atom feed24 messages in at.iem.pd-listRe: [PD] video querying
FromSent OnAttachments
Tina ShahFeb 9, 2006 11:08 pm 
chris clepperFeb 10, 2006 10:19 am 
Yves DegoyonFeb 10, 2006 10:45 am 
Tina ShahFeb 10, 2006 11:02 am 
Yves DegoyonFeb 10, 2006 11:04 am 
day 5Feb 10, 2006 12:00 pm 
cybo...@nocturnalnoize.comFeb 10, 2006 1:35 pm 
daniel kosenkoFeb 15, 2006 3:15 pm 
SylvainFeb 15, 2006 3:39 pm 
Yves DegoyonFeb 15, 2006 3:54 pm 
Renato FabbriFeb 15, 2006 4:54 pm 
Pall ThayerFeb 15, 2006 5:21 pm 
derek holzerFeb 15, 2006 5:37 pm 
daniel kosenkoFeb 15, 2006 8:37 pm 
Hans-Christoph SteinerFeb 15, 2006 9:56 pm 
Hans-Christoph SteinerFeb 15, 2006 10:01 pm 
Mathieu BouchardFeb 15, 2006 10:06 pm 
hard offFeb 16, 2006 3:29 am 
vincent riouxFeb 16, 2006 3:48 am 
Jonathan GoodwinFeb 16, 2006 3:57 am 
B. BogartFeb 16, 2006 7:51 am 
Frank BarknechtFeb 16, 2006 7:58 am 
Mark PolishookFeb 16, 2006 6:15 pm 
daniel kosenkoFeb 19, 2006 10:52 am 
Subject:Re: [PD] video querying
From:Yves Degoyon (ydeg@free.fr)
Date:Feb 10, 2006 10:45:55 am
List:at.iem.pd-list

chris clepper wrote:

The video portion can probably be done in GEM using a few objects like pix_mean_color, pix_diff, pix_histo, etc.

The sound part is trickier because there is no way to get sound from a movie file into the Pd audio DSP chain.

errr, you never tried pidip ? pdp_yqt, pdp_live~, pdp_theorin~, you need a good machine though...

ciao, sevy

Having the audio in a separate sound file and reading it at the same time as the video file could work but not have perfect sync. FFTs would make sense for the analysis.