 |
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Hello,
This is an old scene with some slight improvements. Getting the rainbow gradient
right is tricky. Making a mathematically linear gradient doesn't seem to give
visually linear results. The primary colors appear to dominate the secondary
colors, causing uneven transitions. This scene uses a partial fix but it's still
not perfect. I have seen what appear to be perfectly balanced rainbow gradients
before - does anybody know the trick to getting it just right?
Regards,
Dave Blandston
Post a reply to this message
Attachments:
Download 'atari.jpg' (74 KB)
Preview of image 'atari.jpg'
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Am 24.01.2011 18:17, schrieb Dave Blandston:
> This is an old scene with some slight improvements. Getting the rainbow gradient
> right is tricky. Making a mathematically linear gradient doesn't seem to give
> visually linear results. The primary colors appear to dominate the secondary
> colors, causing uneven transitions. This scene uses a partial fix but it's still
> not perfect. I have seen what appear to be perfectly balanced rainbow gradients
> before - does anybody know the trick to getting it just right?
What assumed_gamma are you using?
While assumed_gamma 2.2 gives visually more pleasing dark-to-bright
gradients, assumed_gamma 1.0 generally seems to give more pleasing
color-to-color gradients.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"Dave Blandston" <nomail@nomail> wrote:
> Hello,
>
> This is an old scene with some slight improvements. Getting the rainbow gradient
> right is tricky. Making a mathematically linear gradient doesn't seem to give
> visually linear results. The primary colors appear to dominate the secondary
> colors, causing uneven transitions. This scene uses a partial fix but it's still
> not perfect. I have seen what appear to be perfectly balanced rainbow gradients
> before - does anybody know the trick to getting it just right?
Somewhere (in the POV docs or elsewhere) you can find the relative intensities
for RGB values. I have no idea where to find that information anymore. I know
for a fact that green overpowers the other two.
Another thing you can try is to find an image of the gradient you want, and fill
up a color_map with the colors from it. You can use eval_pigment (slow), or you
can make a function from the image and test it directly. You might still have
gamma issues, but the result should be close to what you want.
~Sam
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"Samuel Benge" <stb### [at] hotmail com> wrote:
> "Dave Blandston" <nomail@nomail> wrote:
> > - does anybody know the trick to getting it just right?
>
> Somewhere (in the POV docs or elsewhere) you can find the relative intensities
> for RGB values. I have no idea where to find that information anymore. I know
> for a fact that green overpowers the other two.
....It seems that I was remembering a method for converting a color image to
grayscale, though it might still be applicable here since we're talking about
the visual influence of colors...
From the POV docs, in reference to grayscale output:
"The brightness of a pixel is calculated in the same way that color images are
converted to grayscale images: height = 0.3 * red + 0.59 * green + 0.11 * blue."
Maybe this helps, maybe it doesn't :/
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
> Hello,
>
> This is an old scene with some slight improvements. Getting the rainbow gradient
> right is tricky. Making a mathematically linear gradient doesn't seem to give
> visually linear results. The primary colors appear to dominate the secondary
> colors, causing uneven transitions. This scene uses a partial fix but it's still
> not perfect. I have seen what appear to be perfectly balanced rainbow gradients
> before - does anybody know the trick to getting it just right?
>
> Regards,
> Dave Blandston
A simple way: Add some more intermediate tints to your colour_map.
If you need, say, the green to be visualy narower, shift the neiboring
entries closer a bit.
Alain
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Thanks for the suggestions. Using more intermediate values is an effective fix.
I may tinker around with some of the other ideas too - this is an interesting
topic to me.
Regards,
Dave Blandston
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
