Inaccurate chopping, autochop useless?

This one is bothering me for a while now, i wonder why noone else ever said anything about it, maybe im doing sth wrong.

1. Do any of you use the auto feature when chopping? I tried a few times on the easiest sample possible. When i resample drum hits, for example snares i record 4 of them with different velocity. Sample looks like this:


Press chop, then auto, default settings (50/50, tried other settings too), press "do it" and the chops look like this:


Zoomed in:


Completely useless?

2. I f i chop manually theres also sth going on. I zoom all the way in and make sure the start of the chop is on point but when i convert to sliced samples it always chops about 2400 samples earlier then selected.


Noone ever bothered?

yupp i am confused about auto chop too. it has never been able to chop it the way i intend it to. i kinda gave up long time ago really.

i think it's really only to get the chops in the ballpark.
i use autochop and then make sure it chops every note into 1 segment. then i just flicker from segment to segment to make sure it's close enough to the transient.

i chop very quickly.

Thats what i do most of the time too, but useless feature -> more os space

I´m more worried bout the second point, i think it wasnt always that way...

Thats what i do most of the time too, but useless feature -> more os space Very Happy

oh wow
THAT is a good idea, i hadn't thought of that.

i imagine it's quite heavy, as there would appear to be a significant calculation involved...but then again that's only speculation - it could very well just be a stagnant equation that doesn't take much code...

Especially since it doesn't seem to work correctly!

well, i mean it works somewhat correctly. it at least throws a split in the general vicinity of a transient

but you're right - it's no recycle--it doesn't need to be. i've yet to find a sample that's so complex that i can't chop it in the mpc without using AUTO. in the end, it'd take longer exporting it to the pc and chopping in recycle and then going back and loading all the samples again and THEN making a program. no way man.

EZ, if you want it more accurate, then improve the algorithim. but do not get rid of the autoslice feature!!!

i agree autochop should be improved not removed. if it were to work better it could be immensely useful.

Yeah I wouldn't mind it if it worked. I haven't used it much but I remember it being slightly wonky and off the last time I tried.

The calculations for doing something like autochop aren't that difficult so it shouldn't take up much os space. If the algorithm was improved it definitely would work better.[/quote]

There is no way to "improve" the autochop function. The nature of what it is trying to do is what makes it not accurate... kind of like when you are trying to manually chop a sample that has no definate transients and you are not sure exactly where to chop it. Auto chop, by nature, is a function that is better performed by humans than by a machine. Sometimes you have to be ables to just try a few different spots in order to get it right... there is no way a machine can do it perfectly...

Now, that said, I'm sure there are (or soon will be) a few programs for the computer that can "intuitively" chop a sample for you, but this sort of software is far beyond the scope of the processing power of the MPC.

I say... Fuk an AutoChop... I would rather chop it myself...

TBonus76 wrote:There is no way to "improve" the autochop function. The nature of what it is trying to do is what makes it not accurate.....

yeah, but look at rinseout's example, that's ridiculous!! should work better than that, if that's it then get rid of it i say. i don't use it anyway, for chopping into small pieces, i find this to be helpful:

wow, is it really always that bad?

seems like a bug, really

-I only chop even intervals on the mpc... auto-stuff i do in recycle

so i never really noticed that its that bad.

i'm kinda remembering it working better than that when i did try it...?