[cairo] Image Format / Pre-Multiplied Alpha Related Feature Request

[William Kappler]

I was thinking that when you suggested it, but I don't know if that's a big
problem or not. I suppose it depends on how exactly the information is used
- although I really wanted to avoid concerns like that.

Oversampling would definitely help with gradients/boundaries, but I am a
bit worried that abutting regions of solid color might unexpectedly fail to
match up. And if someone for some reason tried to use Cairo output as
non-image data in a shader... I'd rather not think about that situation in
the first place!

I'm just putting this issue on my backburner for now. If Cairo doesn't have
some sort of internal solution by then, I'll try to come up with an
external workaround. Thanks for the suggestions.


:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-
William E. Kappler II
LinkedIn <https://www.linkedin.com/in/williamkappler> · Blog
<http://williamkappler.blogspot.com/> · Project Website
<https://github.com/WilliamKappler/onathacar/wiki>

On Sun, Jan 31, 2016 at 1:03 AM, Lawrence D'Oliveiro <
ldo at geek-central.gen.nz> wrote:

> On Fri, 29 Jan 2016 12:12:29 +1300, I wrote:
>
> > Try oversampling.
>
> I realized, this may work to some extent for gradients and other such
> areas of varying pixel values, but it won’t properly handle areas of
> constant colour. Averaging 4 or 16 pixels of the same truncated value is
> not going to recover the original, more accurate value.
>
> Still, it might be of use with photorealistic-type imagery...
> --
> cairo mailing list
> cairo at cairographics.org
> http://lists.cairographics.org/mailman/listinfo/cairo
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.cairographics.org/archives/cairo/attachments/20160201/26a9be1c/attachment.html>