Converting using slices, what am I doing wrong?

I have an 82bpm bass loop that sounds nice https://soundcloud.com/giacecco/good-ba ... al_sharing . Then, I convert it into a new drum program after slicing, playback at 82 bpm and it sounds bad instead, like this https://soundcloud.com/giacecco/bad-bas ... al_sharing . I'm using more than 50 slices, by threshold using default values (50, 100), so the resolution of the resulting program is high and I did not expect artefacts. This is not the first time I go through this process, but the first time that it seems to create a bad result. Can you give me hints about what I may be doing wrong? Thanks!

G.

If your aim is to slice that bass loop into its individual notes (hence using threshold) then I cannot see why you have over 50 slices as I can only get 25 individual notes there. Personally I would not use threshold slices, I would just go for the manual chop approach. Here's mine

https://soundcloud.com/mpctutor/bounce- ... al_sharing

Here's the project file, I've included the chopped sample (25 non destructive slices), a chopped program and the MIDI file (82 BPM)

https://www.dropbox.com/scl/fi/iwqbpsbu ... 9c7w1&dl=0

Thanks and - by the way - I purchased your book already some time ago, thank you for your work.

I believe that the issue was somewhere else. I was starting from a sample that is 81 bpm and there must be something wrong in the process I was using to adapt it to my 82 bpm sequence. See my steps:

1) select a part of the original full sample to have an exact loop of 4 bars, Process > Discard
2) Process > Time Stretch from 81 to to 82 bpm
3) Process > Extract to another sample foobar.txt
4) Without switching sample to foobar.txt, chop the original using regions, with a reasonable number of regions (e.g. suitable to mute / unmute for live performance)
5) Convert > New program using slices

If I jump step (3), I get a program that does not sound good, as in my original post. But if I do (3), I get a program that sounds good... problem is, it doesn't make sense because I don't create the program starting from the extracted sample!

Perhaps you can confirm something: is the file always saved after a "Process" operation?

Your method sounds overly complex, I just chopped it as it was, I think I took a tiny bit off the end so it looped better and hit detect bpm just to get a tempo for the sequence. But I would say no need to time stretch, not when it’s just a 1 bpm difference and you are chopping down to individual notes

Time stretching is also going to degrade the quality of the loop