Brian McFee

This is quite a bit more complex than it is in, say, stft. The core VQT function has a recursive downsampling step for computing the response at each octave (wavelet...

Yeah, it could definitely be implemented in pseudo-cqt, since that's just an stft + basis projection anyway. I agree that some lookahead is fine here, but the issue is going...

Your traceback provides the answer to this question: some of the dependencies (notably numba) do not yet support python 3.12, so we can't either.

Looks like there's now a release candidate for numba, so they'll beat us to the next milestone. I've added this to the 0.10.2 docket.

> Note that Python 3.12 doesn't come with setuptools installed by default, so it may be good to add that to the requirements, otherwise imports of pkg_resources will give errors....

This is actually a bit subtle, but I believe it is behaving according to spec. The problem is that the trimming logic depends on being a certain amount (`top_db`) below...

I'd prefer to not introduce magic numbers, and leave the code as is. However, i agree that the documentation should be expanded to cover this point.

> I have checked the source code and there is a clear expectation that a signal containing only zeros should give an empty array in output. See [line 510](https://github.com/librosa/librosa/blob/8dae20c66353190749486abf6b332d93c4ab2b46/librosa/effects.py#L510) and...

:wave: @dana-gill just checking in to see how you're doing on this? It's the last PR we need to put in before pushing out the 0.10.2 release, so if there's...

Ok, no worries! I'll make some today to sort out what actually needs to be there from what's a nice-to-have, and take this one out of the critical path.