op + or alphablend?

Roman Thilenius's icon

redoing some old stuff of mine.

i am using jit.op @op + in a serial array to mix still images against each other.

but the 8 bit resolution of the + operator becomes quite inaccurate very soon (say like... layer 8 is already performing + 12% instead of + 12.5%)

can i get op + to work in 32 bit somehow? or would you rather do that using the alphablend object?

(i do not need the alpha layer but have it included one everywhere, so you could just mix 2 of the "char 4" images using [alphablend] without the need to artificially create any alpha layer content)

Roman Thilenius's icon

no, sorry i am using xfade, not alphablend. xfade supports any type.

Matteo Marson's icon

Hi Roman,

I'd suggest to use jit.gen, because inside it the operations are floating point

Max Patch
Copy patch and select New From Clipboard in Max.

Roman Thilenius's icon

that way of performing addition with matrixes looks intriguing.

but can you explain why the first bang only gives me paprika?

Rob Ramirez's icon

An unfortunate side effect of adapt mode in matrix ops. Set adapt 0 to workaround this if needed:

Max Patch
Copy patch and select New From Clipboard in Max.