|
[Dec. 13th, 2015|04:22 pm]
Jason
|
Internet spiral led to some papers on synthesized birdsong:
https://ccrma.stanford.edu/~tamara/publications/asa_02.pdf
http://www.csis.ul.ie/dafx01/proceedings/papers/kahrs.pdf
you can hear audio samples of the former, here:
https://www.cs.sfu.ca/~tamaras/
The zebra-finch to me kind of like air wheezing out of a balloon... which I think is still indicative of good progress, in a way, in that it does sound like a real-world acoustic process. |
|
|
| Comments: |
From:  eub
2015-12-16 09:17 am (UTC)
| (Link)
|
Hey, that's my officemate one summer!
The audio clips sound like they are the excitation source and they'll want some filtering... ah, described as "syrinx model", so I think it's like you're listening to your glottal pulse train before it goes through your series of tubes.
Hmmm the papers talk about a beak model also, so I think they may be somewhat filtered. Dunno, I think they sound fantastic just as interestingly-complex-synthesized-sounds. I'ma try to email her and see if she has an implementation she's willing to share. | |
