|
 |
The Google has specified an official color for Its Android mascot: #A4C639.
When Karl Ostmo submitted AndroidRobot to the Object Collection, he included the
official color in a comment, but used MediumSpringGreen in production. He did
not explain the discrepancy, but my best guess is that his linear conversion of
#A4C639 didn't look very official when rendered.
I resubmitted AndroidRobot with the official color last July, but applied gamma
correction to the color under the ASSumption that normal people set
assumed_gamma 1. The other day, recalling last September's epic arguments in
povray.advanced-users, and still somewhat convinced that Warp is a normal
person, I decided that my ASSumption was inappropriate, and added the
uncorrected color as an option. To assess the different religious views, I ran
a test scene with three variables:
- the POV #version
- assumed_gamma
- the official color with and without gamma correction.
The new sRGB feature in 3.7 gives interesting results. When assumed_gamma is
2.2, the sRGB looks the same as the uncorrected color. When assumed_gamma is
greater than 2.2 (not shown in the attachment), the sRGB is /lighter/ than the
uncorrected color.
Notes:
- All 8 frames were rendered with 3.7 RC6. The version behavior was
controlled with #version statements.
- I used radiosity and left the finish { diffuse } at its default
value.
- For easy visual comparisons, the border color is literally #A4C639.
- I initially tried omitting assumed_gamma as an uncorrected control.
However, this caused a parse error in frames 5 and 6, due to the way
androidrobot.inc computes the colors, so I used assumed_gamma 2.2 as
the control.
- The colored balls are linear conversions of #FF0000, #FF8000,
#FFFF00, #008000, #0000FF, and #800080, respectively.
- Pay close attention to the hues of the orange balls. Orange is the
only one of the colors with differential values between channels.
- The green and purple use fractional values, but there is no
differential between channels. The result is differing lightness
values, but no differences in hue.
The Google wishes to remind you that the robot is based on work created and
shared by the Android Open Source Project and used according to terms described
in the CC-BY-2.5 license.
Post a reply to this message
Attachments:
Download 'android_gamma_montage.png' (361 KB)
Preview of image 'android_gamma_montage.png'
|
|
|
|
On 06/29/2012 05:53 AM, Cousin Ricky wrote:
> - I used radiosity
I think that's skewing the results.
AFAIK using assumed gamma 2.2 and the uncorrected color (in pov3.7)
should give the same end result as using assumed gamma 1.0 and
srgb-corrected color.
However, since you are applying the srgb only to the robot, everything
else becomes brighter with assumed gamma 1.0, making the radiosity
lighting also brighter, making the color of the robot likewise look
brighter.
Post a reply to this message
|
|
|
|
Warp <war### [at] tag povrayorg> wrote:
> On 06/29/2012 05:53 AM, Cousin Ricky wrote:
> > - I used radiosity
>
> I think that's skewing the results.
Here's another run, with the default diffuse and ambient increased so that they
add up to 1.0.
Post a reply to this message
Attachments:
Download 'android_norad1_montage.png' (243 KB)
Preview of image 'android_norad1_montage.png'
|
|
|
|
On 06/29/2012 10:10 PM, Cousin Ricky wrote:
> Warp<war### [at] tagpovrayorg> wrote:
>> On 06/29/2012 05:53 AM, Cousin Ricky wrote:
>>> - I used radiosity
>>
>> I think that's skewing the results.
>
> Sans radiosity attached. Looks kind of flat.
I think the default ambient is now brighter by a power of 2.2, which is
why the shadows are so much brighter. You might want to adjust that as well.
(Perhaps the easiest way is to simply have ambient_light 0 in all the
cases. That way it has no effect on anything.)
The image will still not be identical, though, because the shading will
be different with a different assumed_gamma. However, the color ought to
look about the same at the brightest parts.
Post a reply to this message
|
|
