how NOTE ON timestretch is possible in REALTIME in jj os

now with a new, improved, consumer friendly thread title!!!!!!!

so this came to me jsut a sec ago in the prod. tech forums, but i'm gonna post it here cuz this is a bitchin tip for the jj os since we can control entire pgms via qlink.

ok. first, set you qlink to note on, layer, and track.
set each pad so that each layer is triggered by a different position. this can be done by setting the bottom to 1 and the top to 4, and the middle 2 to 2 and 3, or you can just divide 127 by 4 and do it that way, whatever. that's the time consuming part, so i highly suggest that muhfckrs make a default "layer" pgm like the CHROMATIC pgm that's kicking around.

here's the next step. for the example, lets say you have a nice and easy sample to chop - 16 notes, all 8th notes so they're the same length. 1 for each pad.

chop the original sample and assign it to the pads on layer one. this is where most people stop. however, we're better than that. here's where it gets really fckin cool, while still preserving polyphony.

now go back to the original sample and timestretch it by 20 so the grain is JUST noticeable. chop this sample and assign it to the 2nd layer, so that layer 1 is the orignal chop and layer 2 is the timestretched chop.

now go back and timestretch the original (or the already timetretched sample depending on how funky you want to get) by another 20, making it 40, the grain is defined and audible. chop THIS sample and assign it to the 3rd layer. ya'll see where i'm going.

do it one more time by 80 bpm or so, so the grain is just rediculous. assign this to the 4th layer.

notice how you can control the timestretch in realtime by sweeping the qliink and tapping the pads! it's not fully realtime since you can only edit your sample's timestretch by an interval of 4, predefined by the timestretch difference between each sample chop, but record a normal 1-16 in a sequence and then go back w/ a qlink sweep overtop, bottom up or top to bottom. notice how the qlink will slide the entire pgm to a different tempo, while preserving the programming!

ALSO, you can get much deeper with this trick using simult. simult effectivel gives you 4 more layers, so divide 127 by 8 instead of 4. simult each pad to its corresponding pad in B rather than A.

realtime timestretch son. awesome effect, totally possible on the 1000. takes some time to do, but highly worth it.

you lost me almost from the very beginning. All the layers and stuff was just too much for me. I'm still confused with the q-link on the 2500. Besides tune and velocity, most of the other perameters don't do anything when I try to make adjustments.

people should NOT be sleeping on using layers not as a way of playing 2 sounds at once, but as a way of making 1 pad control 4 sounds

this is perhaps the most slept on feature. roxstar - sit down and learn it. you can figure it out in an hour or so, and you'll get better each time

i'm not gonna let this go obscure, this is one of the most powerful things i've learned in a hell of a long time. where's distorted, tbonus and gunmeNtal!!? where's my deep programming brethren? give this a shot, you'll trip balls

nicely thought out

i'll take on your challenge. brb.. reading

i see how this would work... but i dont see a use for this in a mpc context... thats a lot of pre-production for what intended use? an example being; you sample a phrase, you chop it up into a program... you take it the extra step by assigning layers etc... and save this as your pgm 'instrument' so that you can timestretch it to any projects tempo in realtime? i suppose this would be pretty handy for a sequence that contains a tempo change & sample phrases.

I think this is a neat trick, but as you said, it's not realtime timestretch, and one issue is that you have to use up considerably more sample memory than than a true realtime timestretch system would. If you have the memory to throw away that doesn't matter, but the main difference is that there's no automatic link between sequence playback tempo and the switching of the samples.

If you could get JJ to implement 2 parameters on a single slider (i.e. tempo and layer switching) you might get a bit closer to a "real fake", but then you also have to get the whole tempo range to line up properly with the samples you've stretched, and obviously if the tempo is tweaked in between the timestretch sample tempos it all falls apart.

But like I said neat trick, I'm sure there's some limited application for it...

realtime timestretch son. awesome effect, totally possible on the 1000. takes some time to do, but highly worth it.

Worth it? Realtime timestetch is available on practically all major digital audio recording software...just a few clicks of the mouse versus a time consuming method on the 1000???

jahrome wrote:

realtime timestretch son. awesome effect, totally possible on the 1000. takes some time to do, but highly worth it.

Worth it? Realtime timestetch is available on practically all major digital audio recording software...just a few clicks of the mouse versus a time consuming method on the 1000???

worth it? of course it's worth it!!!, you don't base the value of something on something else, my favorite program does this much easier but so what
this tip isn't about my favorite computer program it's about the mpc. the dumbest thing a person can do is look a gifthorse in the mouth and talk shit while knowledge is being dropped and if you know everything in the world already but you don't know how to 'just listen' or respect a m/fcka then you don't know shit.

Some good stuff there Antonym. Some of you may not have seen this: -

http://www.youtube.com/watch?v=--GzodmsZtI

G4s

Thanks Nym!

I was reading down the post wating for Jahrome to pop up saying 'Hey, this sucks / I love Akai / My MPC2500 is bigger than your MPC1000 / Pong is ghey'.

And there he goes!

I think 'you can do this on a computer' is a new rhetorical low, even for him. Well done, Jahrome.

Can't wait to see his reaction when JJ introduces the 'full sized propellant actuated device' update he's promising for v5

"full sized propellant actuated device" ?

what's that?





and btw, dont mind jahrome, he's a clown

this tip isn't about my favorite computer program it's about the mpc

spot on
i'd rather do this in the mpc than in any other program since as of yet i don't have a standalone wave warper program and i dislike loading ableton solely for that purpose.

used more subtly, this trick is very powerful for note variation. ESK used it (before i did, even! i've only thought about it so far!) and sent me the song, a really dope beat. he did a lot with pad layers, but also some layererd timestretched stuff was going on as well. i didn't notice at first but going back i could hear some quiet timestretching going on but it wasn't all the time - it's kind of like synthesizing note variants, different hits of the same original source sound.

this helps a lot with keeping a tired sample fresh. as long as it changes a little in key ways (filter, velocity, attack, decay, effect, or in this case, timestretch) you preserve the sound. it's all in being a good programmer.

now i'll address some issues brought up by credible members.

one issue is that you have to use up considerably more sample memory than than a true realtime timestretch system would

very true. however, it's still not enough memory to have to start worrying about anything. plus, you could always get even more grime in there by having the original loop pitched up and then just pitching down all the layers, old sampler style. memory is the least of my concern, i'm more worried about the amount of "preproduction" as someone else said.

it'd be nice if jj would give us more placement options after a chop, ie layer, pad bank, etc. that would make this a breeze. however, i feel like it takes maybe 5 minutes or so. if you make yourself a "LAYER" pgm it's even quicker. i highly suggest doing this if you intend to start using pad layers to their fullest.

electrik muz was talking about actual tempo changes. that's something that i was not thinking about necessarily for this trick, i was more thinking about keepinng the seq tempo the same throughout the trick's implement ation and just gettin a strange grain as you make your slider, not necessarily a slowdown of the entire sequence (althouh you could do that too via tempo sweep, but at that point i'd have to say it'd be best done on a pc or something)

this is more just another trick out of the bag, good for note variation and interesting fades. often i like to filter out somethin, but lately i've been looking for other ways to fadeout rather than just a straight volume drop. imagine a timestretch drop coupled with a volume or filter drop...the sound just gets that much more maleable as it fades.

it's also really good for more glitchy stuff. i've been going kind of crazy with sample manipulation lately, and this goes right in. between reverse layers, timestretch layers, waveform simult layers, there's a lot you can do.

Allow me to retort

Jahrome wrote:
Worth it? Realtime timestetch is available on practically all major digital audio recording software...just a few clicks of the mouse versus a time consuming method on the 1000???

Wow...I was talking sh!t here So sensitive......


Ok. So must people purchased MPCs because of the work flow. Others purchased it because Nym said you should purchase it I guess you guys are the ones that are ticked by my response.

The method Antonym suggested is a cumbersome process. I don't think it is...I know it is. If you really need realtime time stretching, an MPC is not the way to go. Software is a better alternative. If you don't like software then try out one of Roland's products.


And by the way...I will have a copy of the JJ OS. You know...trouble shooting bugs is what I do..so why not have their buggy OS that everyone pays $30 for?? Also, some of the added features that the JJ OS has always been possible with older MPCs.