Am 26.05.2005 um 22:33 schrieb Pierre-Olivier Latour:
It's already the case: for each iteration, the CI kernel program
executes for every pixel, but the CI Kernel patch that wraps it
evaluates its inputs only once - the inputs are like constants for
the kernel program.
hmm, i suspect I haven't made myself really clear.
In my example "MulleSubdivideImage" (Walk down red "Image
Subdivider" (3) -> "Iterator"), there are (differently arranged and
simplified to conserve bandwidth) two Core Image Kernels:
if I hook up the "xOff, yOff and factor" to outputs of "iterator
Variables" (or its descendants) the image kernel runs faster than the
"Image Texturing Matrix". If I use some passed in values (from
parent) as input variables, performance drops down to the level of
crap -er- crop :)
I would like to understand why that is so and how I can remedy that.
Ciao
Nat!
------------------------------------------------------
I like the word Irksome, it sounds like it should mean
more than it actually does. -- Chris Redgrave
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Quartzcomposer-dev mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/quartzcomposer-dev/email@hidden
This email sent to email@hidden
