Why doesnt it sync?

again, could you clarify on your exact technique of what you're doing?

are you recording these one at a time or all at once?

ive done it all.

1. I record a 8 bar loop, 4 times, 4 takes, into SF. then I cut all the 4 recordings at the same spot on the wave file, at the begining, the first hihat.

then I import all 4 tracks into cubase. into different tracks. and the I zoom in... and then u see how the tracks are all "different". so each take of the same loop is recorded diffrently,

ex. 1 snare might be 0.5 milliseconds before or after the other one etc etc... all the drums dont line up perfectly when u zoom in super tight...

as u can see in this picture...

http://img325.imageshack.us/my.php?image=sync12oh.jpg

2. ive also tried syncing the mp with cubase, the sync works, but when I record, its the same problem as above. ^^

Ive come to the conclusion that something is messed up with my mpc, since I just record straight the same stuff, but they are all different....


peace.

and since all takes are different, I cant use cubases delay compensation option.

is the zoomed picture showing the beginning of those files, or somewhere in the middle?

does the error seem consistent as you go forward in time (ie. the 2nd file is always a little ahead or whatever) or does it drift around?

What happens if you try recording from the mpc's multiple outputs into multiple inputs on your soundcard? do those line up then?

And I'm still unclear on the first method you tried,, how did you cut them at the beginning? (you said at the first hihat) ie by eye? Using 'snap to grid'? By zooming in?

Anyway, Im 99 percent sure there is nothing wrong with any of your gear.
Rather you are seeing the normal inaccuracy of midi synchronization.

Its very very common to see errors of +/- 1 millisecond or even more when it comes to midi sync. Mostly this is to do with the Operating System of your computer, they are not well designed for tight signal timing. (Audio is different, i won't get into it.)

In fact, if you told me you had a pc midi interface that was guaranteed to be off by 0.5 ms at most, I would buy that for a couple hundred dollars .

On the upside, the human ear cannot detect so small an error, afaik. I have seen research that suggests 5ms as a threshold, so even being conservative I think maybe 1ms is the absolute minimum where anyone could hear the difference. Phasing, however, is a valid problem if you need to have the same samples on different tracks for some reason...

The zommed is somewhere in the middle.

its not consistent, its like all mixxed, sometimes the 1st is ahed, somthimes the 2ed, etc.

I dont have a multiple soundcard, I got the Juli@, im recording from the digital outs on the mpc to the digital in on the Juli@

this is how i cut it.

http://img215.imageshack.us/img215/52/sync32yr.jpg

http://img116.imageshack.us/img116/1406/sync46ay.jpg

so they are all cut at exatly the same spot, the first hit.

the thing is that it sould not be about midi, since im not using it, im just recoding straight into my wave program. no midi, just record and play.

I know its not a big thing, and u probably wont hear it, but when ur recording alot off samples, and they all dont fit the way the should, it will change the feeling, the swing off the song...

peace.

Okay, now I'm getting what you're doing.

If I understand you right, what you are measuring is the fundamental timing inaccuracy of the mpc in playing back notes. Looks to be +- 0.5 milliseconds.

Now we all know the mpc wasn't 'sample accurate' so basically that's how inaccurate it is. And I'm not too surprised,,, honestly 0.5 milliseconds is a respectable figure.

ps. the mpc was running off its own clock, not synced to anything, right? (at least with your sound forge test?)

ps. the mpc was running off its own clock, not synced to anything, right? (at least with your sound forge test?)

yes sir.

mm ok. so its just the mpc then?

could u do me a favor? could u try to to do the same thing I did, and see if u have the same inaccuracy?


peace.

could nobody else do it? please, it would be a big help

holla.

I've done it a thousand times..... tried EVERY setting! still the same problem

Hello,

I am lazy by nature so I haven't tested it yet

However I did do this quicker test which concisely illustrates that the mpc makes small random timing errors all the time::

http://www.mpc-forums.com/viewtopic.php ... 6&start=30

so don't worry, your machine is not broke.


But remember, one half of a millisecond is _absolutely not_ detectable by the human brain as a rhythmic deviation, so its really no biggie. It is only detectable in the sense of frequency cancellation, such as the phasing test I described.

It would be nice if the mpc series was sample-accurate but it is not. Just learn to not care. Listen to some old Kraftwerk:: their rhythm sounds tight and I'm quite sure they were using hardware sequencers with an error of at least 1/2 millisecond.