how to make 32 voices feel like 64 (or 8, for that matter)

i have seen cats crying about "i have bug lol my mpc drops notes"
when the prolem is they just don't use their mpc rite.
polyphony is a crowded desert watering hole, and your shit has to wait its turn.

EDIT: le rat suggested that we clarify some diction here.
a sample can be MONO or POLY
it can also be MONO or STEREO
when referring to a sample's channels, it's MONO if it has 1 channel, and STEREO if it has 2 channels - LEFT and RIGHT. this means that there are 2 simultaneous streams of audio playing at the sample time.

each stream of audio takes 1 voice. a stereo sample takes 2 voices, one for each stream.

in PGM mode, there are settings to make a sample MONO or POLY. this doesn't refer to the channels, it refers to its "voice" type. MONOPHONIC pads will cut themselves off if played twice before the first instance stops playing. POLYPHONIC pads will play overtop and overtop itself.

you have 32 voices in the MPC. use them wisely and it should be enough, especially if you've got a computer to do some of the extra work.

how to make 32 voices feel like 8:

* use long stereo samples (esp hat samples) and leave them set to POLYPHONIC playback

* trigger loops with a single midi note (will automatically reduce your poly by however many channels the loop is, and since the loop is your oldest note, it'll get dropped when you hit your polyphony cap

* use crossfade loop willy nilly (tho odds are you're not a dunce if you're using CFL and you won't have poly prolems)

* set a long release on your chopped samples

how to make 32 voices feel like 64:

* set snares, kicks, and hats to mono. in fact, set most pads to mono some of yall record w/ quantize off. sometimes your mpc records 2 notes 1 tick away from each other and this will eat 2x the polyphony - we're talkin 8 voices if you were using a CFL'ed snare. hungry sumbitches. if you set your shit to mono, the lower velocity hit will cut off the higher, and it'll be obvious, and you will have to edit.

* trim the tails of your longer samples, don't just ADSR out the stuff you don't want. if you have a ride with several seconds of silence or near silence at the tail, that shit will GULP your polyphony faster than fat chick does diet cokes. esp. if you didn't follow bulletin 1.

* DONT put all your drums on diff. tracks. youll proly get sloppy and lose track of your shiz

* use a different pgm for every chopped sample including drum breaks (but NOT including layers). this way you can either set the pgm to mono (NOT for drums) or better yet use mute group all smart-like. for your chopped samples, if you're playing them back in the order they were chopped, set all yer even number pads to mute group 1, and all your odd number pads to mute group 2. this way, the adjacent pads won't cut each others' tails off, but instead they will kill the penultimate pad. if you don't believe me...

* solo your tracks and watch your polyphony meter. you'll see if a pgm is using more voices than it should, and you can find out why.

EDIT: ocular's tip, a good one:

"Also worth mentioning is targeting pads to be muted by a master pad rather than blanketing pads under specific mute groups. This can be more flexible in certain cases while still keeping the polyphony count in check."

special note, there doesn't even have to be any sample assigned to the empty pad to use it as a mute trigger. however, if i'm chopping a loop from vinyl, i like to drop some vinyl noise on there and modulate its play start position so that instead of just dropping to digital silence, it drops to a more realistic sound.

Great advice.

You been here before?

been lurkin for a year or two, mostly just to track down bugs
but decided to start posting because this place isn't what it used to be
not much discussion on technique anymore

That's true.

Keep it coming.

Awesome post.Any stereo sample tips?

Good stuff, Crothers, and thanks for coming out the shadows and contributing Polyphony seems to trip up a lot of the newer/ non-technical users.
Also worth mentioning is targeting pads to be muted by a master pad rather than blanketing pads under specific mute groups. This can be more flexible in certain cases while still keeping the polyphony count in check.

nice, ocular. i added that into the original post.

TYPO, regarding STEREO, these tips can all apply
first, however, make sure you NEED stereo
if the sonic characteristic of you sample is different on both channels, keep it stereo
if it's similar, but just sounds quieter on one side than the other, drop it to mono and recreate that diffference with a pan and save yourself up to 2 voices per trigger if you're using CFL

I have a different workflow so I don't share all of your thoughts but if your thread is dedicated to newbies let me add the basic knowledge.

We need to make a difference between the mono/polyphony setting of a program and monophonic/stereophonic samples.

16 stereo samples = 32 voices of polyphony if played at the same time at the beginning.

As suggested it's better to make some samples monophonic if needed.
In general my personnal preference is to have a POLY program and to have several MONO settings at the pad level inside that program.

One useful tip for saving polyphony is to resample the MAIN OUT of the MPC.
That's another story but I've noticed recently that SIMULTING long samples sometimes makes the result sounds slightly different when sequenced over lots of bar so I'm more and more into resampling my stuff.

PEace

We need to make a difference between the mono/polyphony setting of a program and monophonic/stereophonic samples.

thank you, i'll edit my first post with this information. i just kinda jumped on in there without easing the new folks in, huh?

In general my personnal preference is to have a POLY program and to have several MONO settings at the pad level inside that program.

yeah. i find setting a program to MONO is too severe. better to keep pgm poly and use sneaky settings at the pad level as you said.

regarding SIMULTing long samples, i have not heard of this. would you explain your technique?

I was too tired to edit my post one more time earlier but there's the mono/polyphony setting of a program and the setting of individuals pads inside that same program.

I think there are very subtitle differences at the pad level setting if the program is POLY or MONO. I don't remember well but I thought I wrote a short post regarding that when JJ introduced the feature. Anyway there is nothing to care or worry about. It's just a detail.

My current workflow is almost based on simulting stuff. For instance I've noticed few times when simulting a 4 bar loop with the same low-passed loop there is sometimes a very slight change in the simulted sound once I've rendered the song. I've read on that forum that the MPC wasn't sample accurate so maybe it's because of that? I don't have much knowlegde regarding that subject so when i think there's a strangeness in my sound I resample.

Well maybe i'm too nitpicking with my stuff and i'm focusing too much on details. That's not excluded