On Wed, 18 Dec 2002 10:35:01 +0800, Bill Brehm wrote:

> Thanks.
> 
> I was looking in the file output section, so I didn't stumble across
> hf_gray_16 in the global settings. I tried it and it seems to work; at
> least I'm getting gray output. While I can tell the difference between
> color and grayscale, I can't with my eyes determine the number of
> grayscale bits. So I have a few additional questions.
> 
> 1a. If I need only 10 bit grayscale (or any other number between 8 and
> 16), will Bits_Per_Color=10 only allow 10 bits of grayscale info to be
> written into the file, assuming Output_File_Type=N?

5.2.2.3.1 claims that POV-Ray will output 5-16 bits.
I think I remember reading that PNG files actually use either 8 or 16 bits
per channel but outputting less eg. 10 bits will make the files smaller
since the extra bits are all zeros and compress very well.

> 
> 1b. If I use Output_File_Type=N, I'll get a PNG file which can accept up
> to 16 bits of grayscale info. If I use Output_File_Type=S, will I get a
> BMP file with only 8 bits of grayscale info? (This will be helpful to me
> too in some cases, as I've been generating 24 bit color images and
> converting them to grayscale with a separate utility and maybe now I can
> generate the 8 bit grayscale BMP files directly.)
> 

Not sure what happens to BMP output with hf_gray_16

> 2. I was not aware of ImageMagick. I will download it tonight and try it
> out.
> 
> 3. The reason I am using grayscale is because I need to generate some
> images to test what a monochrome analog or digital camera will see in a
> piece of machinery I'm working on, in order to test some computer vision
> algorithms. This leads me to a question that's not directly POVRay
> related, but someone out there may know. I see that hf_gray_16 causes
> POVRay to use the luminance formula to convert the color image to gray. I
> understand the reason for the .3, .59 and .11 parameters is because that
> matches the human eye's sensitivity to the three primary colors of light.
> Will a monochrome camera, as used in most equipment with machine vision,
> have a similar sensitivity? To put it another way, will a color camera
> looking at a color scene and having its image converted to grayscale with
> the luminance formula produce the "exact" same image as a monochrome
> camera looking at that same color scene?

There is some variation in the contribution of the 3 colour components in
different systems 0.3,0.59,0.11 seems fairly common but a particular
camera may give a slightly different result.

> 
> 4. What is the purpose of a height field image? Any example scenes
> available on this news server that show its application?
> 

The best answer to this question is in the documentation 6.5.1.5
Nearly every landscape will have at least one heightfield.

-- 
PoD.

Post a reply to this message