Alain <aze### [at] qwertyorg> wrote:
> The top strip show almost no difference fot the first 5 shades, and to 
> much contrast between the last 5.

  Your monitor is probably calibrated to be really, really dark.

  Sometimes people post really dark images in p.b.images and I have hard
time distinguishing anything in those images, even though others say that
they can see it just fine. That's probably because my monitor is calibrated
to be a bit too dark. I assume with your monitor you would only see black
and nothing else.

-- 
                                                          - Warp

Post a reply to this message

Warp schrieb:

> http://warp.povusers.org/pov_imagemap_test/index.html

>   Granted, the perceived brightness of this checkerboard pattern does not
> match the 0.5 grey of the POV-Ray 3.6 image. However, neither does it match
> the one rendered with POV-Ray 3.7. In fact, the latter deviates *more* from
> that perceived brightness than the former. It's way too bright. At least on
> my screen.

The 3.7 shot looks /perfect/ on my display, no matter whether I display 
it in Internet Explorer, Windows Explorer preview, or Photoshop 6.0. 
Given that I did invest some time into calibrating my primary display, 
this tells me that yours/ must be /way/ off.

Now, that isn't generally a problem: You can work with a display system 
gamma of 1.0 if you like. But in that case make sure your image viewing 
software supports non-standard display gammas, and is actually 
configured to do so.

So stop your griping, disengage "demand mode", calibrate your image 
viewing pipeline properly, and /then/ come back with any residual problems.


As for entering gamma-pre-corrected colors into POV-Ray, you might use 
something like this:

   #macro UnGamma(C)
     #local G = 1/2.2;
     <pow(C.red,G),pow(C.green,G),pow(C.blue,G),C.filter,C.transmit>
   #end

   #declare MyGrey = UnGamma(rgb <127,127,127>/255);


Now go and do your homework please.

Post a reply to this message

Warp schrieb:
>   Sometimes people post really dark images in p.b.images and I have hard
> time distinguishing anything in those images, even though others say that
> they can see it just fine. That's probably because my monitor is calibrated
> to be a bit too dark.

Your image viewing pipeline must be really, really badly calibrated; 
from the shots you put on the 'net, I would instead expect you to 
experience most images as way too bright.

Post a reply to this message

clipka <ano### [at] anonymousorg> wrote:
> Warp schrieb:
> >   Sometimes people post really dark images in p.b.images and I have hard
> > time distinguishing anything in those images, even though others say that
> > they can see it just fine. That's probably because my monitor is calibrated
> > to be a bit too dark.

> Your image viewing pipeline must be really, really badly calibrated; 
> from the shots you put on the 'net, I would instead expect you to 
> experience most images as way too bright.

  My experience is the exact opposite: Often images which others report are
fine (even if a bit dark) look almost completely black in my monitor.

  Search my posts in p.b.images about this subject if you like.

-- 
                                                          - Warp

Post a reply to this message

clipka wrote:

> 
> The 3.7 shot looks /perfect/ on my display, no matter whether I display 
> it in Internet Explorer, Windows Explorer preview, or Photoshop 6.0. 
> Given that I did invest some time into calibrating my primary display, 
> this tells me that yours/ must be /way/ off.
> 

I need to run my colorimeter on my display again, but the 3.6 grayscale 
looks better than the 3.7 grayscale on my screen. The 3.7 grayscale does 
look too bright to me. I'll have to make sure my display at home is 
calibrated to see how it looks...

> So stop your griping, disengage "demand mode", calibrate your image 
> viewing pipeline properly, and /then/ come back with any residual problems.

See? this is the whole problem with gamma correction, if everyone 
doesn't have a properly calibrated screen, and doesn't properly embed a 
color profile in the image, or the viewing app does not handle the 
embedded profile correctly, then you get all sorts of different results.

-- 
~Mike

Post a reply to this message

clipka <ano### [at] anonymousorg> wrote:
> Warp schrieb:

> > http://warp.povusers.org/pov_imagemap_test/index.html

> >   Granted, the perceived brightness of this checkerboard pattern does not
> > match the 0.5 grey of the POV-Ray 3.6 image. However, neither does it match
> > the one rendered with POV-Ray 3.7. In fact, the latter deviates *more* from
> > that perceived brightness than the former. It's way too bright. At least on
> > my screen.

> The 3.7 shot looks /perfect/ on my display, no matter whether I display 
> it in Internet Explorer, Windows Explorer preview, or Photoshop 6.0. 
> Given that I did invest some time into calibrating my primary display, 
> this tells me that yours/ must be /way/ off.

  You are seriously claiming that my display is way, way too bright.
That really makes me laugh out loud.

  If anything, my display is too dark. Often images posted by people in
p.b.images look almost black. When I complain about it, others say they
can see it just fine.

  Now you are telling me that I have to calibrate my monitor to be something
like 25% darker than it already is. Everything in the darker 20% or so
darkest color space would become completely black and invisible.

  Are you also seriously claiming that the starfield image in the pov3.7
rendered image looks just fine? Does that mean that the original looks like
a black square with some tiny white pixels here and there in your monitor?

> So stop your griping, disengage "demand mode", calibrate your image 
> viewing pipeline properly, and /then/ come back with any residual problems.

  I don't appreciate your condescending tone of voice. I'm trying to discuss
about an issue here.

> As for entering gamma-pre-corrected colors into POV-Ray, you might use 
> something like this:

>    #macro UnGamma(C)
>      #local G = 1/2.2;
>      <pow(C.red,G),pow(C.green,G),pow(C.blue,G),C.filter,C.transmit>
>    #end

>    #declare MyGrey = UnGamma(rgb <127,127,127>/255);

  I don't get it. The image POV-Ray 3.7 is producing by default is way too
bright, and you want me to brighten the colors even further? (For example
rgb 0.1 becomes rgb 0.35, rgb 0.5 becomes rgb 0.73.)

  Yeah, that will look marvelous. After your UnGamma converts the original
rgb 0.1 to rgb 0.35, POV-Ray will render the image with a further gamma
correction, making the final pixel have an rgb of 0.62. That will really
fix all problems.

-- 
                                                          - Warp

Post a reply to this message

Mike Raiford schrieb:
> See? this is the whole problem with gamma correction, if everyone 
> doesn't have a properly calibrated screen, and doesn't properly embed a 
> color profile in the image, or the viewing app does not handle the 
> embedded profile correctly, then you get all sorts of different results.

So, what's the alternative?

Note that working with linear colors is not a solution: Things being as 
they are, it just adds to the many nonstandard display gammas (1.0 in 
this case), opens up more cans of worms than going for the standard 
gamma of whatever OS you happen to be using (which would be 2.2 for 
Windows), and is therefore prone to being part of the problem instead.

Post a reply to this message

clipka wrote:
> Warp schrieb:
> 
>> http://warp.povusers.org/pov_imagemap_test/index.html
> 
>>   Granted, the perceived brightness of this checkerboard pattern does not
>> match the 0.5 grey of the POV-Ray 3.6 image. However, neither does it
>> match the one rendered with POV-Ray 3.7. In fact, the latter deviates
>> *more* from that perceived brightness than the former. It's way too
>> bright. At least on my screen.
> 
> The 3.7 shot looks /perfect/ on my display, no matter whether I display
> it in Internet Explorer, Windows Explorer preview, or Photoshop 6.0.
> Given that I did invest some time into calibrating my primary display,
> this tells me that yours/ must be /way/ off.
> 
> Now, that isn't generally a problem: You can work with a display system
> gamma of 1.0 if you like. But in that case make sure your image viewing
> software supports non-standard display gammas, and is actually
> configured to do so.
> 
> So stop your griping, disengage "demand mode", calibrate your image
> viewing pipeline properly, and /then/ come back with any residual
> problems.

Last time I tried to "calibrate my display" to show gamma correctly (that
is, to show 0.5 and {0.0, 1.0} checkerboard the same), *everything* looked
too washed out, not only pictures but mainly GUI elements.

Post a reply to this message

clipka wrote:
> Mike Raiford schrieb:
>> See? this is the whole problem with gamma correction, if everyone 
>> doesn't have a properly calibrated screen, and doesn't properly embed 
>> a color profile in the image, or the viewing app does not handle the 
>> embedded profile correctly, then you get all sorts of different results.
> 
> So, what's the alternative?

How do other applications deal with color?

Embed the sRGB profile with 2.2 gamma, perhaps, would be an alternative. 
Just make sure the image is displaying as intended in the sRGB space, 
with an app that properly handles color correction.

> Note that working with linear colors is not a solution: Things being as 
> they are, it just adds to the many nonstandard display gammas (1.0 in 
> this case), opens up more cans of worms than going for the standard 
> gamma of whatever OS you happen to be using (which would be 2.2 for 
> Windows), and is therefore prone to being part of the problem instead.

Agreed. Like I stated, the whole thing is a terrible mess. I think 
there's some transform to the color values from the renderer output to 
the color space (As suggested above) that would be needed in order to 
get a proper image. Then, it simply isn't just gamma, but also how the 
red, green, and blue values relate to each other. Getting color 
management right is a freaking mess.

I could suggest render linearly, create a color profile for that, load 
the image in PS, convert to the sRGB profile, the save as a JPEG or PNG, 
but that would be a major PITA ;)

Note that I suggest sRGB here, because it's the profile that should be 
"right" for the majority of the web users. Most computer equipment these 
days operates in that color space.

-- 
~Mike

Post a reply to this message

Warp <war### [at] tagpovrayorg> wrote:
>   Now the lower half of the rendered images is a checkerboard pattern of
> alternating black and white pixels.

  There was a warning somewhere that a checkerboard patter should *not* be
used for this kind of test because it doesn't work properly with CRT monitors.
Instead, alternating horizontal lines should be used instead.

  I changed the images to use horizontal lines. Now the POV-Ray 3.7 image
background looks equal in brightness, while there's a remarkable difference
in the POV-Ray 3.6 image.

  It was not, after all, a problem with my monitor gamma. It was a problem
with the test being skewed by how CRT works.

  So yes: I now see how the gamma correction of 2.2 changes the rgb 0.5 to
a shade which closely matches half brightness emulated with alternating
black and white lines. (The gray stripe at the bottom still looks to me
like biased too bright, but maybe it's a question of getting used to it.)

  There's still the problem of the image map incorrectly brightening, though.

-- 
                                                          - Warp

Post a reply to this message