Analog + Digital syncing question(s). HELP!

[mquilling]

I'm playing around with some techniques, and phase, as always, is the bane of my existence.

I've recorded some tracks using Logic and my MBox (not-so)Pro.
I'm using an 8-track Yamaha cassette multi-track as a poor man's summing mixer if you will, or just to get some different tonal characteristics. I record the already recorded digital files, verb bus, or what have you to tape, and then send that back into Logic and lay that back in to the mix. And as I (and i'm sure anyone else reading this) know phase becomes a real b*tch, to say the least!

I seem to be able to nudge somewhat close to in-phase at the beginning of the track, but about 2 minutes into, or specifically during backing vocal passages towards the end it seems to be drifting further and further into near-vertigo inducing phase problems.

I'm thinking this may be due to slight tape speed vs. actual time differences that merely become more apparent over time. I'm pondering cutting up the digital file of the tape track into smaller sections, and trying to sync that way, but i'm open to any other suggestions.

I realize this is a somewhat awkward, and to this point not so desirable way to work, but I'm stubborn and determined. Yes I realize I can "emulate" analog, and so forth. But I want to try and make this work, however laborious. I like the "glueing" effect of this technique (minus phasing) a lot and think it's worth figuring out.

Are there any tricks anyone can suggest for syncing in this manner? All you seasoned engineers out there, any of you run across this problem 20-25 years ago?
Are there any digital tools that may help in this scenario?
What did engineers do when time-clocks weren't available?

Thanks for reading, and if you posses any helpful thoughts or knowledge on this please help me slay the beast of phase yet again.

[kslight]

Ideally you would want to transfer the tracks to your computer simultaneously, you are probably dealing with the mechanical limitations of your tape deck. I have experienced this same issue on various levels of tape machines if I transferred to digital but not all tracks at once.

If you are bound and determined to use your deck with an Mbox (limited I/O) then what I would suggest is putting a marker hit (cowbell maybe doesn't matter) at 0:00 on every track before you transfer, and maybe even at every minute if the tape deck is drifting that bad, then put everything to tape, bring it back into your DAW, and manually sync away.

I have a feeling however that this problem would not be so noticeable if you took in the stereo out of your deck into your DAW instead of individual tracks, or if you transferred every track to digital at once (that is the only way to know that the deck is playing at the same speed and it's minor variations which cause time issues).

[sessionsatstudiom]

Like KSlight said if you can transfer the tracks at the same time then you would not have syncing issues.

Another way to do it would be to get a midi interface that can generate SMPTE time code. Put time code on track 8 and make your yamaha the master. Then run the output of track 8 to the MIDI interface timecode in. And sync Logic that way.

in version 2 there still might be drift etc... because the cassette tape is the most consistent format.

Also the Cowbell trick would work but still might have drift.


Just a few thoughts.

[ricey]

i don't think there's any sync strategy that will deal with drift.
but, as you're using logic, maybe you could automate the Sample Delay plug in?
not that i've tried it, it just came to mind...

[mquilling]

Thanks for the input. KSlight-I'll try both suggestions and see how they work. Otherwise I may try tracking down a midi smpte generator. I'll post my findings on here once I get it to work.

[MoreSpaceEcho]

i'm not sure what your method is exactly, but try this: send the tracks from the computer to tape deck, monitor the tape in repro. get the sound you want. while you're sending tracks to the tape, record them back into the computer at the same time. i.e. don't record tracks to tape, rewind, and then transfer back to the computer.

this *should* help keep the file in sync from start to end. i would do the cowbell thing too, but use something higher pitched, like a triangle, something with a really sharp, well-defined transient, as its much easier to line that up exactly. put it at the start and end of each file.

[Z-Plane]

Keep the phasing in as a boutique analogue process! I'm only half joking as I think you might be scraping the bottom of the barrel with this one. I defy you to make a qualitative judgement on multi-tracking to tape-vs-dumping your digital master to tape a couple of times.

Since DAWs first emerged I have tried almost every combination of DAW-to- tape, from doubling up tracks on a Fostex B16 to mixing back from MTR90s. Sometimes the results were pleasing but very rarely contained any wow factor in large amounts, and none of the multi-tracks ever really seemed worth the time. I loved working with tape, but I gotta say sometimes its just tape, no magic guaranteed. I salute your lunatic quest and would second that repro head idea, although unless you are happy heading into lo-fi hiss, don't expect mr. yamaha to deliver anything special.

[Tragabigzanda]

The repro head technique won't work, as the Mbox isn't capable of sending and receiving audio at the same time. Would be great if you got another computer and A/D setup.

I think that triangle at each track start makes the most sense as a starting point (and I would recommend a four beat count-in, just for diligence sake). The SMPTE box could be a good, relatively cheap investment, and you'll be able to keep using it, or sell it with little difficulty, later down the line.

Or you could just invest in a CLASP system.

[MoreSpaceEcho]

The repro head technique won't work, as the Mbox isn't capable of sending and receiving audio at the same time.

really? i've never used one. how the hell do you overdub?

[Tragabigzanda]

ODs are no problem, because the headphone out works; it's the mic/line combos that can't handle simultaneous I/O. At least that's true of the original Mbox, not sure if it changed with later models? So the OP couldn't send his bass track out channel 1, hit tape, then come back in channel 2.

[jhharvest]

You don't actually need to buy a separate SMPTE generator box, if you have no other use for it. You really only need a pre-generated recording of the SMPTE LTC signal (such as the ones from here).

Run that to a track on the multitrack deck while you dump your audio tracks. When you play it back you need your DAW to jam to the TC track. Of course this limits you to seven audio tracks at a time (or maybe six, depending on how much crosstalk there is between the tracks - LTC doesn't sound very nice).

[Tragabigzanda]

You don't actually need to buy a separate SMPTE generator box, if you have no other use for it. You really only need a pre-generated recording of the SMPTE LTC signal (such as the ones from here).

Whoa, I had no idea this site existed, much less even considered this an option. Great find!

[Tragabigzanda]

How would you set up your DAW to jam to an LTC track without a SMPTE box?

[Gregg Juke]

What KS said; if you're determined to do it, then probably hand-syncing by mouse-click is what you'll have to love. (Although Scott's idea, and JH's website are beyond my brain's processing ability right now-- I'll have to download those PDF's sometime and pretend I'm Einstein).

This kind of thing has come up in video-post before, and you just have to account for drift over time by re-syncing.

GJ

[jhharvest]

How would you set up your DAW to jam to an LTC track without a SMPTE box?

You can use SMPTE reader.

(Let's ignore the fact that I didn't consider that Logic doesn't have SMPTE LTC support and I meant to post that tool to begin with... )

---

Edit2: Nevermind what I said at all. Turns out Logic does do varispeed from MTC.

Edit: In fact, now that I think about this I'm pretty sure this won't work for the simple reason that Logic mostly likely won't jam even to MTC. It might sync to it at the start of playback but that's it. For this to work you'd need to output word clock based on the LTC. The same problem would exist with external conversion boxes too.

There could be a very hacky way of doing this. There is libltcsmpte that reads the timecode. Robin Gareus, who has done some work on jplay2, has also written a Jack example application for the library. Jplay2 itself can varispeed based on Jack transport. I'd guess Mr. Gareus would have a solution on how to take your LTC wav and audio wavs and spit out new audio files at the correct speeds. Very hacky.