[jfrancis]

I.

The pdf docs say that in a diffuse material, [R + G + B] / 7.65 should be less than 100.

This is because 255 + 255 + 255 = 765.

Why are colors being described in 0..255 space as if they were 8-bits? Is this for the comfort of Photoshop users?

I would stop doing this and specify all colors in floating point 0..1 space.

I don't think calling white (255 255 255) is a good way to go.

II.

How is it that colors are specified in Maxwell as RGB triples? Shouldn't we be seeing spectral waveforms? or 12 samples? or at least 9? Is this for the convenience of the user? Isn't the point of Maxwell to get away from thinking of colors as only 3 spectral samples, R G and B?

III.

If you have a good system for converting from 3 spectral samples to some higher dimensional space and back down to 3 again, could you partner up with Corel Painter and work out multi-spectral pigment color mixing for them? (just a thought)

[tom]

Hi jfrancis,

You are absolutely right and these are just temporary settings for the materials.
With the new material editor, you'll have more control over color range.

Best regards,
Tom

[bugyboo]

that's right

we'all so anxoius to see "the new material and GUI for maxwell"....

[jep]

jfrancis,
So glad you brought that up - I get banding everywhere! Floating point is the only way to go. I'm assuming all the banding I'm seeing is a result of this temporary 8bit space.

Tom, heya! Will we have the ability to save out floting point images from Maxwell in other formats than .mxi?

[Pietro Spampatti]

good and coherent observation, but in 1.0 version don't forgotten the usability and a clear User Interface.... I think than the right way will be a preset of UI (simple, normal expert) with the possibility of personalization

P.

[stonelli]

I think than the right way will be a preset of UI (simple, normal expert) with the possibility of personalization

P.

Good point

[Becco_UK]

I prefer R/G/B as I can relate this to other software I use but for those wanting something more 'techy' then Pietro Spampatti makes a good point.

[Mihai]

Changing to H|S|V is not a big adjustment step and it would be a system which could work with "spectral energy" since the hue would represent the color spectrum.

[jfrancis]

becco_UK:

But there are multiple "wavelength recipes" for each RGB triple.

Take yellow, for example.

If you specify yellow as (255 255 0) then how do you talk about the yellow that is NOT made of red and green? How would you specify the yellow that resides at around 570 nanometers and that cannot be split by a prism into red and green?

http://eosweb.larc.nasa.gov/EDDOCS/Wave ... olors.html

[stonelli]

I prefer R/G/B as I can relate this to other software I use but for those wanting something more 'techy' then Pietro Spampatti makes a good point.

Obvoiusly there will need to be some way to convert the RGB data we get from swatches to whatever data MR needs. Otherwise, how will we ever match the colour chip we want for our walls

[DELETED]

DELETED

[stonelli]

..witch every coyp of maxwell NL will ship one of these i heard

Woa, they must have deep discount

[jfrancis]

You wouldn't see a difference between the two yellows in primary rays, but as things bounce around, reflect and refract, you'd start to see differences in the muddiness or richness of the colors in those areas.