Re: v2.1.0 broke UV layers

On 03/03/14 19:59, Florian Kainz wrote:
> Are you sure there's anything in the file's RenderLayer.UV channels 
> besides zeroes?  The file reads fine in exrdisplay:  no errors 
> detected during decompression (meaning the pixel data decompress to 
> the expected size), and all the other channels contain what looks like 
> a reasonable image.  exrdisplay says that the channels in the UV layer 
> contain only zeros (not random data as you'd expect if the file didn't 
> decompress properly).
>
> As far as the IlmImf library is concerned, the UV layer is no 
> different from all the other layers; I'm not sure what kind of bug 
> would affect only the UV layer, but not any others?
>
> Have you tried saving the file out with a different compression 
> scheme, or with no compression at all?

Thanks for taking a look at it, Florian.  I could have sworn the 
channels were full of data, but in your analysis there's no room for 
mistake.  I tried redoing the render but I don't have any earlier 
openexr installed anymore.  And the file is from blender and I am unsure 
how to check the UV maps in that application.
Sorry for being a bother and reporting a false bug.  I apologize.