Has anyone done this with Pd? I thought I remembered something in GEM, but I can't find it anywhere. Just to have a cursor scroll left to right and assign the different values of pixels to different musical properties...
-Chuckk
Chuckk Hubbard wrote:
Has anyone done this with Pd? I thought I remembered something in GEM, but I can't find it anywhere. Just to have a cursor scroll left to right and assign the different values of pixels to different musical properties...
-Chuckk
Like this?
http://gridflow.ca/gallery/photo_pianoroll.pd.gif
Claude
On 1/23/07, Claude Heiland-Allen claudiusmaximus@goto10.org wrote:
Chuckk Hubbard wrote:
Has anyone done this with Pd? I thought I remembered something in GEM, but I can't find it anywhere. Just to have a cursor scroll left to right and assign the different values of pixels to different musical properties...
-Chuckk
Like this?
Yeah, I think so. How's it work? Thanks for sharing.
-Chuckk
Chuckk Hubbard wrote:
On 1/23/07, Claude Heiland-Allen claudiusmaximus@goto10.org wrote:
Chuckk Hubbard wrote:
Has anyone done this with Pd? I thought I remembered something in GEM, but I can't find it anywhere. Just to have a cursor scroll left to right and assign the different values of pixels to different musical properties...
-Chuckk
Like this?
Yeah, I think so. How's it work? Thanks for sharing.
-Chuckk
Not sure exactly, it's not my patch, but from the screenshot it looks like it copies a scanline of greyscale pixels into an array, then rescales the array exponentially into annother array (frequency<->pitch conversion), then does an inverse fft and sends the resulting audio to dac~; the partials of the output sound being determined by the particular scanline of the image that was just copied to the array.
I think this technique is a form of "scanned synthesis", but I'm not an expert on terminology. At least searching pd-list for that term shows up some interesting results...
Claude
That looks rad. I'll have to try it out.
Kevin
On 1/22/07, Claude Heiland-Allen claudiusmaximus@goto10.org wrote:
Chuckk Hubbard wrote:
On 1/23/07, Claude Heiland-Allen claudiusmaximus@goto10.org wrote:
Chuckk Hubbard wrote:
Has anyone done this with Pd? I thought I remembered something in GEM, but I can't find it anywhere. Just to have a cursor scroll left to right and assign the different values of pixels to different musical properties...
-Chuckk
Like this?
Yeah, I think so. How's it work? Thanks for sharing.
-Chuckk
Not sure exactly, it's not my patch, but from the screenshot it looks like it copies a scanline of greyscale pixels into an array, then rescales the array exponentially into annother array (frequency<->pitch conversion), then does an inverse fft and sends the resulting audio to dac~; the partials of the output sound being determined by the particular scanline of the image that was just copied to the array.
I think this technique is a form of "scanned synthesis", but I'm not an expert on terminology. At least searching pd-list for that term shows up some interesting results...
Claude
PD-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
Claude Heiland-Allen claudiusmaximus@goto10.org writes:
I think this technique is a form of "scanned synthesis", but I'm not an expert on terminology. At least searching pd-list for that term shows up some interesting results...
No, not at all. Scanned synthesis is explained here: http://www.csounds.com/scanned/
Anyways, I remember doing things like reading an image with [soudnfiler] into arrays as raw data. It gives interesting results, especially on uncompressed image formats. But there was no fancy playhead. It was pure data ;) just like the classic cat image.tif /dev/dsp
I recon this is not what Chuckk is looking for...
./MiS
You might want to take a look at [pix_data] if you want to use Gem.
alex
24 Jan 2007 23:04:09 -0500, Michal Seta mis@artengine.ca:
Claude Heiland-Allen claudiusmaximus@goto10.org writes:
I think this technique is a form of "scanned synthesis", but I'm not an expert on terminology. At least searching pd-list for that term shows up some interesting results...
No, not at all. Scanned synthesis is explained here: http://www.csounds.com/scanned/
Anyways, I remember doing things like reading an image with [soudnfiler] into arrays as raw data. It gives interesting results, especially on uncompressed image formats. But there was no fancy playhead. It was pure data ;) just like the classic cat image.tif /dev/dsp
I recon this is not what Chuckk is looking for...
./MiS
PD-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
Thanks for the tip. I've been looking it over, not quite sure yet. -Chuckk
On 1/25/07, Alexandre Quessy listes@sourcelibre.com wrote:
You might want to take a look at [pix_data] if you want to use Gem.
alex
May I suggest this paper: http://www.dafx.ca/proceedings/papers/p_309.pdf
and: http://ccrma.stanford.edu/~woony/works/raster/
The examples are quite fun. :)
Steve
This is awesome. I still don't have it working right, but I'm close. What I was doing was cropping the image to 1 pixel wide, but I think that was a mistake. I need a large block size for resolution, and through some testing, as Paris suggested, I realized I was getting discrete blocks of pitch that changed every few seconds. I suspect I just need more window overlap...
-Chuckk
On 1/26/07, Stephen Sinclair radarsat1@gmail.com wrote:
May I suggest this paper: http://www.dafx.ca/proceedings/papers/p_309.pdf
and: http://ccrma.stanford.edu/~woony/works/raster/
The examples are quite fun. :)
Steve
PD-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
When you get something workable, please post a patch. I've been following your conversation here with much interest.
Thanks, Kevin
On 1/27/07, Chuckk Hubbard badmuthahubbard@gmail.com wrote:
This is awesome. I still don't have it working right, but I'm close. What I was doing was cropping the image to 1 pixel wide, but I think that was a mistake. I need a large block size for resolution, and through some testing, as Paris suggested, I realized I was getting discrete blocks of pitch that changed every few seconds. I suspect I just need more window overlap...
-Chuckk
On 1/26/07, Stephen Sinclair radarsat1@gmail.com wrote:
May I suggest this paper: http://www.dafx.ca/proceedings/papers/p_309.pdf
and: http://ccrma.stanford.edu/~woony/works/raster/
The examples are quite fun. :)
Steve
PD-list@iem.at mailing list UNSUBSCRIBE and account-management ->
http://lists.puredata.info/listinfo/pd-list
-- "Far and away the best prize that life has to offer is the chance to work hard at work worth doing." -Theodore Roosevelt
PD-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
Well, here's where I have it now. I'm still working out how to scale everything correctly, so you'll have to adjust the size calculations and stuff according to the image. The metro object plays from left to right, or you can manually scroll with the y-input to the pix_crop object. I'm not sure of any of this. I wish I could get rid of the staircase effect; might just need more overlap. The scaling settings as it is are for the .jpg of the patch itself.
The scaling settings exist because I have a separate rectangle for the cursor, and its crop and translate take different values to have it show the part of the image it is over depending on the width of the image, and I don't understand exactly how those values work.
-Chuckk
On 1/27/07, Kevin McCoy km.takewithyou@gmail.com wrote:
When you get something workable, please post a patch. I've been following your conversation here with much interest.
Thanks, Kevin
So I'm using this now, and the problem I'm encountering is that I'd like to be able to control the frequency range of a picture without editing the picture itself. I'd like to find a way to have pix_pix2sig~, or some other object, read the image and render it to audio, but padding some number of samples with zeros, so high points in the picture aren't necessarily out of my hearing range.
Would it work, perhaps, to give the cursor user-controllable height, up or down, and color it before adding the image? But if I'm not mistaken, rectangles cannot be textured to in any other way than right at the corners... Any suggestions?
Thanks. -Chuckk
On 1/27/07, Chuckk Hubbard badmuthahubbard@gmail.com wrote:
Well, here's where I have it now. I'm still working out how to scale everything correctly, so you'll have to adjust the size calculations and stuff according to the image. The metro object plays from left to right, or you can manually scroll with the y-input to the pix_crop object. I'm not sure of any of this. I wish I could get rid of the staircase effect; might just need more overlap. The scaling settings as it is are for the .jpg of the patch itself.
The scaling settings exist because I have a separate rectangle for the cursor, and its crop and translate take different values to have it show the part of the image it is over depending on the width of the image, and I don't understand exactly how those values work.
-Chuckk
On 1/27/07, Kevin McCoy km.takewithyou@gmail.com wrote:
When you get something workable, please post a patch. I've been following your conversation here with much interest.
Thanks, Kevin
-
Alexandre Quessy -
Chuckk Hubbard -
Claude Heiland-Allen -
Kevin McCoy -
Michal Seta -
Stephen Sinclair