plem

In a system like this (see patch), where there can be several resolutions coming in, how do you tune the system for an optimum output resolution?

For instance: the beamer is 1280 x 720. The best pratice would be, in my eyes, to set, somewhere central, a desired resolution of 1280 x 720 and let everything to be transformed to that resolution. But where to set that while running and how to let all the participants in the process obey that settings?

This would maybe also help the next problem: the bigger patch, from which this is a simplification, behaves somewhat unpredictable at times when different resolutions are brought in (flickering, a kind of whitish flare, hd looks like low resolution and things like that). And my idea is that that might end when all the resolutions would be the same.

The cost would be that maybe the graphic processor would have to work harder, because there will not be any low resolution parts left.

But the principle question of cause, and a little bit a newly one, is: what determines the resolution of a multi resolution openGL world? And how to influence that?

MixResolTest.maxpat

In a system like this (see patch), where there can be several resolutions coming in, how do you tune the system for an optimum output resolution?
For instance: the beamer is 1280 x 720. The best pratice would be, in my eyes, to set, somewhere central, a desired resolution of 1280 x 720 and let everything to be transformed to that resolution. But where to set that while running and how to let all the participants in the process obey that settings?
This would maybe also help the next problem: the bigger patch, from which this is a simplification, behaves somewhat unpredictable at times when different resolutions are brought in (flickering, a kind of whitish flare, hd looks like low resolution and things like that). And my idea is that that might end when all the resolutions would be the same.
The cost would be that maybe the graphic processor would have to work harder, because there will not be any low resolution parts left.

Thanks to any input.
Peter
plem@xs4all.nl

So the question is:
In a system like this (see patch), where there can be several resolutions coming in, how do you tune the system for an o

resolution-setting-in-opengl

if your jit.movie(~) objects have output_texture enabled (which they 

 should in your patch), then the output dimension adapts to the dimensions of movie file loaded from disk.

jit.gl.slab/pix/texture will adapt to the dimensions of their left-most input, unless @adapt is disabled.

jit.gl.node/camera capture will use the dimensions of their parent window context for texture output unless @adapt is disabled.

it is usually best practice to match your file dimensions to that of your output window, but not necessary. If you'd like to process part of your signal at a lower resolution simply set adapt 0 on the jit.gl.slab/pix/texture at the start of the chain where you'd like to downsample, and explicitly set the dimension with the dim attribute.

if your jit.movie(~) objects have output_texture enabled (which they absolutely should in your patch), then the output dimension adapts to the dimensions of movie file loaded from disk. 

jit.gl.slab/pix/texture will adapt to the dimensions of their left-most input, unless @adapt is disabled. 

Thanks Rob. Clear answer and clear logic that the lefmost input determines the reaolution with adapt at 0. Would that also mean that, as it is in my patch where two jit.pix objects are next to each other on the same level, the leftmost input of the left one determines all resolution?

i'm not quite sure what you're asking. if input A and B are going into pix 1, and input C and D are going into pix 2, then pix 1 calculates using dim A and pix 2 calculates using dim C.

if those both feed pix 3, then pix 3 will calculate using whatever is feeding it's left input.

resolution setting in openGL