MPC 3000 LE Polyphony

Hi everyone,

so the MPC 3000 LE is said to have 32 simultaneous playback voices.

Nevertheless, i found out that there are some small shifts between sounds played back together, and those shifts, or the order in which the sounds are played can even change according to midi notes number of the pads, pad mode (simult/normal) and on which track the timing correct function has been used lastly. (Of course i am not talking about shifts due to swing settings).

For most situation, let say a clap on a snare or a hat on a bongo.., different kind of sounds, it doesn't matter.

But when layering same families of sound it is enough to be embarrassing.

By example, two kicks played at the beginning of a bar can be heard not really together but with a shift between their attacks...

Is there a way to actually play several sounds together or "sounds combination" has to be done elsewhere prior to sample them in the MPC?

Thank you!

DM

This is the word famous groove of Akai mpc3000.

I think it's unrealistic to expect sample accurate timing on an MPC3000 with a sequencer resolution of 96 ppqn. I don't think polyphony is your issue, but unrealistic expectations.

But if the nuances of the machine don't suit what you want to achieve, layer those samples in a DAW and flatten or resample them prior to use in the 3000.

Ultimately with the 3000 you have got what you paid for so don't be unrealistic and get on with it like every other 3000 owner. It is what it is I'm afraid...

By example, two kicks played at the beginning of a bar can be heard not really together but with a shift between their attacks...

Having transients match may guarantee accuracy, but it certainly doesn't resolve any issues of phase.

I layer kicks and snares all the time in my MPC 3000. I've never had this problem you describe.

not every sample is chopped the same, sum could be chopped closer to the sound than others.

i think u need to start tuning your drums or learning to filter/layer, its a phase.. you'll get over it.

Thanks to all of you for your answers!

delisious wrote:This is the word famous groove of Akai mpc3000.

Of course maybe it is!

clint246clint wrote:I think it's unrealistic to expect sample accurate timing on an MPC3000 with a sequencer resolution of 96 ppqn. I don't think polyphony is your issue, but unrealistic expectations.

But if the nuances of the machine don't suit what you want to achieve, layer those samples in a DAW and flatten or resample them prior to use in the 3000.

Ultimately with the 3000 you have got what you paid for so don't be unrealistic and get on with it like every other 3000 owner. It is what it is I'm afraid...

Yes i agree with you. For now i do all layer in the daw and then resample them in the MP

However i don't see where there the sequencer resolution can be involved as all the sounds are played at the same position, and those shifts are smaller than one ppqn anyway. So it could not be corrected by shifting the notes position too.., and even if it could be the order in which the sounds are played can change.., problem remains

Shostakovich wrote:

By example, two kicks played at the beginning of a bar can be heard not really together but with a shift between their attacks...

Having transients match may guarantee accuracy, but it certainly doesn't resolve any issues of phase.

I layer kicks and snares all the time in my MPC 3000. I've never had this problem you describe.

I would like to be able to say the same lol, I might do something bizarre

TreCipher617 wrote:not every sample is chopped the same, sum could be chopped closer to the sound than others.

Yes of course, but what i describe is not related to that. It's actually one sample is played slighly before the other(s).

padlock wrote:i think u need to start tuning your drums or learning to filter/layer, its a phase.. you'll get over it.

Yes of course! These are the phenomenon heard, i hear the phase and i am able to correct it.

In one point of view, what i am talking about right now could be seen as a special function of the MP lol

It is all under control. It is possible, in a strange way, to choose the order in which the MP plays the sounds, and then get as many different soundings combinations of the same samples as the number of sample used in the actual layer! The MIDI Note number of the pads define the order! I swear it lol.

Still it's surprising that the MP doesn't play them at the real same position

With respect, I would suggest this is user error.

I verified last night. Two sounds set to trigger from the same pad and recorded on separate channels did indeed start at the same time - both visually on the waveform and aurally.

Thank you to have taken the time to do a test, then i must be doing something wrong, because the same test showed me different start time, both visually on the waveform and aurally

How about using the same sound copied onto different pads? That would negate any question of start times and attack settings.

In fact, how about recording the MIDI out of the MPC into a MIDI sequencer or DAW and checking offset between the two MIDI notes? Run the sequence for at least 32 bars to see if there is any further drift or whether the delay is consistent.

Report back.

Shostakovich wrote:How about using the same sound copied onto different pads? That would negate any question of start times and attack settings.

In fact, how about recording the MIDI out of the MPC into a MIDI sequencer or DAW and checking offset between the two MIDI notes? Run the sequence for at least 32 bars to see if there is any further drift or whether the delay is consistent.

Report back.

Hi Shostakovich,

thank you for your reply.

So i recorded into my DAW a 32 bars sequence of the same sound assigned to two differents pads.
First pad is panned full left, second pad is panned full right, and the two sounds appear with a delay on the waveform. I recorded the MIDI OUT of the MPC too, and there is an offset between the two notes too.

As i expected it, the MIDI note who has the highest MIDI note number and his assigned sound come first in the recordings.

I took pictures of the waveform and the MIDI notes so you can actually see the shifts.

Here is the waveform:

Here are the midi notes:

Thank you again for your help.

Whatever it is don't let it ruin your life, it's a machine.

Yes of course!!
As you told me i do it in a DAW now, but i was wondering if i do something wrong with the machine

Even without the XY values, that's quite a considerable offset.

So i recorded into my DAW a 32 bars sequence of the same sound assigned to two differents pads.
First pad is panned full left, second pad is panned full right, and the two sounds appear with a delay on the waveform. I recorded the MIDI OUT of the MPC too, and there is an offset between the two notes too.

OK, is the offset consistent throughout the 32 bars? IE, is it the same at the end?

What about sync? I see from the grid you must be syncing the DAW/MPC. Whihc is slave, which is master? Do you get the same result the other way around?

Shostakovich wrote:Even without the XY values, that's quite a considerable offset.

So i recorded into my DAW a 32 bars sequence of the same sound assigned to two differents pads.
First pad is panned full left, second pad is panned full right, and the two sounds appear with a delay on the waveform. I recorded the MIDI OUT of the MPC too, and there is an offset between the two notes too.

OK, is the offset consistent throughout the 32 bars? IE, is it the same at the end?

What about sync? I see from the grid you must be syncing the DAW/MPC. Whihc is slave, which is master? Do you get the same result the other way around?

Hello, yes the offset is consistent throughout the 32 bars and i always get the same result whatever which one is master or slave. Anyway, even if i use the MPC alone i hear the offset between the sound.

Pad with highest MIDI note number comes first.
Simult mode breaks the rule: main pad comes first.
This is for notes recorded on the same track.

If the notes or recorded on differents tracks, the notes of the track recorded lastly come in second postion.


It's hard to me to believe that only my MPC is doing this.., it seems so much like "that's the way the machine works", but i wanted to share this with other MPC user just in case i'm doing something wrong that can explain that