[Req] Allow more colorspaces to enter color data

General questions about Indigo, the scene format, rendering etc...
Post Reply
9 posts • Page 1 of 1
User avatar
galinette
1st Place Winner
Posts: 923
Joined: Sat Jan 09, 2010 1:39 am
Location: Nantes, France
Contact:

[Req] Allow more colorspaces to enter color data

Post by galinette » Sun Jan 31, 2010 5:13 am

Dear Indigo team,

It would be very interesting to be allowed to enter colors as CIEL*a*b* , CIExYz or CIEXYZ data.

Maybe that's more an exporter thing to allow multiple colorspaces, but in that case it would be much better to have one of these complete color spaces instead of RGB. RGB is not easy to use for professionnal color use, as there are several (an infinite number actually) definitions of the RGB color space. Moreover, RGB does not covers the full spectral gamut, and it's a pity to enter limited gamut data in a full spectral renderer.

For instance, I have these kind of colorimeters at work (http://www.colorcontroller.com/en/product-h.asp?id=19) that gives instantly the color of any surface in L*a*b*. That's difficult to correctly and safely convert that to RGB (and sometimes impossible if you're outside the gamut).

Thanks for your interest in making Indigo more suitable for professionnal use!

Etienne
Eclat-Digital Research
http://www.eclat-digital.com

User avatar
galinette
1st Place Winner
Posts: 923
Joined: Sat Jan 09, 2010 1:39 am
Location: Nantes, France
Contact:

Re: [Req] Allow more colorspaces to enter color data

Post by galinette » Sat Feb 06, 2010 3:05 am

Hey, no answers... Nobody is interested by real pro non-clipped colorspaces?

Etienne
Eclat-Digital Research
http://www.eclat-digital.com

User avatar
pixie
Indigo 100
Posts: 2346
Joined: Sat Dec 29, 2007 4:54 am
Location: Away from paradise
3D Software: Cinema 4D
Contact:

Re: [Req] Allow more colorspaces to enter color data

Post by pixie » Sat Feb 06, 2010 4:20 am

galinette wrote:Hey, no answers... Nobody is interested by real pro non-clipped colorspaces?
Of course they are, they just don't understand what this all means... :D

I wonde if by saving in exr one could, somehow, convert into this color spaces. :?

User avatar
galinette
1st Place Winner
Posts: 923
Joined: Sat Jan 09, 2010 1:39 am
Location: Nantes, France
Contact:

Re: [Req] Allow more colorspaces to enter color data

Post by galinette » Sat Feb 06, 2010 4:39 am

Yes, if correctly color-encoded, EXR is able to use negative color values (like negative R, G or B float values...). At least, if the generating software did not clip them. And if the reading & tonemapping software correctly processes negative values, then it should be possible. I have never tried that way with photoshop, but I will...

CIELab tiff output would definitely be the best option for photoshop users (It's robust and well supported, unlike EXR)

For everybody, this can be an advantage if you want to print your art on photo paper by a pro lab. Save to CIELab, tonemap in photoshop, then save in Adobe RGB. You will have much nicer saturated colors on the print, that most monitors are not able to display.

CIELab colorspaces & TIFF format are quite well documented.

Etienne
Eclat-Digital Research
http://www.eclat-digital.com

User avatar
Doug Armand
Indigo 100
Posts: 1038
Joined: Fri May 16, 2008 5:49 pm
Location: London UK

Re: [Req] Allow more colorspaces to enter color data

Post by Doug Armand » Sat Feb 06, 2010 4:40 am

I think it would be a nice option to add Lab support - it's often used by Photoshoppers so couldn't hurt
Doug

Doug Armand

User avatar
PureSpider
Posts: 1459
Joined: Tue Apr 08, 2008 9:37 am
Location: Karlsruhe, BW, Germany
Contact:

Re: [Req] Allow more colorspaces to enter color data

Post by PureSpider » Sat Feb 06, 2010 9:01 am

How can I imagine CIELab to work?
I mean I can imagine RGB as components and HSL/HSV as a "tube" (http://upload.wikimedia.org/wikipedia/c ... linder.jpg).
But how does CIELab work?

User avatar
dougal2
Developer
Posts: 2532
Joined: Wed Nov 15, 2006 8:17 am
Location: South London

Re: [Req] Allow more colorspaces to enter color data

Post by dougal2 » Sat Feb 06, 2010 10:36 am


User avatar
OnoSendai
Developer
Posts: 6244
Joined: Sat May 20, 2006 6:16 pm
Location: Wellington, NZ
Contact:

Re: [Req] Allow more colorspaces to enter color data

Post by OnoSendai » Sat Feb 06, 2010 11:57 am

Hi Galinette,
Are you wanting the extra colour spaces for output data (render save format), or for input data (textures, constant material colours etc..)?

User avatar
galinette
1st Place Winner
Posts: 923
Joined: Sat Jan 09, 2010 1:39 am
Location: Nantes, France
Contact:

Re: [Req] Allow more colorspaces to enter color data

Post by galinette » Sat Feb 06, 2010 9:54 pm

Hi Nick,

Most important for me is input data now, since I mostly work in L*a*b* for glass (and also on spectra). And I have all the hardware to measure L*a*b* on any flat surface.

Output would be also interesting later on for me, as soon as we have invested in monitors able to display AdobeRGB, and for making photo quality color-proofed prints. But we're not there yet.

PureSpider : Lab is roughly a sphere : bottom point is black, top point white, and vertical axis are greys. When you go away from the vertical axis, you increase saturation, and when you turn around, you change hue. However the main advantage of CIELab is not its layout (other shemes such as HLS are OK for this) but it's perceptive, non clipped nature. CIELab indexes all the colors visible by (standardized) human eye. CIEXYZ and CIExYz have different layouts but they are equivalent. On the other hand, systems such as RGB or sRGB describe all the colors reproducible by combining three colored light sources (such as your monitor's pixels). Unfortunately, it's theoritically impossible to reproduce all the visible colors with three light sources, because that would require sometimes your pixels to emit negative light intensities (this is related to linear algebra in spectral space, 3d subspaces of spectral space, and base change concepts). However, some systems like CYMK printers are much more color capable than monitors. But to keep all the color info you must never convert to RGB!

Etienne
Eclat-Digital Research
http://www.eclat-digital.com

Post Reply
9 posts • Page 1 of 1

Who is online

Users browsing this forum: No registered users and 4 guests