Can Film Actually Capture the Color Violet

[Daniel Lee]

To clarify;

 

As above colour is a perception phenomenon. Your eye has no violet cones. Violet influences the response of other cones outside the wavelength of a violet light. The colour you perceive is violet, and may be a false-colour/colour-crossover effect, but it is the real colour violet, as again it is a perception thing. Therefore anything that can respond in the same colour-crossover manner as the eye does can reproduce the true colour violet according to the human perceptual experience regardless of any missing wavelength.

[John Sprung]

OK, try a thought experiment. Take a light meter and three strongly colored filters, red, green, and blue. For these colored light sources you're thinking about, what measurement would you get through each of the three filters?

 

Guess what -- that's how the cones in your eyes work. That's the meaning of the first graph that Ben posted, it's the filter curves of the three photopigments in your eyes.

 

Different combinations of wavelengths and intensities can result in the same readings with the light meter and filters, and look the same to your eyes.

 

Film doesn't reproduce specific wavelengths or combinations of wavelengths, and it doesn't have to. It just makes other combinations that look the same.

 

 

 

-- J.S.

[Chris Millar]

I've read most of this thread and gotten a bit cross eyed on it...

 

I'll attempt an outsiders understanding, not expecting it to be right, but the clarification of it will likely help me understand:

 

Violet wavelengths will be picked up by the different layers in film according to the spectral sensitively of each layers ability to pickup the 'violet' wavelength...

 

So film has captured violet as an input

If you were to film this violet in isolation the resulting colour once the film has been processed and projected 'all things being equal' - i.e. normally - will not give the same look/wavelength as the original violet, but will give at least a relatively violet impression.

 

But, there is some further capability in film to create a 'more' violet output, given another set of exciting factors in its exposure than real/on set violet.

 

A disconnect between input violet and output violet ?

 

And hence the confusion and unnecessary discussion at cross purposes

 

...or, I'm talking out of my bum ?

[Simon Wyss]

Let me try again.

 

Violet is a colour outside RGB. If at all, only the blue-sensitive layer(s) of a color film stock will respond to it. No Red involved, no Green

 

Upon processing, a color negative stock produces the complementary colour to any given one. The complementary colour of Violet is . . . ?

 

It doesn’t exist. No subtractive color photographic system until today is made for Violet. Why? It’s not needed. We see Violet only faintly. It’s off the retina’s RGB sensorium, too.

 

A color positive film would have to have an extra layer of Violet dye in order to let Violet appear on the screen.

 

Things are simpler with black-and-white plates or film. These do react to Violet as far as the colour is transferred to them by the lens. Violet exists in nature, proven. Animals see Violet and Ultraviolet. The answer is thus: Yes, film can capture the colour Violet but that’s where it ends also. There has never been Violet on a cinema screen or a photo paper.

 

Violet is not purple, nor lilac, nor pink. Purple is the blend of Red and Blue, exactly half-way. Violet contains no Red.

[Chris Millar]

ok, I was talking out of my bum

 

So what was all the confusion about ?

 

 

I find conceptual roadblocks interesting - I'd like to develop a filter for them so I can be aware I've hit one before I go and say something dumb in a job interview for that job I really want

 

(job hunting in a new city soon)

[K Borowski]

Is this another sequel to "Dumb and Dumber?"

 

What they call "cyan plus magenta" is arbitrary. You can call it "violet" or you can call it "blue" or you can call it "purplish blue," or you can call it "light urple."

 

 

The friggin' color patches used to say "violet" instead of "blue" and film responds to wavelengths all the way into the ULTRAVIOLET region, so it's obvious that it is going to go past what the eye sees on the way to ultraviolet and register that too. Do you think Kodak went back and changed all of its films to SKIP OVER violet but still respond to ultraviolet, or do you think they arbitrarily changed the names on their color cards?

 

 

Since none of the names are scientific and accurate, why would we have two pages of silly debates on "whose violet is the right violet?" It's like using a pair of Goddamn calipers to measure distances on a bumpy golf pitch that someone just turfed in an all terrain vehicle.

 

 

Then, just for fun, let's translate it back and forth across German and English, debate whether the human eye can see "violet" (not relevant here) and introduce who knows how many other variations in different languages, people, and culture's different terms for different ranges or perceptions of color.

 

Let's argue whether violet can influence the human eye (again, not the subject here at all whatsoever) 's perception of blue. Since photographic blue IS violet, who cares?

 

 

 

So many people are misreading and going so far off topic that I'd say the signal-to-noise ratio on this thread is, conservatively 10:1 at this point. Can someone just post a pre-1974 Kodak color card that says "violet, green, red" instead of blue green read so this horse can die?

[K Borowski]

http://www.ebay.ca:80/itm/ws/eBayISAPI.dll?ViewItem&item=310325768130

 

It will cost you $8.50 CDN to find out whether or not this guy is my partner in crime as to my whole, unsubstantiated on Wikipedia, claims to something that happened before the invention of the World Wide Web. Kodak's changing the name on their color chart from "VIOLET" to "BLUE" must have originated outward, into space at the speed of light, so at this point, it's only been implemented in a small segment of our galaxy :-p

 

Now I have spent at least 45 minutes (45 minutes too long) on this silly thread.

 

 

 

 

Watch, someone will *still* come back and say I am wrong, just like I can have a saturated low contrast film and I should quit copy-pasting off of wikipedia, (not rolling ECN and ECP at work every day in between offering free advice to people that DON'T READ IT).

 

 

I suppose that may be the most frustrating thing someone can do, completely ignore something I freely contribute an immense amount of time explaining.

 

 

 

 

Anyway, one more tidbit of mental masturbation: You are never seeing ANY true blues, greens, violets, OR reds, you are seeing Kodak or FujiFilm or AgfaGaveart or Lucky or Ferrania or Technicolor proprietary cyan, yellow, and magenta dyes. (Contrary to popular belief, Fuji does NOT have a red-dyed cyan layer, it's a cyan interlayer that neutralized consumer fluorescent color shifts).

 

Probably the ONLY systems that are anywhere near close are dye transfer printing or IB technicolor prints. Doing it chemically with some arbitrary complementary colors (magenta and cyan taking the bulk of the density in a filter pack and yellow, being so thin, probably having the most spectral impurity) or arbitrarily using additive filters with their own impurities, by their very natures, produce a great deal of inaccuracy (or imprecission?) that none of us, except maybe a select few female readers, could ever see unless pointed out to us,

 

 

BTW, how many of us can REALLY say that it makes any difference whatsoever whether colors exactly match between a subject photographed and a subject projected, when the eye automatically adjusts for changes and color temperatures, and can easily be tricked into seeing "white" from a wide variety of temperatures of a black body equivalent of different heated Kelvin temperatures?

[Chris Millar]

Karl,

 

I've known you for a while here on this site - I've seen a progression toward (even more) anger lately - quite a profound anger that's not going to help you at all. And for the most part, probably not help the people it's directed at as they'll just switch off and disregard the information you're trying to express.

 

All this extra cortisol you're pumping in is only going to prematurely age you, or worse, get you a heart condition or something similarly nasty ... Realise that the internet has potential for death for those who choose to try to 'get it right', a process probably more futile than your bungy callipers.

 

You might be right (ouch, I can feel your anger at the conditional there already!) - but ... where did that get you ?

 

Take care bro !

[K Borowski]

It's frustrating seeing advice offered almost four mos. ago after a 16 hour day of picketing utterly and completely ignored, by the same people who accused me of "copy-pasting" off Wikipedia.

 

The writing isn't on the wall in this case, but it is right on the 1970s Kodak color patches.

 

Seeing semantics and philosophies applied on an arbitrary name to an otherwise very precisely defined photographic color is mind-boggling to me. How can so much debating go on so long and be just totally WRONG and off the mark?

 

 

 

I'd have been grateful to have been given the gift of having just one of my early color instructors in school telling me "Photographic Blue is a mixture of cyan &MAGENTA (violet)." Would have saved me many frustration sessions of learning that the hard way.

 

PHOTOGRAPHIC BLUE = VIOLET

 

 

There's no simpler way to state it. Pull up some Kodak Q-Cards on a color-calibrated monitor and you will see that they are NOT what we would consider blue, they are violet.

 

But, as with other things on the internet, it appears not to be about knowledge and application of that knowledge to our field, but about who can p*ss the farthest.

 

 

 

As for my anger at two pages, four months of ignorance, proceeding my first response, I choose to express it here, rather than to underlings and coworkers on the set and in the lab. I've had a lot of pleasant phone calls and internet conversation with people who are absolute a-holes to work with. Or maybe I should bottle anger at one person up and take it home to my family?

 

 

 

I feel I am giving anger exactly where it is due here. You choose instead to tell me to "take care" but it is clear you mean the exact opposite, further implying you'd never want to work with me! And that my anger is going to cost me work, and then kill me. Only if there are another two pages in late December is it going to kill me. . .

 

Maybe it would look better, be more diplomatic, if I didn't call out information that is simply erroneous supposition. Should I let it go another four months and contaminate the fall semester's film students?

[Matthew W. Phillips]

All this extra cortisol you're pumping in is only going to prematurely age you, or worse, get you a heart condition or something similarly nasty ...

 

Are you kidding? This stuff is the only thing that gets Karl through the day! I think its healthier for Karl if we let him vent. :)

[Chris Millar]

I feel I am giving anger exactly where it is due here. You choose instead to tell me to "take care" but it is clear you mean the exact opposite, further implying you'd never want to work with me! And that my anger is going to cost me work, and then kill me.

 

 

Karl, I didn't mean the exact opposite at all...

Far out man - I really do mean TAKE CARE

No sarcasm.

 

 

I don't think we'll ever work together (geographically) but since you mention it, yeh, why not ?

 

 

Maybe don't think you realise that I speak from a position of reflection at my own self, we're not to much different, sure you know a flip load more about this stuff than I, but then I could well get angry at you for a bunch of misconceptions and poop thinking about the things I know a lot about.

 

 

... one of those things is where the anger of the type you're displaying will get you

 

 

It seems like I'm just stoking the fire so ... s hit dude, take care

 

 

no hard feelings ?

Edited August 30, 2011 by Chris Millar

[K Borowski]

No hard feelings. Sorry I guess I did read you wrong. It's really tough to tell without emoticons sometimes, different dialects, English as a second language for many speakers.

[Mei Lewis]

I never realised that film might not record certain colors or that there'd be such an argument over whether it does or not. It got me thinking that maybe this is all related to a phenomenon I've noticed shooting certain sorts of lights with a digital still camera.

 

Violet colors can sometimes get really ugly when they blow out, and they seem to cause one channel (blue I think) to clip very much before the others and when otherwise the scene seems not very bright.

 

See the lights above and around the face here:

 

100% crop:

 

It only seems to happen with violet/purple colors. Other colors look like I'd expect, for example the orange/yellow lights here look fine:

 

 

 

 

This usually happens at concerts and in nightclubs so there's a possibility that it's the black lights they use but I don't think it's that because in most cases they're not having that kind of effect to my eye. They lights aren't making UV paint, teeth etc glow.

 

To my eye the lights just look like violet/purple lights.

 

Another thought I had was that the lights are unintentionally emitting UV which I can't see but which the camera sensor does.

 

 

I'd always assumed this was some artifact of how individual color channels clipped first, but maybe it's realtade to the current discussion. I'd love to hear your thoughts.

[K Borowski]

With digital, there is a different set of problems, as the sensors are natively sensitive to IR wavelengths (above red, rather than UV sensitivity).

 

Cross-contamination aside, digital and film definitely render different colors with varying degrees of success (pleasantness of reproduction). Digital tends to do a TERRRRRIBLE job with grene grass, especially when it is overexposed, in my opinion, but I actually think it does a better job with purples. There are other differences I can't think off of the top of my head. IDK why purple, mostly blue with some red would bring out what I'm assuming is IR contamination. I'd think, since there are twice as many green sites in a digital sensor, this channel would be most susceptible to IR contamination, followed perhaps by red as its wavelengths are closer to IR (kind of like UV shows up as blue haze on film). Any overexposure, though, is where I'd think it'd make it through the IR filter onto the chip.

 

 

My problems with "violet" are more that is a matter of out-dated color theories (painter's colors are red YELLOW and blue or the color wheel: Red Orange Yellow Green Blue Indigo Violet), so it's more a matter of English words and ancient, incorrect theories on color than a debate on what is happening scientifically. Talking about "violet" is just as meaningless as talking about "orange" being captured by film or digital. Scientifically, the colors you deal with that film and digital can be sensitive to are ultraviolet, blue, green, red, infrared.

[Mei Lewis]

...Scientifically, the colors you deal with that film and digital can be sensitive to are ultraviolet, blue, green, red, infrared.

 

How do you think it's best to specify colors scientifically?

 

I used to think you could do it by just stating their frequency, but now I think it's more complicated than that. Would you want to specify entire frequency distributions?

[John Sprung]

How do you think it's best to specify colors scientifically?

 

I used to think you could do it by just stating their frequency, but now I think it's more complicated than that. Would you want to specify entire frequency distributions?

 

It's customary to use wavelengths instead of frequencies, but yes, that's how it's done. See the earlier posts in this thread from Ben Syverson and Hal Smith, they have graphs of intensity or transmission versus wavelength. Rosco has transmission vs. wavelength graphs to tell you exactly what their gels do.

 

 

 

 

 

-- J.S.

[Phil Rhodes]

Of course almost all practical "colors" (as in real-world observations of objects or phenomena) involve a cluster of frequencies, so you'd have to graph it. Almost no light is truly monochromatic.