as far as i understood the help patches i should use pix_imageInPlace instead of pix_image because of better performance. i do not get the usage of pix_imageInPlace it does not really work for me, where is the problem?
OK, it looks like [pix_multiimage] works a bit more like [pix_image], only without the need to load them from disk each time, as they are all loaded into RAM.
[pix_imageInPlace], on the other hand, loads the images to texture RAM (and as such is not used with the [pix_texture] object, as this is already done). As such, these images cannot be processed with subsequent [pix_*] objects
i was comparing the performance of LZW compressed tiffs and uncompressed ones, and had the impression that without compression it runs a bit better. do you agree?
At a guess, when you load the compressed images to memory (of whichever sort), the computer decompresses them (so that the image in memory is uncompressed). If you are using [pix_image], it would need to do this often, whereas with the other objects it would only happen once when you load them. So, when you get things working with objects other than [pix_image], by this logic, it would no longer make any difference wether the images were compressed. I think.
I've not actually directly used these objects significantly, though, so this is a bit speculative.
Cheers,
Peter