Timing optimize mode for JJ OS2xl?

Hi Guys,
I'm wondering what steps I can take to tighten up the timing on my MPC1000 with OSxl when tracking out to the DAW? The timing is 'tight enough' during normal usage but I'd like to get it super tight once I have it in the DAWfor mixing.

I usually track out just one track at a time by muting all the other tracks and this seems to tighten things up considerably, but I'm wondering if there are any other proven steps I can take to improve matters. I've tried turning the all the FX off but it doesn't seem to make any great difference. I have mutes set to events vs. volume as this seems to have caused people grief as well.

Tips? Tricks? ...or perhaps it is hopeless.....

thanks in advance for your time!
patchen

As someone suggested in a topic about timing the other day, you can sync and track sequences separately instead of the whole song. I think it could help too.

Yeah, I do just one track of one sequence at a time (not in song mode), I also don't bother with sync just in case that might cause any anomalies. (setting the click to count in only in playback is a great way to line up tracks in the DAW btw).

Seems like I get drift in the order of a few milliseconds, not sure of the exact amount offhand but i seem to remember it being +/- 3 or 4 ms or so for each hit, which if one is early, then the next is late could equate to what, 8ms error between hits?
I know the MPC is not sample accurate so therefore the timing will never sound quite as precise as creating beats in the DAW will but I'd like to optimize as much as possible.

Seems like I remember my old MPC2000xl being tighter but it's been some years. You know what they say, we sometimes look at the past with rose tinted glasses.

You seem to be an expert so do you mean you don't recommend stuff based on MIDI Sync (I was thinking about MTC)?
I know with JJ we can customize the metronome sound, there's no such timing issue with the click metronome replayed by the MPC?

Sorry to be a little off-topic regarding your original question.

This smells of a version of JJOSXL below v2.01. Make sure you don't use any JJOSxl below version 2.01 and above 1.38 because the timing is messed up in those versions (when switching sequences).

@jimmie:
I'm not talking about switching sequences, just playing back a single looping sequence. I'll check when I get home later what version I am running but I think it is the most recent one, but I remember the same issue in XL1.47

@Le Rat:
To be honest, I have never synced the MPC to anything so I couldn't speak as to how it behaves in
That situation.
I've just found out soemthing very revealing. I just recorded a very simple test sequence with five tracks that cause the mpc to peak at around 28 voices (as can be seen from the global screen)
I then recorded some fairly active q link pitch bend on one of the tracks. Doing this alone is enough to get some timing instabilties that are audible, if only just.
Now, I set both the fx processors to reverb and whooa nelly., the damn thing just about chokes completely. Sounds like a frickin' rubber band is stuck in the clockworks!
So, now I know for certain that using the fx processors can affect timing, I"m wondering if there is anything else I can disable to relieve strain on the cpu as well... I can upload a project file if you want to see what I'm talking about. Perhaps it's worth bringing up with jj directly as we'll...

Don't worry I trust you and yes you're right the effects are what is stressing the CPU the most.
It seems there was timing and stability issues from os2xl 1.38 but just wait and let's see when bugs on os2xl 2.XX series will be fixed.

But yeah you can send a project to JJ if you want him to check for potential issues.

By the way the bug Jimmie makes reference to was a problem in MAIN screen, NEXT SEQ and SONG mode so it's worth upgrading if you didn't already.

confirmed in OS 2xl 2.05. Timing issues persist when maximum polyphony is used up and and both filters are engaged. Timing becomes severly unstable when one reverb is selected, mpc1000 freezes completely when 2 reverbs are engaged.
Bug report submitted to JJ.
I post scripted a couple feature requests as well.
1- a setting that limits the amount of polyphony a program can use. So for example the options for a program could be; MONO, POLY, POLY8, POLY4, etc.
2- an offline function that 'thins' controller and q-link data to make it less dense.

These functions could be outside the scope of what can be done on the mpc, but the don't call him the Jenius for nothing!!

Stilling listening if anybody has any other tips on how to 'turn cpu hungry shit off' to tighten up the timing...?....?

IMHO, we now have some AWESOME features in OS2xl, perhaps it is time to stop adding new features for a while and to go back and do optimizations to the OS so that we can always have a tightly running machine. Perhaps more features that focus on letting us 'turn off' things we don't need that take up cpu.

It's late and my head hurts, i could ramble for hours but I'll just stop before i get too far behind....

patchen wrote:confirmed in OS 2xl 2.05. Timing issues persist when maximum polyphony is used up and and both filters are engaged. Timing becomes severly unstable when one reverb is selected

man im relieved,,i was starting to think that i was driving nuts here,,as i mostly use external fx but sometimes i added a reverb here and there,,so the changes on the timming were not too obious but u could feel them when listening the sequence meanwhile working on synths..damn i was really concerned haha

well i guess i rather stop using the internal fxs and stick on the externals until we have this bug fixed.. and try not to reach max poliphony

yeah of course we have lots of new features.. i love the multitimbral and mixer cc control..but would be good idea now to focus on timming stability

I did some real tests in the DAW instead of just listening the other day, the only time I've found where timing is actually affected is when you have used up polyphony, have reverb set to one slot and another effect like 4band Eq to the other slot and then switch to the pad mute screen. This causes sloppy timing. I think the other timing issues i was 'hearing' were really just inconsistent note stealing happening because of all the voices i was using.
The other issue of putting reverb in both fx slots will definitely make your machine croak if you use up all your voices so don't do it!

When the timing is solid, it's reaaally solid, which is nice to find out. I was only seeing drift of max 1-1.5ms in between beats, usually much less than that. That's plenty good enough for me and waaay better than you are going to see using midi with hardware synths, etc.