windowing extended objetcs, mistakes on help files and how to make them into vanilla patches
Hi there, I was studying the window functions for FFT, so I was trying to build into vanilla pacthes the window functions that come on Pd-Extended [windowing] package. I got to do them all, but kaiser~ and gaussian~
Do any of you feel capable to perform this ver nerdy taks?
Well, anyway, here are some inconsistencies/mistakes I found ont the description of the formulas that are present in the help of those external objetcts:
blackman~ instead of 0.8 it should be 0.08
welch formula: 1-abs(x) - not sure, but doesn't seem to be right or
clear. I used this one: http://paulbourke.net/miscellaneous/windows/
- barlett~: 1-x^2 doesn't seem to be right or clear either, I used this
one: http://mathworld.wolfram.com/ApodizationFunction.html
- connes~: cos(pi*x/2) - it's the same as cosine~ so it can't be right, I
used the above link again.
Cheers Alex
seem to have nailed the gaussian window, but with different width values than the object.
2012/10/16 Alexandre Torres Porres porres@gmail.com
Hi there, I was studying the window functions for FFT, so I was trying to build into vanilla pacthes the window functions that come on Pd-Extended [windowing] package. I got to do them all, but kaiser~ and gaussian~
Do any of you feel capable to perform this ver nerdy taks?
Well, anyway, here are some inconsistencies/mistakes I found ont the description of the formulas that are present in the help of those external objetcts:
blackman~ instead of 0.8 it should be 0.08
welch formula: 1-abs(x) - not sure, but doesn't seem to be right or
clear. I used this one: http://paulbourke.net/miscellaneous/windows/
- barlett~: 1-x^2 doesn't seem to be right or clear either, I used this
one: http://mathworld.wolfram.com/ApodizationFunction.html
- connes~: cos(pi*x/2) - it's the same as cosine~ so it can't be right, I
used the above link again.
Cheers Alex
I don't quite follow, but if you think there are bugs in the windowing objects, you should file a bug report.
.hc
On 10/16/2012 05:03 PM, Alexandre Torres Porres wrote:
seem to have nailed the gaussian window, but with different width values than the object.
2012/10/16 Alexandre Torres Porres porres@gmail.com
Hi there, I was studying the window functions for FFT, so I was trying to build into vanilla pacthes the window functions that come on Pd-Extended [windowing] package. I got to do them all, but kaiser~ and gaussian~
Do any of you feel capable to perform this ver nerdy taks?
Well, anyway, here are some inconsistencies/mistakes I found ont the description of the formulas that are present in the help of those external objetcts:
blackman~ instead of 0.8 it should be 0.08
welch formula: 1-abs(x) - not sure, but doesn't seem to be right or
clear. I used this one: http://paulbourke.net/miscellaneous/windows/
- barlett~: 1-x^2 doesn't seem to be right or clear either, I used this
one: http://mathworld.wolfram.com/ApodizationFunction.html
- connes~: cos(pi*x/2) - it's the same as cosine~ so it can't be right, I
used the above link again.
Cheers Alex
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
I seem to remember sitting in Hans' apartment making the help files for the windowing library. There were none, and I found the objects useful as drop-in FFT windows (which of course is the point) and for other PSOLA stuff... I had flu in New York. Perhaps that made me wikipedia-lazy! I suggest that the bugs are in the descriptions rather than the objects, or alternatively the formulae have been modified to make them more "pd standard" in the objects. I'm sure nobody will mind if someone makes this obscure but useful library more mathematically correct. Ed Gemnotes-0.2: Live music notation for Pure Data, now with dynamics! http://sharktracks.co.uk/
From: Hans-Christoph Steiner hans@at.or.at To: pd-list@iem.at Sent: Wednesday, 24 October 2012, 2:48 Subject: Re: [PD] windowing extended objetcs, mistakes on help files and how to make them into vanilla patches
I don't quite follow, but if you think there are bugs in the windowing objects, you should file a bug report.
.hc
On 10/16/2012 05:03 PM, Alexandre Torres Porres wrote:
seem to have nailed the gaussian window, but with different width values than the object.
2012/10/16 Alexandre Torres Porres porres@gmail.com
Hi there, I was studying the window functions for FFT, so I was trying to build into vanilla pacthes the window functions that come on Pd-Extended [windowing] package. I got to do them all, but kaiser~ and gaussian~
Do any of you feel capable to perform this ver nerdy taks?
Well, anyway, here are some inconsistencies/mistakes I found ont the description of the formulas that are present in the help of those external objetcts:
blackman~ instead of 0.8 it should be 0.08
welch formula: 1-abs(x) - not sure, but doesn't seem to be right or
clear. I used this one: http://paulbourke.net/miscellaneous/windows/
- barlett~: 1-x^2 doesn't seem to be right or clear either, I used this
one: http://mathworld.wolfram.com/ApodizationFunction.html
- connes~: cos(pi*x/2) - it's the same as cosine~ so it can't be right, I
used the above link again.
Cheers Alex
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
-
Alexandre Torres Porres -
Ed Kelly -
Hans-Christoph Steiner