This forum is now locked and moved (back) to Plogue's forum
programmatic region variations vs actual sample variations (efficiency)
  • Hi everyone and thanks for having me here! I have a seemingly stupid question, but I guess it's more stupid not to ask :-)

    I'm making an SFZ lib, and I have an option whether to use several recorded samples for the same note or just one sample for each note. I would then use velocity layers and round-robins. For the fewer sample case, I would have to do some tricks like different cutoffs, slight detuning, offsets, etc...
    Aside from the size of more samples combined, is it not naive for me to assume that the fewer sample spread over many regions would require less memory? (i.e. effects are done on the fly, probably taking a little more CPU). I'm also leaning towards this because I have little patience in preparing more samples (I'd rather leave it to the pros :-) ). Maybe the answer is obvious, but I would still be glad to get a clear yes/no.

    I'm aiming for a less powerful notebook that I can conveniently place near my digital piano or carry around with a small midi keyboard (since the ARIA website also mentions iOS, maybe an iPad too someday...?). I had been experimenting with sforzando for quite a while now, and also have GPO4.


Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!