Dear List,
I would like to realize an installation in GEM but I don't know if there are appropriate objects or If I have to teach myself how to program them. If you could give me advise which direction to go I would be very pleased.
Two cameras would send their frames to two image stacks which contain say 30 frames (the more the better). The actual frame is subtracted from the last one and the resulting frame of this simple kind of movement-dedection would serve as a parameter to compose from the two stacks the actual video output reading each single pixel in a random manner from the two stacks mixing both by using an appropriate function.
Does anyone have an idea how to realize that with existing objects, and if not, where do I find good documentation to learn how to program GEM.
Thanks a lot,
Florian
- there is a tutorial and help files comming with gem, they are
exellent. http://gem.iem.at 2. you may find some more in the net, like: http://taproot.dyndns.org/~cgc/tutorial_gem/tutorial8.html http://www.ekran.org/pd/ 3. i've made one for bloody beginners in german: http://kunstundmedien.burg-halle.de/km/PD-workshop.zip 4. for yur stuff you might want to have alook in the helpfiles of pix_background and pix_movement there is also doc/11.Gem/04.Video/03.movement_detection.pd
Note that Gem is very strong when it comes to putting video on openGL objects and moving them in space and such, the strength is in vector operations, scaling rotating stretching lighting etc..
Pixel processing, which sounds like what your aiming for, is probably better suited to PDP/PiDiP and/or gridflow.
That being said there are some pix_ objects to buffer pixes, but I've not really used them personally.
B.
Florian Grond wrote:
Dear List,
I would like to realize an installation in GEM but I don't know if there are appropriate objects or If I have to teach myself how to program them. If you could give me advise which direction to go I would be very pleased.
Two cameras would send their frames to two image stacks which contain say 30 frames (the more the better). The actual frame is subtracted from the last one and the resulting frame of this simple kind of movement-dedection would serve as a parameter to compose from the two stacks the actual video output reading each single pixel in a random manner from the two stacks mixing both by using an appropriate function.
Does anyone have an idea how to realize that with existing objects, and if not, where do I find good documentation to learn how to program GEM.
Thanks a lot,
Florian
PD-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
Hello
Two cameras would send their frames to two image stacks which contain say 30 frames (the more the better). The actual frame is subtracted from the last one and the resulting frame of this simple kind of movement-dedection would serve as a parameter to compose from the two stacks the actual video output reading each single pixel in a random manner from the two stacks mixing both by using an appropriate function.
should be possible with stock GEM modules, I did an simular installation with pix buffering and subtraction etc, but a single usb cam. My only concern is connectig 2 cameras, I believe that this gives trouble with any modules but I would like to be proven wrong.
Cheers,
Malte
Malte Steiner wrote:
Hello
Two cameras would send their frames to two image stacks which contain say 30 frames (the more the better). The actual frame is subtracted from the last one and the resulting frame of this simple kind of movement-dedection would serve as a
should be possible with stock GEM modules, I did an simular installation
yes
with pix buffering and subtraction etc, but a single usb cam.
if you just want to get some simple movement detection there is [pix_movement] (and [pix_movement2] which does exactly that: subtract to (or more) subsequent (if this is the correct word) frames; so no need to get into the strangeness of [pix_buffer]...
My only concern is connectig 2 cameras, I believe that this gives trouble with any modules but I would like to be proven wrong.
i am not sure about osX (Gem either tries to grab the first available device, which would be ok; or the first listed device (which would reduce your options to one camera) on linux i have worked with several cameras within one Gem-patch, there is something like a "device" message.
mfg.as.er IOhannes
On Apr 25, 2005, at 6:51 AM, IOhannes m zmoelnig wrote:
i am not sure about osX (Gem either tries to grab the first available device, which would be ok; or the first listed device (which would reduce your options to one camera)
This is indeed the case on OSX. Only the first device is used by pix_video. I could not get additional devices working when I had two to test, and now I only have one. I figure 1 input is better than none to start with.
There is a decent chance I might have a whole lot of time to work on the Quicktime code of GEM in the near future. This is on the list to do.
-
B. Bogart -
chris clepper -
Florian Grond -
IOhannes m zmoelnig -
Malte Steiner -
Max Neupert