I know it's hardly possible with a DAW,

This isn't true. Ardour can sync to incoming MTC to within about 100 samples of the master, using continuous varispeed coupled to a DLL. It locks in about two MTC/SMPTE frames.

I'd certainly love to see Reaper being a high performance dubber. But for that it needs to be able to sort out machine control input and run from timecode.

You might consider Ardour for this purpose, especially since it does destructive recording if asked to (something I don't believe Reaper does). Harrison is now selling the X-Dubber specifically for this purpose, which is just Ardour running on multi-processor opteron system with their network I/O that carries hundreds of channels plus timecode, machine control etc. See this URL for photos and details of that system:

NAB: "Proof of Concept" Linux system, plus tracks from Fox's "24"

[Airon]

I wouldn't mind a dubber like that for any weekly or feature. For dailies and many other high throughput projects where in-the-box power is plenty, it's probably going to be no-fuss high speed mix exporting like Nuendo(or the incoming Reaper), or simple printing with Protools systems.

Nuendo 4 seems to get rerecording, and thus live printing, which is what I do with Protools all the time right now. But still nobody's giving me function group automation arming and pass oriented automation recording. That's still Harrisons thing, but who knows what N4 is doing.

Congratulations on getting Ardour up to such a level btw. Harrison must think very highly of the project to put it to such use.

[thenoodle]

I'm not sure what Justin is saying when he says the plan is to make Reaper only trigger to midi and not "full sync".

For any type of pro integration, a daw needs to be able to sync (chase and lock)to at least incoming common smpte frame rates (via available timecode inputs available at the sound card) , even if via midi time code. That's the requirement for slaving a daw to tape ..or to other daws ....or outside synchronizers or anything else.

I'm not sure what Justin is stating in the reply to the slave sync question or why the code would be such a big deal to implement in a couple of days. After all, EVERY major piece of recording sequencer/audio recorder has the ability to slave sync at least at the frame level, if not sample. This goes for Cakewalk/Digidesign/Steinberg...AND....my twenty year old ATARI 1040s with old old old Mastertracks programs AND the old ancient Doctor T KCS programs from last century...AND every standalone black box recorder made from yesterday back to 1987.

They all slave and they all do it well. After the number of requests that have come in to Justin, I'm thinking he believes it's harder to implement than it really is.

[Justin Frankel]

Originally Posted by thenoodle

I'm not sure what Justin is saying when he says the plan is to make Reaper only trigger to midi and not "full sync".

For any type of pro integration, a daw needs to be able to sync (chase and lock)to at least incoming common smpte frame rates (via available timecode inputs available at the sound card) , even if via midi time code. That's the requirement for slaving a daw to tape ..or to other daws ....or outside synchronizers or anything else.

I'm not sure what Justin is stating in the reply to the slave sync question or why the code would be such a big deal to implement in a couple of days. After all, EVERY major piece of recording sequencer/audio recorder has the ability to slave sync at least at the frame level, if not sample. This goes for Cakewalk/Digidesign/Steinberg...AND....my twenty year old ATARI 1040s with old old old Mastertracks programs AND the old ancient Doctor T KCS programs from last century...AND every standalone black box recorder made from yesterday back to 1987.

They all slave and they all do it well. After the number of requests that have come in to Justin, I'm thinking he believes it's harder to implement than it really is.

I do hope you're right!

So how does other software handle when it ends up off by a frame? Do they drop frames or insert silence?

[plush2]

Originally Posted by Justin Frankel

I do hope you're right!

So how does other software handle when it ends up off by a frame? Do they drop frames or insert silence?

Usually there are options regarding this in the software. I would think you could drop frames when ahead and use your slick varispeed to catch up when behind or just use varispeed for both. This kind of compensating would be the most tape-like in behavior and the easiest to make minute adjustments with I think.

A comment from Dawhead would be nice as he is the one here to most recently implement a really slick software sync. Most of us haven't done that since....well, ever.

[dawhead]

Originally Posted by Justin Frankel

So how does other software handle when it ends up off by a frame? Do they drop frames or insert silence?

ObDawHead: basically, you run a DLL that constantly adjusts to the apparent speed of the master. when enough frames have been dropped, you have to flag an error, but its not really a problem to drop one or two MTC quarter frames. that said, i've yet to see any system where you drop any MTC "packets", ever.

for SMPTE, transmission is even more reliable (its just a single channel of audio), so the errors are really all at the master end, and that ol' man DLL, he jus' keeps on rolling along ....

the hard part of sync is handling the many varieties of Frames Per Second settings - drop, non-drop, pull up, pull down and so forth. i was very happy to deputize the development of that code to a guy who totally understood it, having implemented it before for Jazz++ (the first and oldest MIDI+audio sequencer for linux) and a few other systems.

this all assumes you can do global varispeed. i don't know if reaper can do that sort of thing.

[MsM]

I'm setting up a system where I sync an MTR to a DAW. As far as I could read from this thread

sync to blackburst vs sync to word clock

post #5, in this case Reaper should be the SMPTE timecode master. Is it video/post use specific to have the tape machine as the TC master? Or is this a different view on chosing sync TC master? To me it seems that slaving the DAW would indeed introduce jitter, as the post referenced above says, but I'm just starting experimenting with this.

Thanks
M.

[otto.r]

AFAIK the proper way to sync tape to a DAW is for the DAW to be master, and the tape machine to be slave. That keeps the sample rate of the DAW constant. The other way round is a kludge.