That link doesn't seem to work. Do you know ahead of time which regions each image is valid for?
If you know the region for each image, then you can use crop() to trim down to just that region, and then add() will work. The problem is that there isn't a reliable way to automatically figure it out from the image itself (if there is no window designation in the metadata), because there's no way to distinguish in-region black pixels from ones outside the intended region. On Jun 20, 2014, at 1:18 AM, Fredrik Averpil <[email protected]> wrote: > Hi again Larry, > > Sorry for the confusion... I see now that the files I provided in the dropbox > link did correctly composit and were not exhibiting the issues I'm having... > so I just now updated the files inside the folder with an AOV that do exhibit > the issue, called "renderid" along with a "diffuse" AOV that does not exhibit > the issue: https://www.dropbox.com/home/Public/email/oiio_ImageBufAlgo > If compositing all "diffuse" image slices together using ImageBufAlgo.add(), > the result (composited_diffuse.exr) looks exactly like you would expect. No > issues here. > If compositing all "renderid" image slices together using ImageBufAlgo.add(), > the result (composited_renderid.exr) looks wrong as the white background (or > "canvas" if you will) gets added along with the slice's rgb values. > > The slices come out of the V-Ray renderer so I cannot change the way they get > written. But after having received your input, it sounds like they are not > defining the data window correctly – or they should keep the surrounding > color around the slice black (like they do in the "diffuse" case), and not > white (like in the "renderid" case). > > My initial question was about whether it would be possible, using OIIO, to > somehow combine these kinds of "renderid" slices where the surrounding > background is not black and still get expected results. But I'm guessing > that's not possible as they haven't defined the data window to equal the > slice dimensions. > > I'll contact the V-Ray devs and see if they can address this. > Thanks for your help, Larry – and I'm sorry about the confusion. > > > Regards, > Fredrik > > > > > _______________________________________________ > Oiio-dev mailing list > [email protected] > http://lists.openimageio.org/listinfo.cgi/oiio-dev-openimageio.org -- Larry Gritz [email protected]
_______________________________________________ Oiio-dev mailing list [email protected] http://lists.openimageio.org/listinfo.cgi/oiio-dev-openimageio.org
