Brian McFee

I'm very much in favor of this idea. I don't think we should have display automatically generate analyses, but otherwise this seems pretty straightforward.

> I just had an idea about calibrating the thickness of the waveform: how about we just leave it as it is natively? Natively, it's unbounded. Signals need not be...

> Does that mean that we won't be able to compare the absolute amplitudes of two different segments on the same plot? Not at all. It just means that if...

@lostanlen since #1063 is getting close, any chance of reviving this one? Or if you can share the prototype that made the plot above, I can polish it up.

I think I won't have time to put this in 0.8.0. Punting to 0.8.1 unless someone else wants to pick it up in the immediate future.

Thanks @nullmightybofo for reviving this one. > I am not sure whether there should/could be some kind of interaction (selecting range, scrolling?) for these plots, which presumably would require some...

Following up on this one, it was pretty easy to adapt our `waveshow` function to do something like this. Specifically, the code below does the following: - Interpolate the provided...

Did a bit of background reading, and I think the right way to go about this is to define `melodyplot` as a wrapper function which constructs a [`Transform` object](https://matplotlib.org/stable/api/transformations.html#matplotlib.transforms.Transform) from...

Got an almost working prototype over here using transforms. It's a much cleaner implementation, but the use of nans is causing a glitch in the envelope plot. ```python class TransformMelody(Transform):...

Follow-up on the above: the wrapping effect did not happen in the first implementation because the nans were apparent from the outset when the envelope plot was constructed. In the...