Hey guys -- I have a lot of large black and white masks for a scene. If I combine three at a time, as RGB channels, then split them via functions once in TG, would that reduce the memory consumption or is it the same cost?
Anyone?
I don't know, why don't you try it yourself? :)
I would think it should be the same memory cost. TG2 needs to store everything in memory at full bit depth, no compression, and all channels. So whether it's 1 image with 3 channels or 3 with 1 each, it should be similar. There should only be a minimal difference due to file container overhead.
- Oshyan
Excellent news. Thanks, Oshyan.
Hang on, if I currently have 3 images, that's 9 channels. If I combine them it reduces to 3 channels, so the overall cost is less. Right?
Currently reading http://forums.planetside.co.uk/index.php?topic=2519.0 as I work on this. I've combined my masks for a test and will post updates.
I was assuming you were using 1 channel images. But come to think of it, I'm not certain TG2 supports them, at least not for all formats. Maybe TIF, SGI, and/or EXR.
- Oshyan
How do you reduce an image to a single channel? When I split to RGB then save one, on opening it seems to have 3.
Meanwhile, can anyone help me figure out how to seperate a RGB image into the separate channels. I've tried constant Colour Red ---> Subtract Colour ---> Blend by Shader but get weird results. Is there a function to do this better?
I think I've found it: Red to scalar, Green to scalar, Blue to scalar seems to work so far. What does scalar mean?
Here's some info about scalars:
http://en.wikipedia.org/wiki/Scalar_(mathematics) (http://en.wikipedia.org/wiki/Scalar_(mathematics))
http://en.wikipedia.org/wiki/Scalar_multiplication (http://en.wikipedia.org/wiki/Scalar_multiplication)
Thanks. I've converted some of my masks and updated the tgd. Running my first test render now. If it doesn't crash I'll consider it a success. Meanwhile I'll read up on the links you posted.
Damn. Back to the drawing board.