Brilliant explanation of color etc

[Robert Ardill]

I haven't gotten very far, but this made me react: "It also suggests there are colors a Canon camera can distinguish that we cannot."

He's talking about capture metameric error, AFAICT. If that's right, the author's definition of color is different from mine. I interpret color to mean normal human tristimulus response to input spectra. It looks like the author is using color to mean response of some arbitrary three-part capture system to spectral input. To my mind, color is defined by human response. If two spectra match for, say, the 1931 observer, they are the same color. If they don't, they aren't. It's a little wiggly because there is more than one standard observer, but you get the idea.

Jim

I think it would be worth reading through to the end of the notes (there's a link to the next 'applet' as he calls it, and one after that ...) as page 1 is just the introductory page.  There are also other 'applets' that cover color mixing etc., also well worth looking at (for those of us who have little knowledge of color theory, that is).

Also these are additions to lectures and lecture notes for an undergraduate course, so the author is being a bit loose in using terms like 'color', including what some animals can see and which lie well into the IR or UV frequencies (and so well outside of human vision).

Robert

[bjanes]

This is something that Bill might be able to explain to us better.  My understanding is this:

If that's correct, then the gamut of a person's vision and the gamut of a camera will not be the same ... at least not without correction to the camera's SPD.

But perhaps Bill, or someone who understands these things fully, could clarify things for us.

Robert

Robert,

If I am the Bill to whom you are referring, thanks for the compliment, but your understanding in these matters appears greater than mine.  I am in no way an expert, but am interested in these color matters. Jim Kasson has entered into some of these threads, and he really knows his stuff.

Bill

[Simon Garrett]

...these are additions to lectures and lecture notes for an undergraduate course, so the author is being a bit loose in using terms like 'color', including what some animals can see and which lie well into the IR or UV frequencies (and so well outside of human vision).

Quite so.  This material is part of Stanford University's Digital Photography course for enrolled Stanford students, not a complete set of material for our use.  The fact that there is a great deal of information online for Stanford students that we can read is our good fortune and not Marc Levoy's intention. 

PS - you may know that the Lytro light field camera that Ren Ng and others developed originated from the same department (Computer Graphics) at Stanford, who are pretty strong on computational photography. 

[Jim Kasson]

I finished looking at all the applets and reading all the text. I think the applets are marvelous, and the text quite good.

Of course, I have a few quibbles.

the reference white is set for you by the manufacturer, typically to a slightly blueish-white called D6500.  

They mean D65, not D6500. Actually, D65 defines a spectral power distribution, not a color, but you can turn it into a color by specifying an observer.

In closing, one question you might be wondering about is: how were the X, Y, and Z axes chosen? They were clearly chosen to ensure that the spectral locus fell into the all-positive octant of XYZ colorspace. Beyond this, the answer is complicated and beyond the scope of this applet; the interested reader is referred to (Wyszecki 1982)

I consider referring non-color scientists to Wyszecki and Stiles to be cruel and heartless. There are many more approachable books, although W&S is an indispensable reference if you really know what you're doing.

It’s great that they stayed in three dimensions as long as they did. I think they should have talked about the math to get from rho,gamma,beta to XYZ (not at all difficult for any Stanford student), and they should have defined xy and u’v’ chromaticity space instead of talking about "XYZ chromaticity space".

I’m disappointed that the gamut mapping exercise was in two dimensions, especially since printers are explicitly called out. They do note and sort of apologize for this, however.

I love the way they make it easy to create 4 and 5 primary simulated emissive displays.

None of these comments are meant to take away from the worth of the material. It is very well done.

Jim

[Robert Ardill]

Robert,

If I am the Bill to whom you are referring, thanks for the compliment, but your understanding in these matters appears greater than mine.  I am in no way an expert, but am interested in these color matters. Jim Kasson has entered into some of these threads, and he really knows his stuff.

Bill

Well I wish I did know more, Bill (and you are the Bill I was referring to) ... but my knowledge is certainly not on a par with yours.  I'm just trying to figure out how to get the best from my images and every now and then I come up against a brick wall, so I do a bit of digging to see if I can get through to the other side.

It's really great that Jim is helping us out ... but to be honest, sometimes it's hard to follow people who know so much (this is just a little hint to Jim to try to keep it simple .  

Robert

[Robert Ardill]

I consider referring non-color scientists to Wyszecki and Stiles to be cruel and heartless. There are many more approachable books, although W&S is an indispensable reference if you really know what you're doing.

Jee! (is that the correct Americanism?).  How wonderful to find someone with a sense of humour!!

I’m disappointed that the gamut mapping exercise was in two dimensions, especially since printers are explicitly called out.

Yes, this is something I also felt was sorely missing.

Robert

[Jim Kasson]

Oh, and another thing. Allowing the viewer to act like a subject in the color matching experiment is just great.

Jim

[Robert Ardill]

Oh, and another thing. Allowing the viewer to act like a subject in the color matching experiment is just great.

Jim

Yes, actually it is quite fascinating.  In my case at least it seems that I am far more sensitive to hue than to intensity at the green part of the spectrum, whereas at the red end I seem very sensitive to both (or perhaps it's my monitor :



At 510nm I can vary the green intensity wildly and see no difference, while at 605nm changing any of the primaries makes quite a difference.  I would have thought that as we are (supposedly) more sensitive at the green frequencies that small differences there would be more noticeable.

Seems there is more here than meets the eye!

Robert

[MarkM]

Yes, actually it is quite fascinating.  In my case at least it seems that I am far more sensitive to hue than to intensity at the green part of the spectrum, whereas at the red end I seem very sensitive to both (or perhaps it's my monitor :

At 510nm I can vary the green intensity wildly and see no difference, while at 605nm changing any of the primaries makes quite a difference.  I would have thought that as we are (supposedly) more sensitive at the green frequencies that small differences there would be more noticeable.

No monitor will reproduce 510nm spectral light. The closest it will get is somewhere around 100% green in your monitor's color space. If you look at a chromaticity diagram of you're monitor's space (or sRGB for a ballpark non-wide gamut monitor) you'll see there's a long distance between the monitor primary and the edge of the diagram. I suspect what you are seeing is simple clipping. Beyond a certain intensity your monitor isn't going to display a different color — it will just give you the closest it can. On my monitor if I move only the green slider, the color stops changing at about .18.

[Robert Ardill]

No monitor will reproduce 510nm spectral light. The closest it will get is somewhere around 100% green in your monitor's color space. If you look at a chromaticity diagram of you're monitor's space (or sRGB for a ballpark non-wide gamut monitor) you'll see there's a long distance between the monitor primary and the edge of the diagram. I suspect what you are seeing is simple clipping. Beyond a certain intensity your monitor isn't going to display a different color — it will just give you the closest it can. On my monitor if I move only the green slider, the color stops changing at about .18.

There you go ... isn't it great when we can get a clear explanation that's based on science and reason and not opinion!!

Thanks

Robert

[Jim Kasson]

I consider referring non-color scientists to Wyszecki and Stiles to be cruel and heartless. There are many more approachable books, although W&S is an indispensable reference if you really know what you're doing.

I've thought some more about this point. It's really not that unusual in academic circles to go from simple to extremely complex in one jump.

One of my favorite examples, which may be dated at this point, is going from "See Spot Run" toy programming languages straight to C in high school programming courses. To my mind, having a newbie programmer writing code in C is like starting a neophyte juggler off with chainsaws.

Why do we do this? I can think of several reasons:

It's what the teacher, who is presumably expert, knows. Possibly the teacher isn't spending a lot of time evaluating intermediate texts, unless she happened to write one. By the way, another Stanford prof, Brian Wandell, wrote an excellent mid-level text on the psychology of vision, called Foundations of Vision. To me, it's right up there with Cornsweet's classic Visual Perception.

To give the student an idea of how complicated the field really is, and maybe scare off the dilettantes.

To give those students who will go on to make a career in the field an introduction to the tools they will use on the job.

That's the reasons I can come up with that aren't snarky.

Jim

[Robert Ardill]

I've thought some more about this point. It's really not that unusual in academic circles to go from simple to extremely complex in one jump.

One of my favorite examples, which may be dated at this point, is going from "See Spot Run" toy programming languages straight to C in high school programming courses. To my mind, having a newbie programmer writing code in C is like starting a neophyte juggler off with chainsaws.

Why do we do this? I can think of several reasons:

It's what the teacher, who is presumably expert, knows. Possibly the teacher isn't spending a lot of time evaluating intermediate texts, unless she happened to write one. By the way, another Stanford prof, Brian Wandell, wrote an excellent mid-level text on the psychology of vision, called Foundations of Vision. To me, it's right up there with Cornsweet's classic Visual Perception.

To give the student an idea of how complicated the field really is, and maybe scare off the dilettantes.

To give those students who will go on to make a career in the field an introduction to the tools they will use on the job.

That's the reasons I can come up with that aren't snarky.

Jim

Yes ... having been a lecturer in Engineering I can think of quite a few snarky reasons

But this guy sounds like a decent sort, so maybe we could be kind and assume it was a slip of the pen.  Or maybe he's got a sense of humour.

Robert