MPC3000 Sync to a Computer - Just line up click?

So I've tried a few experiments with syncing my MPC 3K to the DAW. I've tried MIDI Clock, MTC, creating an audio SMPTE stripe, FSK (http://www.rv0.be/sync-unit/legacy/sync-unit-dc), etc..

I only have an Apogee Duet, so for multi tracking, I need things to be tight.

Interestingly, I've had the best results by doing none of the above Instead, just turning on 'count in' on the MPC, and recording a couple bars of click track for each audio track. Then, I just line up the tracks by the click transients.

I then read on the Innerclock site, that MPC 3K internal sync is much tighter than external (depending on which mode you're in -- song or pattern)
(http://www.innerclocksystems.com/new%20 ... itmus.html)

I'm sure some people do this, but as I've read so many threads on sync discussion, this seems like a pretty simple and reasonably accurate way to go,

This works with the SP1200, and I'm not positive, but possibly with the 3000.

If you program whole songs on your mpc and you prefer the feel and drift of an older mpc, you can record the midi that the sp/mpc spits out of its midi outputs into a daw on a midi channel. Then separate the midi notes, use them to trigger your mpc and track out individual parts.

Hm but wouldn't spitting MIDI notes out the DAW be subject to the usual DAW midi timing issues?

viacom wrote:This works with the SP1200, and I'm not positive, but possibly with the 3000.

If you program whole songs on your mpc and you prefer the feel and drift of an older mpc, you can record the midi that the sp/mpc spits out of its midi outputs into a daw on a midi channel. Then separate the midi notes, use them to trigger your mpc and track out individual parts.

dtirer wrote:Hm but wouldn't spitting MIDI notes out the DAW be subject to the usual DAW midi timing issues?

Yes, lining up a click as a timing reference is the best way to do it, IMO.

Then set the DAW to the tempo of the recorded audio, if you want the DAW tempo to match.

Recording into a DAW using sync is unreliable and counterintuitive. Expecting a computer to synchronize individual tracks from a different source when you have your own set of good eyes and ears and can just run a line in never made sense to me.

I always did it "manually" too. Recorded a hi-hat in an empty sequence for each individual track and lined them up manually, never had timing issued like this. It got interesting when recording individual tracks to tape like this, as timing throuhout the recording can vary slightly from the tape for each track (just really slightly), that sometimes gave things a nice groove...

To be clear, what I meant by this is not to use your Daws timing, or tempo. Just record the midi in (not quantized and not synced to the daw) so you have a recorded snapshot of your full sequence/song. Think of it more like your recording audio with no grid or click. Essentially your capturing a live performance from your mpc. Then you can separate the midi channels accordingly and drop your individual tracks one by one all being triggered from your "snapshot". This would be most effective if your recording a whole song sequenced on your mpc. If your just doing a couple bars then don't bother.

Most daws now have pretty high ppq. Logic has something like 960, but if your still worried about accurately capturing the midi, just double your tempo in your daw, or set it high and forget the tempo in your daw altogether. It's not going to line up anyway.
Of coarse, you would still get some "artifacts" from your daws ability to capture midi recordings, however minimal compared to your mpc (96 ppq), but at least your tracks will be getting triggered from one accurate and consistent source when your tracking them in.

You stated that it has to be tight, so keep in mind none of these older mpcs have perfect clocks, and never play the same twice, even if your only recording 2 bars. Sure the 3000 sounds tight when all playing together, but it drifts. If you line up all the start points of your individual tracks to the start point of a 2 track, by the end of the 2 bars nothing will be precisely in time with each other or that 2 track. To get it truly tight to the 2 track (down to the sample) you'd have to edit all the kicks snares hats etc individually one by one, start to finish. If your tracking out a whole song this is a huge waste of time.

I don't really like to use my daws modern grid, it's too perfect, and what he mpc does when it's just playing on it's own, all together, sounds better. Drift and all. the old slower computers of these machines and their inconsistency is part of their charm, and add to the human feel.
IMO If you don't have the budget to get a high quality 8ch interface and can only use 1 or 2 inputs per pass, this is one faithful way to get your tracks into your daw in time with each other as close to the original single performance of the MPC as possible.

dtirer wrote:Hm but wouldn't spitting MIDI notes out the DAW be subject to the usual DAW midi timing issues?

viacom wrote:This works with the SP1200, and I'm not positive, but possibly with the 3000.

If you program whole songs on your mpc and you prefer the feel and drift of an older mpc, you can record the midi that the sp/mpc spits out of its midi outputs into a daw on a midi channel. Then separate the midi notes, use them to trigger your mpc and track out individual parts.

viacom wrote:To be clear, what I meant by this is not to use your Daws timing, or tempo. Just record the midi in (not quantized and not synced to the daw) so you have a recorded snapshot of your full sequence/song.

Then, as questioned, you have the jitter in the computer USB MIDI timing effecting the timing of your tracks, both on the way into the computer, and the way back out of the computer, which is the problem the OP wants to avoid.

viacom wrote:Of coarse, you would still get some "artifacts" from your daws ability to capture midi recordings, however minimal compared to your mpc (96 ppq), but at least your tracks will be getting triggered from one accurate and consistent source when your tracking them in.

No, you also get USB MIDI jitter, going into the computer via MIDI, and back out via MIDI, it does not give an accurate and consistent source.

If you are going to do that, you would probably be better off loading a MIDI sequence from the MPC into the computer, and use that to triiger the MPC, at least then you would only have USB MIDI timing effecting the timing of the track in one direction.


All this would have to be tested, but the clocks in older MPCs like the MPC3000 possibly give you better timing over multiple passes than the timing jitter introduced by going in and out of a USB MIDI interface.

Your points are valid Simon.

To be fair, I think the recorded jitter going into the Daw would probably be captured across all midi notes.. So you might not get too much inconsistencies between tracks there. I could be wrong.
Your right about jitter affecting the tracks differently to track parts pass by pass using a slow USB interface. But over long Periods of time ie the course of a 3 minute song, I think the drift of an mpc would be considerably off next to the jitter from a USB midi interfaces which might only be in the realm of samples off. How much this matters really depends on how precise you need it. And let's face it, if your that concerned about it, you (no one in particular) should probably save your money and get a proper 8ch interface.

I'm not saying this is the ONLY way to get some measure of consistency, and nothing is going to be perfect save editing every single sample after the fact, but it's definitely one way of doing it that will get you in the ballpark quickly.

All good.

My assumptions about MIDI jitter affecting timing, and your assumptions about drift between MPC audio output over time, are just assumptions, they really need to be rigorously and properly tested to determine what the actual affects on timing are.

Well according to the Innerclock site (and that guy seems to have spent a whole lot of time measuring , the 3K has jitter as low as 3 samples (depending on which mode you're in). I've done some tests over 32 and 64 bars (maybe not entire song lengths) but the jitter never seems to deviate from that. (5 samples in pattern mode)

I suppose sending MIDI in and out from the computer seems like it would add more timing issues, but I'll certainly give it a try.

A larger interface would obviously be ideal, or I suppose a Sync Gen. But was just curious about others methodologies.

EDIT:

And also, I know the timing won't be perfect and I don't necessarily need that. But it needs to be good, and not 10s to 100s of samples like I've experienced when sequencing hardware from the DAW.

The smart money is on keeping it simple.

I got one of the first sync gens and it was one of the best purchases ive ever made. never could get sync tight enough with modern computers( not true about my old amiga and st) and was having to fix way to much stuff and make trade offs I wasnt happy with. and you get din outs for your non modded roland boxes so that is cool too.