Hi,

Would be great to have options for internal data format for CHOPs (similar to pixel format in TOPs)

At least ability to process 64bit floats or/and 32bit integers.

Obvious example:

AudiofileinCHOP with:

Play Mode - Specify Index

Index defined in samples starts give a precision errors at least after 6 minutes at 48000 sample rate.

Not obvious:

64 bit processing like summing, filtering and so on is a standart for modern DAWs.

It is obviously nessasary for non destructive processing samples recorded at 32bit depth and arguably nessassary for lower bit depths audio.

Cheers,

Alex

1 Like

It was my impression that all CHOPs are in fact doubles (64bit). Derivative can of course confirm. There is an option when looking at a CHOPâ€™s viewer and right clicking to â€śShow Exact Valuesâ€ť which will reveal the higher precision numbers stored. Integers I donâ€™t think are actually a thing in CHOPs, everything is a double float basically as I recall.

[edit] Oops, now I see your error is regarding the parameter for indexing, sorry. Could you use the seconds units format instead? You may be right then about this limitationâ€¦ not sure.

Here is â€śShow Exact Valuesâ€ť and it is actually 32bit float (with 8 didigts mantissa).

In case of 64bf - there will be 15 digits

Using frames for indexing - it is limited with frame length step (800 samples for 60 fps), so it is not possible to play exact part with per sample precision. Or you get similar precision error if using fractional frames.