At the moment I'm preparing a show in which I'd like to pixel map all of the bricks of a wall. So each brick would behave as 1 pixel: light up in its entirety. The approach I'm going for right now is taking a slice for each brick (right now I'm at slice No. 376 ) and resolume gives a warning of low memory, even if no content is playing and no clips are loaded into the program even. I consulted the facebook group and they said there wouldn't be any problem with such an amount of slices but clearly there is. Could anybody maybe suggest another way to pixelmap these bricks?
To clearify: after I mapped the bricks, I'll size down all of the input slices so that they become roughly 1 pixel and hence the output will change in its entirety too.
Slices definitely aren't 'free'. They do take up some memory, and will need to be drawn at some point. So by the time you get to number 376, things will have added up.
Performance wise, it makes a huge difference how many pixels you are actually drawing. 100 slices with a 10x10 input, outputting each at 50x50 doesn't make a dent. 100 slices with a 10x10 input, outputting at 1920x1080 each will slow things down considerably. So when preparing a complex map, make sure you keep both your input and output sizes somewhat similar to what you will need in the final result.
Memory wise, the Polygon slice is probably the most complex of the bunch. It's designed to keep track of an input shape, and then let you warp each point of that shape individually. Since you're going to be using 1x1 inputs anyway, you don't need access to the individual warping points. Basically you're storing quite a bit of data that you never will use. In such a case, it's better to use regular slices and then just mask the output or use the warper to get the shape you need.
Thanks for the reply. I was going for the slice squares and mask later on -approach already, only I'm still using polygon slices simply because regular slices are a huge pain in the ass to map! They always start out spanning the complete composition, both in input and output. Is there a way to draw these slices instead of having to reshape them from huge to small each time, or is this the only way?
By the way, awesome that you guys included spout as an output, now I can use a virtual screen to do this slicing and send it back through spout to a layer, where I apply a complex mask.
Is there a way to draw these slices instead of having to reshape them from huge to small each time, or is this the only way?
You can hit CTRL-D to duplicate a slice with the correct scaling. When you hit CTRL-D after a translation, it will duplicate the translation as well (like Illustrator).
Alright, still a ton of work but it already helps alot! Thanks!
One last question: The actual reason I'm doing this is because I'm trying to pixel-map bricks in a wall. So my input slices are preferrably 1x1 pixel and my output pixels are about 60x30 pixels. According to what you just said I assume that that'l be a butload of work for my computer, but intuitively this doesn't seem like a lot of work. It's just coloring a square with a certain color value. But of course if he calculates the value of each pixel individually (as he would do if he had to interpolate to enlarge or shrink an image) I can understand why it would.
And yeah but I think (hope) that the end result will be dope. But maybe as a suggestion for the next update: drawing rectangular slices with similarly shaped output slices, would be awesome.