Y8 Color Space and Avisynth 2.5/6

[bill_baroud]

Hello,

I recently implemented some DeBayer algorithm paper (http://www.dei.unipd.it/~mutley/Paper_pdf/Menon_Andriani_IEEE_T_IP_2007.pdf) and wanted to make an Avisynth Filter out of that.
Remembering some discussion about Y8 color space support, i was thinking it was now in avisynth, but it looks like it's still planned for Avisynth 2.6 which isn't ready yet. Is that correct ?
In that case, do you have any suggestion on how i could work around this limitation ? I just need two (contiguous) memory arrays as parameters to my debayer function.

Thank you for your help !

[IanB]

Assuming the data is per frame based. If it is clip static just store it in a var with env->SetVar/SetGlobalVar/GetVar.

And yes Y8 will be in 2.6. There may also be an untyped byte array format for frame associated meta data like the motion vectors from mvtools.

If you need a 2.5 compatible filter either use the luma channel of YV12 and ignore or fill the chroma planes with 128 or use RGB32 with height=1 and width=number of dwords you need.

[bill_baroud]

Debayering is just taking some 8-bits array image and convert that to RGB24, pretty straightforward but i need some way to input my 8-bits data. I think i'll go the YV12 way, since it's planar, i can just leave the UV channels like you said and i won't have much modification when Y8 will be around (i should have think of that myself, doh!)

And yes, i'm doing a 2.5 version, since i don't know when 2.6 will be available and i have time now :)

Thank you for your quick response !