How does the mpc work memory wise when extracting a sample?

Here's the deal. I was working with the sample from the beat battle and I had several chops, I had one small chop that I was trying to extract as a new sample by pressing SHIFT then F2 to extract it as a new sample. This brought up the screen that shows "new sample name" but when I pressed DO IT it gave me an "insufficient memory" message. I have 128mb installed, when I went to record mode it said I had something like 7 min in mono and 3 and a half or so in stereo to record, the sample list window showed that I had 39.7mb of free memory left. So I started experimenting for a minute. I was finally able to extract the 78kb chop once i deleted unused programs and freed-up 56.4mb, the actual sample is 56.1mb. If this is true then that would mean that I could have re-sampled the chop without having to free-up memory with no problem but extracting it wouldn't work. Does that make sense? Could it be that the mpc needs to have enough free memory to temporarily store the entire 56.1mb sample again while it extracts the chop even if the new sample ends up being only 78kb? Not sure if this is how the mpc works when extracting samples so I figured I'd bring it up so someone can drop some knowledge on the subject. Thanks guys.

yeah I have had similar problem. I do believe it might be a bug, but not sure. So might have to report it. Which JJOS are you using

I'm on os2xl v.3.16 and using a 2500. It was interesting to me that eventhough I had pleanty of free memory, it wasn't until I had just a little bit more free memory (56.4mb) than the actual size of the sample (56.1mb) that I was finally able to extract a very short sample. Maybe it's not a bug, maybe it's just a limitation in that the mpc does temporarily create another copy of the original sample while it extracts from it so even if you are extracting a small peice from a very large sample you still need to have enough free memory to cover both the new small extracted sample and the temporary copy of the original sample that is being used for the extraction. Maybe I'm crazy, maybe it really doesn't matter, but it's odd. I guess an easy way to check this theory out is to load a long sample, one that takes up more than half your free memory but still leaves you with a good amount of memory left, then try to extract a small peice from that sample and see what happens.

B.A. wrote: mpc does temporarily create another copy of the original sample while it extracts from it so even if you are extracting a small peice from a very large sample you still need to have enough free memory to cover both the new small extracted sample and the temporary copy of the original sample that is being used for the extraction.

I believe this is explanation plus we have only 115.5mb to work with wav's