Not so much a specific suggestion or request as pointing out that an original Kinect is easy to find in the $10-$20 USD range (plus another $5-$10 for a 360 to PC adapter) RPi support already exists, and the possibilities of depth data as a control source are pretty vast.
Getting a ton of use out of auto_waaave so far, going to start collecting the stuff I need to put together a spectral_mesh over the next few weeks.
Great suggestion
one of my future projects that I’ve got a decent start on would benefit from a kinect but can work without it
looking forward to seeing what others do!
It looks like people have done a lot with the Kinect/RPi combo in hobbyist robotics, but not much else.
One more thing that makes me wish I had any coding experience at all.
I guess it would probably be possible to use it with Processing, but it would be really cool to integrate it into the waaave_pool ecosystem somehow.
the only obstacles i can think off off hand would be frame rate issues due to the size and rate of information coming in and placing the depth data into a texture for processing along with the video data. ive been like low key interested in experimenting with those but in the long run its seems way more computationally efficient to just use brightness as depth cue like human perceptual systems do!
Yeah, I’d be way more interested in using the depth data to control other parameters than using it for real time 3d scanning, and also using the raw IR camera feed to have a cheap IR video camera. I guess depth-based keying would be pretty useful, too.
word yeah, either way for controlling parameters it would still need to get encoded as a texture for it to work as video rate control source (as opposed to the midi controller which is all frame rate control source). might look more into it once i start exploring more into the second gen waaave pool stuffs after the big video waaaves update is finished
Makes sense, I was thinking there might be a lower level way to use the depth data stream but I don’t really know.
It’d be fun to just use it as a camera and be able to choose between the main camera, IR camera and depth map (IIRC the depth map is leaves the Kinect itself as a separate grayscale video stream) but not so big a deal it would be worth prioritizing it above anything else you’re working on.
yeah theres probably some low level way to extract info from kinects, is more of a question of what form it is getting decoded/encoded. to do anything seriously fun with wp that you cant just do with any other kinds of control sources tho itd need to get encoded as a texture somehow so it can be used on the parallel processing shader side of operations and im about 50 percent sure that that process itself would significantly impact the frame rate
Somehow I didn’t notice until just now that this got posted in current projects. Ooops!
you can edit the category after posting i believe, this one could fit in several but i don’t think it’s that off topic in here.
Didn’t notice that, I’m used to forums with minimal editing. I think it fits fine here now.