Forum to discuss all matters relating to the MPC1000 and MPC2500 operating systems created by 'JJ' (all versions).
User avatar
By arebee Wed Nov 07, 2007 3:56 pm
hey, ive been on about this for sometime, and i get buffed off each time by unconvincing arguments.....

the track mute page is cool, i use it all the time, allways have done, but lately im experiencing wierd issues. some sequences grind to a halt, that never used to do this (in fact some of them were written in older JJos's and showed no signs of this problem - otherwise they wouldnt have got created in the first place).

the problem is one of voice poliphony, like how many samples the mpc can play at once yeah?

ive got a 4bar seq., 6 tracks have breaks on them (all in oneshot/mono voicemode)....all other tracks are empty.

if i play all tracks at once (unmuted) the mpc can do this, and the voice monitor shows up to maximum of 10 voices being used up overall in the seq. so basically my entire usage at the maximum point is about a third of the voice limit.

as soon as i mute even one track!!, or all of them - voice monitor gradually rockets up to the limit of 32 voices, and the sequencer begins to slow under the pressure (ultimately -unstopped my mpc will crash).

now, ive been told that the extra voices used are due to the fantastic 'mute/stop of oneshot sample:immediate' setting (mode+track mute+'window'). and that its so that the track can be actually playing (using up voices in a silent mode) in order to be immediately unmuted.... i totally get this idea, but why is the poliphony use flying off into the maximum before 4bars is even reached? as i understand it, the most voice use i should hit is the same as when all my tracks are active right? (in this case just 10 voices).

i know that i can turn off 'immediate' and have it the easy (keep my mouth shut and grumble to myself) way, but then my oneshot breaks arent being cut up as i had intended - and i said this earlier - and the most important argument i think - these seqs were made with this purpose at some point a while back, and the mpc didnt behave this way when i first made them (in an earlier JJos) otherwise they would never have got finished..........so i know that the code for it was working well at some point i the os history.

so why am i being robbed of polyphony? im getting to the stage of back tracking - convinced im gonna find an os version that didnt do this, then stick with it.....until JJ finds this bug n flattens it.

gizzed off rB)
User avatar
By arebee Wed Nov 07, 2007 4:03 pm
yo, its deff a bug, cos in os480, the seq plays perfectly, and the polyphony (voice monitor) only peaks at 11.

i fkin knew it.....y'all saying ' nah thats the way it is - its sposed to do that'

JJ cmon sort it out dude.

am currently moving up thro' os's to find the one where it went wrong.
rB)

By stale bread Wed Nov 07, 2007 6:11 pm
i for one appreciate your tenacity over breakin all this shit down, please keep on pushin.
User avatar
By Bugfix Wed Nov 07, 2007 7:53 pm
I didn't notice that issue, because from some point I have all pads set to "note on" everytime
(for many reasons, including this poor immediate/through thing) but,

maybe and this is only a blind shot now, when you mute tracks events don't respond to mono voice overlap setting and they are in fact playing as poly?? causing polphony robbing.
Which would be crazy and definatelly a bug. I'll check this.
User avatar
By arebee Wed Nov 07, 2007 9:59 pm
the problems start at 485 and above,

for one, im def staying at 484 now, it works just fine, polyphony monitor looks good, doesnt peak out, mpc doesnt grind down n die anymore, and hey, i still get to use the most popular 'cult of JJ' removal yet (drum roll...)

PAD MODE!..... hah, laughing now, until that is - when i check what bugs im gonna have to stomach they fixed since 484 up to 491.....

but for now im happy - at least i can work on what i was focusing on lately.

but JJ's gonna have to pul it out of his hat sooner or later, a super clean ""bug free"" 'uberOS'.......im here, gonna be waiting on that one.

pce
rB)
User avatar

By arebee Wed Nov 07, 2007 10:08 pm
its mad actually the difference, i dont know why but in another sequence ive just loaded (one i wrote a while ago - probably in this era os) and i tried to push the limit to breakin point in os484, and nothings denting the processor.

ive got voice mointor up to 32 loads this time, and the sequencers not even straining...in later os's i barely get to 24+ before the sequencer clock starts to slow n struggle.
User avatar

By hiphoppa Thu Nov 08, 2007 5:24 pm
arebee wrote:its mad actually the difference, i dont know why but in another sequence ive just loaded (one i wrote a while ago - probably in this era os) and i tried to push the limit to breakin point in os484, and nothings denting the processor.

ive got voice mointor up to 32 loads this time, and the sequencers not even straining...in later os's i barely get to 24+ before the sequencer clock starts to slow n struggle.


Hey Arbe, I see all of your issue's (nOOb here) and I feel what you are saying. At what point can I study the voice polyphony as a visual guide? Are you in program mode seeing it or something else?

Thanks,

Gary
User avatar

By arebee Thu Nov 08, 2007 11:43 pm
since jj os 48x's theres a realtime graphic meter called 'voice monitor' in mode+pad10 ('other' mode), it shows which of the 32 voice engines are active at any time, so at a glance u can assess how many voices u r using up, and theres also a couter that shows the peak use.
User avatar
By arebee Thu Nov 08, 2007 11:46 pm
jj replied to my emails today to say theyve checked and understood the problem and will have it fixed in the next update.

awesome, i hope theyve finally nailed this one for good, was a slippery customer i'll say.