Justin Salamon

@pseeth I'm curious - is there an actual *need* to cast `ann.sandbox.scaper` to `jams.Sandbox`? It doesn't look like it would impact `generate_from_jams`? Also, I'm wondering why `ann.sandbox.scaper` is a jams...

Since the JAMS API doesn't guarantee any checks/casting for sandboxes outside of the specified locations as noted by @bmcfee, perhaps it would be safer to update the code to stick...

Also, we should specifically add tests that combine clipping + reverb, to make sure the two operations work well in for the order in which they are applied (in all...

Sounds perfectly reasonable, sure.

Applying reverb individually and summing will not give the same result as applying reverb to the entire scene, that's expected. We're planning to add support for proper room acoustics simulation...

Thanks for the suggestion - I think it's a great idea. To make this API proposal truly flexible and extensible I think it would be important to support custom transformations...

Cheers @turpaultn ! We could definitely add support for non-uniform discrete sampling, e.g. via a new `choose_weighted` distribution tuple. IIUC in the example above you're providing the probability for each...

@pseeth seems like something we must fix in the context of #55 ? Were you able to get a MWE to debug?

@epicycles much has changed since this issue was reported (we're now on 1.6.2 !!). Any chance you can test your code against this version and let us know if there...

I believe so, but of course we need to handle all possible distributions (e.g. including "choose"). Is this logic not already implemented in ensure_satisfiable_source_time?