[rescue] Spaceballs.

Joshua D Boyd rescue at sunhelp.org
Sat Dec 8 14:32:57 CST 2001


On Sat, Dec 08, 2001 at 02:15:51AM -0600, Scott Newell wrote:
> >Also, changing color spaces altogether (say from YUV or HSV to RGB) is also 
> >just another matrix multiply.  This means that I don't need to do a lot of 
> 
> Does that matrix operation also translate the cubic RGB color space into a
> hexcone for HSV?  I'm having trouble visualizing some of the possible color
> space system transforms as 3d geometric transforms.

I think it works, but I'd have to double check.  Matrices are only for linear
transformations, so the word Hexcone worries me, but I know I've seen matrices
given for HSV operations...

An interesting article on the topic is at: 
http://www.sgi.com/grafica/matrix/index.html

BTW, a rotation in RGB color space is the same as a hue rotation in HSV 
apparently.
 
> 
> >Now other areas for color work would be using FFTs on them.  I've long 
> >understood FFTs in the context of sound, but I'm slowly getting a handle on 
> >what they mean in the context of graphics.  FFTs can be used for interesting
> >effect for picking the detail of an image appart from the shading.
> Basically,
> >(I got this from an article in a game programming magazine interestingly), I
> 
> Was the game developer article a couple of months ago discussing image
> compression with digital filters to keep edge detail?

That's an interesting idea, but the article was mainly about cleaning up 
lighting problems in textures.  I saw it is Game Developers Magazine.  I don't
know if they also posted in to the Gamasutra web site.

> >can use a FFT to split an image into to layers, a high frequency layer,
> and a 
> >low frequency layer.  The two layers summed are the original image.
> However, 
> >now I can edit them seperately for interesting effect.
> 
> Now extend the idea of a transform to your 3d color space of a scene...what
> does that look like?  Interesting...I'd never really considered a 3d FFT,
> much less whether it can be meaningfully applied to a color space.

Apparently FFTs can be applied to any number of dimensions.  However, I also
understood an FFT as something that required 2 inputs, a time and a value, and
returned to inputs, a frequency for that value.  In fact, most descriptions 
I've read about FFTs state that it is a transformation from the time domain
to the frequency domain.  Well, fine.  But, what is the time domain for a 
single image?  Which is to day that I'm far from understanding the forces at
work here.

Anyway, hope this provides people with some ideas.  Followups should probably
go to geeks.

-- 
Joshua D. Boyd



More information about the rescue mailing list