Henrik Enquist comments

Results 306 comments of


                                            Henrik Enquist

Control for pitch adjust

done

Param target_level sanity check - 4 * buffer size instead of 2 * buffer size

Thanks! I forgot to change the limit in the check, will do!

Param target_level sanity check - 4 * buffer size instead of 2 * buffer size

Fixed in https://github.com/HEnquist/camilladsp/commit/ca25bdaaadf6f9e95425101edec622c4206e05ad

Change FftError variants to have named members

This seems like a good idea, and I agree it should wait until the next major version bump. Let's keeps this issue open until then as a reminder.

Safe method is actually unsafe

Good catch! The bytes_per_frame get determined when initializing the AudioClient, so it could be stored then and later forwarded to the AudioCaptureClient. I think this should be safe. The only...

Iterator-based version of process_into_buffer()

I have been considering something like this and I think it's a good idea. I just didn't find the time to do anything yet. It would be really great if...

Iterator-based version of process_into_buffer()

`audio` is definitely interesting. That branch also has some very useful refactoring to reduce the duplication in the asynchronous resamplers. The new polynomial types would benefit from the same treatment....

RFC: Accepting more format changes

It would be really nice if the gadget number of channels and format were exposed in controls too, since that would allow us to stop as soon as a change...

RFC: Accepting more format changes

> IMO once the device is opened with some hw_params, the params cannot be changed within that session. So any change of any of the params would be preceded with...

Make `realfft` an optional dependency?

It would be no problem to make the FFT resamplers optional. The fast ones don't use the FFT and are pretty small, so omitting the FFT should reduce binary size...