Set pixel RGB value in a grid/matrix?



  • Hello,

    I wonder if it's possible to generate/create a matrix or grid where the RGB value of each pixel can be set sequentially [timebased], starting at pixel 1 > pixel... 
    I don't see the solution, but it would be great if someone would.
    Thanks in advance, Klaas

  • Tech Staff

    That's pretty complex; Data Array actor is a good place to start along with the Measure Colour actor.

    I can't help much more that this at the minute.


  • What resolution do you want?



  • I know it's damn complex stuff I got myself into, 

    I have the impression that the data-array actor is limited to 999 values, which is not enough for a 300x1080pix grid [maximum size].
    Final output resolution should preferably be 1920x1080 and 3 grids generating values at the same time [without burning up the machine].
    Color would be set by audio input of the low/mid/high values, scaled to 0-255 RGB
    You can see a mock-up of what I'm trying to achieve [here](http://klaasverpoest.com/crippled-symmetry/)


  • The short answer is no. I would not bother trying this with isadora, the only hope you may have is through a shader inside an fflg plugin. CPU based pixel manipulation is pretty slow at any rate. Doing it in a single thread (Isadora does this) will be very slow. Isadora is not setup for pixel access anyway.

    It seems you want to algorithmically generate images, at the moment your approach is to create an equation to manipulate every pixel manually based on input. I would suggest either re-imagining this to be a little more abstract and using some other effects to achieve something else that represents this relationship or switching tools to do this properly. You could still run Isadora and write yourself some software that can take input data and generate your image and send it over syphon. There are quite a few options for making pixel based algorithmic video (Isadora is not really one of them).
    On your proposal for using FFT (Audio frequency analysis) think a little more about your input, you could run an fft with 1000 bins, but this is not so fast anywhere and by the time you have 1000 bins the relationship of the sound to video is pretty hard to see. With only 3 bins of FFT (you say low mid and high) you have a lot of repetition and pixel access is not really needed, maybe I misunderstand you though. I dont see how using only low mid and high audio will give you a complex and detailed image such as the one you linked to.


  • Couldn't you use the OFXFFGL addon for this Fred ;-)



  • Unfortunately I can get it working with the newer OF but not with Isadora. I will have a go at it when I get some time, (so far the plugins I make with it load in resolume but not Izzy).


  • Tech Staff

    @Fred,

    Would you be willing to share one/two of the OF based FFGL files with the dev team?
    I think it would be great to workout why they are running in some environments and not others


  • I had a chat to Mark about it already. It seems that he adheres more strictly to the standard than other hosts. I dont have a way around this nor more time to dedicate to it. We did chat about making an OF to Izzyplugin bridge- something that would be very useful, but version 2 needs to be finalized first.


  • Tech Staff

    That would be very cool.


Log in to reply
 

Looks like your connection to TroikaTronix Community Forum was lost, please wait while we try to reconnect.