clipka <ano### [at] anonymousorg> wrote:
> Edouard schrieb:
>
> >> (B) If (after fixing (A)) you see checkering in the preview window
> >> and/or the output file when rendering the scene file below, then your
> >> Display_Gamma (POV-Ray 3.6 and 3.7) and/or File_Gamma setting (POV-Ray
> >> 3.7 only), respectively, is wrong.
> >
> > Well that produces a broken file for me if I choose to output a PNG file (i.e.
> > the gray value isn't half the white value). HDR output works though (as I assume
> > it's in linear space).
>
> Careful!
>
> How did you /check/ that the gray value isn't "half the white value"?
>
> If you squint your eyes to "blur" the black-and-white-striped squares,
> and it then /appears/ to have the same brightness as the gray value,
> then that gray value /is/ half as bright as the white.

I can see it's displayed correctly. That's great. But you've taken the decision
that the file should be output with that display gamma baked in, and that's not
something that's desirable under all circumstances (like, for example, any
manipulation of the image afterwards).

> Note that Photoshop lies to you about brightness (and so do your eyes).
> Photoshop also uses physically wrong math when blurring images (at least
> the old version I own does).

No - Photoshop is doing a sensible thing, and that's to work in linear space
(exactly the same way that POV does). Photoshop is a production tool, and having
gamma applied half way through your workflow will just screw everything up.

If you assign a linear profile to the image (which you started to be able to do
in Photoshop 6), then Photoshop not only displayed the image correctly, but all
operations (like blurring it) also work exactly as you would want. I've attached
an image with a linear profile and one half blurred as an example.

Scaling the image also fails if there isn't a linear profile attached, and lots
of users here like to scale their final images down and don't really consider
that post-processing. Baked-in gamma makes that work incorrectly.

> The /physical/ blurring by squinting your eyes, however, does indeed
> comprise a physically correct "computation" of (black+white)/2 (taking a
> few steps back may do the same job, and some people may just need to
> take off their glasses), which obviously gives you a grey of /truly/
> half the white value (*), which you can then visually compare to the
> grey of /allegedly/ half the white value.

I'm arguing that for a digital artist who is using POV as part of their
workflow, having tools like Photoshop's blur give identical "output" to what you
see when you do the "squint" method, is a good thing.

I'm sure lots of people will be happy to bake the gamma into their file if they
are producing "final" images with POV, and not intending to post-process them in
any way (as long as they know scaling the image is then a no-no). But there are
plenty of others who use POV in a workflow that includes scaling and other post
processing, and we would be better suited with having linear ICC profiles
attached to the images. There is no "right" way, it all depends on how the user
is intending to use the image.

(Well, once ICC support is universal, then I would argue that a linear image
with the correct ICC profile is "right". We're not that far from that are we? On
my Mac, the OS, the browser, and image preview app, and the pro software like
Photoshop all get a linear ICC profiled file right every time...)

Cheers,
Edouard.

Post a reply to this message

Attachments:
Download 'gamma-test-linear.png' (136 KB)

Preview of image 'gamma-test-linear.png'