adding yin segmentation to CaMu

[diemoschwarz]

On 23/04/2023 10:54, Christopher Trapani wrote:

Hi guys — I am trying to build a camu.process.yin patch which will apply the yin
markers processing that we developed for our Seg-Lab. Far better for grains with
good frequency estimation.

I'm in need of a little help / collaboration, though. The mubu.process modules
define the markers well, but could one of you give me a hand or a hint about the
next step — namely, writing a mubu.process that will reanalyze the audio
according to the markers generate by yin (or resegment the raw yin data) for a
table of familiar descr values ?

I've attached a prototype here, integrated into tutorial 5. Ultimately we could
add this (and maybe other segmentation methods!) to the package...

Thanks — C

[einbond]

Yes agreed! Could this relate to the more general issue of applying external markers to mubu.process? #82

[diemoschwarz]

Hi guys, segmentation has advanced a lot in Mubu! There is an improved example of yin segmentation in Matt's 4th mubu tutorial.
I'll add a quick mubu.process.yin patch, but, indeed, we should move to process2 which allows external and hand-edited segmentation.

[christophertrapani]

So, just to understand: that would be a feature request, to create a new process2 object? And it's not possible with (some combination of) current MuBu objects?

Meanwhile, I am watching the tutorial to see if there are lessons we can apply to CaMu...

[einbond]

Sorry my mind was rusty on our previous process2 discussions and I now see this is already under issue #48 . I'll close this and we can continue there!

[diemoschwarz]

yin seg. discussion continues on issue #86