Contrast and its Relationship to Color

From TIGwiki2
Jump to: navigation, search

This is a discussion thread pulled from the TIG archives for March, 2007. Some of the discussion may be a little out of chronological order. Rob Lingelbach 15:00, 20 March 2007 (PDT)

From Rob Lingelbach March 14 2007

Something I think all experienced colorists know instinctively deserves a scientific explanation, if someone here would like to attempt one. Take a color image, make it look the way you think it should, and then extract all the color, making it pure black and white. Unless you're specifically looking for low contrast, you will now need to dial in significant amounts of contrast (drop the blacks, raise the whites). (the only other exception I can think of to this formula is if the original image was extraordinarily high in contrast when it was in color.)

We know that detail is a function of luminance, not color, but I'm interested in a more complete explanation of this phenomenon.

And something related occurs when originally-shot black and white is colorized without regard to original contrast, such as with the Turner library as shown on that cable channel of colorized films. The balance of shadows and highlights goes all wacky.

Rob -- Rob Lingelbach http://www.colorist.org/robhome.html


From Steve Hullfish, March 17, 2007

I'd have to do some more research, but I think there's already a pretty good description of what happens in this instance. I've seen it discussed in some of the Photoshop color correction books. Retouchers have an actual formula for making a color image into a black and white image while retaining the proper contrast.

I remember even hearing someone discuss how to do this correctly on an Avid, using the basic color correction capabilities it has.

In 4:2:2 video, it may have something to do with the color sampling being different. (Y Cb Cr).


Steve Hullfish Verascope Pictures (630) 717 9069 office (312) 399-4237 cell www.veralith.com


From Tom Nottingham March 17, 2007

I am sure most good engineers could give you a technical explanation of this phenomena, and for someone on the level of Dave Tosh (whom I really admire and respect), it shouldn't really be difficult to articulate the reasons. My experience comes from over 25 years in the chair, and also a dose of several years on the other side in a film laboratory. Since my background is a technical one of film and television, and not one of engineering, please forgive me if I gloss over the electronic theory, and dwell on the metaphysical.

The characteristic you observe is the primary reason I dislike the practice of shooting on color stock when finish is intended to be black and white. Quite simply, the gamma response of color stock is not even close to that of Panchromatic Black and White stock. Of course unless you know what to look for, or happen to be able to see the results projected side-by-side, the casual observer might not even notice the difference in gamma when contrast is restored in Telecine. The experienced operator understands that true black and white images neither posses true black or white, but consists only of grey scale. A classic film such as Casablanca is a good example. During most of the film, blacks rarely go below 3 - 7 ire (or above 10 ire in NTSC), and the whites rarely clip except for specular highlights. Black and White transferred any other way just does not look natural. Properly exposed and processed b&w stock has a gamma curve of somewhere around .3 to .4. Gamma is defined as the slope of the straight-line portion of the D log E curve when measured on a densitometer.

A slope of 1.0 is a curve that produces equal increments of density for equal increments of exposure. Contrast index is the difference between the toe and shoulder of the straight line portion. D min is the clearest (no density) frequently referred to as base + fog, and D max is the densest. A properly exposed and processed black and white film rendering of a normal contrast scene has a contrast index of about 1.8. Black and white stocks are frequently referred to as having an extended shoulder which means they can be greatly overexposed and still yield usable images, although the contrast index is greatly affected. It was this very phenomena that Ansel Adams exploited in his development of the Zone System.

Color stock is another animal. Composite color stocks are in reality three separate layers of black and white silver images that are exposed through three separation filters built into the emulsion. These three layers represent the Red, Green and Blue records of the original scene. During processing, the silver is bleached and removed, leaving three layers of dye representing the complement of the original RGB. That would, of course, be yellow, magenta and cyan dye. Since there is no "black printer", contrast and luminance are derived by a combination of the three. Positive prints (for projection and made from OCN) can have a contrast index as high as 2.8 to 3.2. Three-strip Technicolor or Kodachrome can have contrast indexes of 3.8 to 4.1. Gamma frequently falls between .6 and .8 - still less than 1.0, but considerably higher than black and white stocks.

Chromatic values consist of three components: Saturation, Hue, and Luminance. All three interact to make up a color stock's contrast, with colors like Red and Yellow normally having high luminance values, and others such as Blue and Cyan normally having much lower luminance contents. This can be seen when RGB outputs from Telecine are viewed separately on a black and white monitor. The Blue record normally has the lowest contrast, and the Red record has the highest. Green is considered "normal" which is why the original Rank Encoders derived luminance from the Green channel. This phenomena can be see in the Lab as well as in various TV systems - it is the nature of color as it occurs in the environment. Of course, there are always exceptions. Blue can have a high luminance content in night-club scenes with strong color lighting, but the in the normal scene, the Blue record has fairly low gamma.

In Telecine, a normal scene is set up with blacks at zero, and whites well into clip. This is especially true of commercials because clients never think there is too much contrast or chroma. Heavy chroma content adds considerably to the luminance content of the various colors, but does not add to the basic "black and white" value of the scene. Where it is true that modern-day color correctors such as the DaVinci 2K derive luminance using a complex formula, and also allow the operator to control it separately, when chroma is eliminated, the four channels become one of the same value. Any luminance added to the original scene due to color saturation or hue is now removed - leaving nothing but the luminance scale. Since contrast is the difference between the lightest and darkest values in the scene, removing chroma reduces contrast. Bright colors, such as Red become darker grey values, and dark colors, such as Blue become brighter. This effect can be seen using a switcher to dissolve between 75% color bars, and its mono component.

When transferring black and white films for colorization, care must be taken by the operator to build sufficient contrast into the luminance values because of the basic differences between color and black and white stocks. Also, a color added later cannot remember the "chromatic" luminance values from the original scene, so this must be taken into account in order for the final result to look like it was shot on color stock. By not adding contrast in Telecine, the same scene when viewed later with chroma added becomes very dark and muddy. A lot of the black & white classics from the 30s and 40s were shot on Orthochromatic stocks - which were only blue sensitive. Even though the studios frequently over developed these stocks to increase contrast and increase film speed, it did not entirely overcome the flat nature of Orthochromatic films. They also did not represent the higher luminance values of warm colors.

The effect of Chroma and Hue on a scene's contrast is really powerful, and is automatically taken into consideration by a good colorist whether he is aware of it or not. After all, luminance is only one third of the picture.

Tom Nottingham Retired Colorist


From Richard Kirk March 17, 2007

Tom has explained in some detail why black and white film should have a different appearance to colour film. There is also the Callier effect: the silver grains diffract light where soft-edged dye clouds - which will give you a steeper tone curve than your densitometer may suggest, because the scattered light may be lost in a projector but not in a densitometer.

However, if I read Rob's post correctly, I think he is asking a slightly different question: when we take out the colour from a picture, why do we want to change the tone curve to be more like black and white film? Are we hard-writed to prefer more contrast, and this has shaped the development of black and white film, or has black and white film modified the way we see things?

When I was young, and watched black-and-white 405-line TV, I used to be aware that there were two sorts of signal - two different tone curves, though I doubt I would have called it that at the time. You noticed the switch between the two most when the adverts came on. The better of the two was what you got when they showed black and white films, and the more expensive adverts. it felt more like watching the TV with the lights out - something we as children were discouraged from doing. The duller tone curve was for live TV and the cheaper adverts. I am guessing that one was done on film, and the other may have tapes or done live. While my perceptions may have been shaped by black and white films, because we did go to the cinema on Saturdays and only the main feature and the Pearl & Dean adverts, and the cartoon if you got a good one were in colour - the Pathe News, the Look at Life and the short feature would be in black an white. If I was being conditioned, I guess it was to think of black and white as the 'cheap' option. Nevertheless, I did seem to want to have the black and white tone depth.

How do we see black and white images? We could imagine we are looking at a planet lit with uniform white light, and filled with colourless objects and people. In fact, we forget about the black and whiteness and just accept the strange stimulus as ordinary vision because the motion, the edge features, the light and shadows are all familiar. The brain is good at re-mapping stimuli and responses in this way. You stare at the field of dots and suddenly you can see the dalmation in the leaves. There was a documentary on the early days of the BBC, where they described a very early outdoor broadcast system experiment. The people in the van were testing the camera under low light conditions, when a policeman came over and asked what they were up to. They re-wound the tape and showed him the moving image - "here is the street lamp, here's that tree, and here you are coming out from behind it...". The policeman looking at luminous blobs on a flat screen could not relate it to the scene outside at all, and then all of a sudden his brain did something and he could.

With analogue black and white TV, it is difficult to change the gamma without lousing up the colour contrast. However, when digital techniques became available, people like me tried increasing the luminance gamma of still images while preserving the chroma. The result is strange, and somehow unsatisfactory. Why? I guess a lot of this may be because we are mentally 'wired' for colour constancy. We tend to think of 'color' as a property of the object, and remove the effects of the illumination. Perhaps, with a black and white image (or even with a duotone) we can shut off our color constancy processing, which leaves us freer to muck around with the contrast. And if you give someone a contrast control or a sharpness control, they will tend to turn it up until they are really damaging the image, because more signal always feels 'better'.

I think we may be taught how to see by our media. We know our sense of a 'pleasing white' varies from daylight at daylight illuminations, to yellow incandescent at low levels - and literature has many examples of people preferring candle-light to incandescents, though I miss my blues too much. Nevertheless, I suggest that our preference for high-contrast images might predate black and white photography. Evidence for this might be found by analysing the effective tone curves of etchings and line drawings in the history of art. I think Charles Poynton may be going to do just that, so maybe sometime soon we shall have a proper answer, instead of just me, guessing.

Cheers. Richard Kirk -- FilmLight Ltd. Tel: +44-(0)20-7292-0400 or -0409-224 (direct) Artists House, Fax: +44-(0)20-7292-0401 14-15 Manette Street London W1D 4AP, UK


From Robert Lovejoy March 18, 2007

Tom is of course correct that black and white film yields a different look than desaturated color stock. However, I have no problem if clients bring color stock and want a black and white look. One approach worth consideration is tweaking the telecine neg match (I remember PEC's!) to emulate panchromatic or orthochromatic stock (see http://www.photo.net/learn/optics/edscott/pss00030.htm ) by varying TK RGB before summing to mono. Instead of first making a balanced color picture before dropping chrominance, play up the spectral differences available in color stock. Using the green channel only is interesting, but when you mix in varying amounts of red and/or blue you get different results. Of course contrast is still going to need (instinctive) adjustment, but the beauty of the suite is the ability to experiment in real time and derive images that please your client. While nothing looks exactly like black and white film, you can certainly get interesting results from color stock by emulating and tweaking panchromatic and orthochromatic response.

Bob Lovejoy Shooters Post & Transfer Philadelphia, PA


From Rob Lingelbach March 18, 2007

On Mar 18, 2007, at 1:42 AM, S. T. Nottingham III wrote:

> I am sure most good engineers could give you a technical > explanation of this > phenomena, and for someone on the level of Dave Tosh (whom I really > admire > and respect),

hear, hear.

I agree with everything you wrote Tom and appreciate the reply. I had the opportunity not long ago to test, for a commercial that would air in b+w, shooting on various color stocks and shooting on b+w stock. Needless to say the b+w stock won, for the reasons you mentioned.

Forgetting film for a moment, and just talking about an 'image', the question still remains for me: why is it we must add contrast to an image that is originally in color, when we pull out the color? I'm in search of a scientific explanation; I've known about the phenomenon since the 1970s, when I first sat in the chair.

Your dissertation on luminance and chroma deserves a place on the wiki, I'll wait for this thread to spin out and combine it with other possible responses in the technical discussions section.

Rob

-- Rob Lingelbach http://www.colorist.org/robhome.html


From Greg Dildine, March 19, 2007

My favorite thing about b&w neg stock was the softer emulsion (or some processing quirk) making a mess of the gate and rollers on the Ursa. -Greg Dildine


From Bob Festa, March 19, 2007

Black and white stocks have no "Rem Jet" backing layer. Thats why B+W = gets such great halation splay around whites

The early Cintel machines were like a blender with this stuff. I can = remember having to vaccuum the deck plate after each reel of film.

See, there are parts of this business I remember fondly :)

________________________________________________________ Bob Festa bfesta@ascentmedia.com Director of Commercial Imaging 310 434 6000 R!OT Santa Monica


From Scott Klein March 19, 2007

Especially nice in 16 MM. We used to thread the film on backwards & then flip it back electronically to fight this shedding effect. Invariably the director would then ask for flex effects which were impossible when running the machine this way so you'd have to flip it back and endure the snowstorm.


From Jim Mann March 19, 2007

Hi All,

Filpping the 16mm B&W Neg was a good trick that worked well, but sadly single pe rf and super 16 put a end to that. I do not believe you can order PlusX or DoubleX in double perf anymore. Too bad on one made the reversable tacho wheel! Real B&W neg jobs are few and far between these days. Just too much hassle to get processing. IMHO


Regards. Jim http://tig.colorist.org/wiki3/index.php/User:Jim_Mann

Jim Mann Freelance Colorist http://tig.colorist.org/wiki3/index.php/User:Jim_Mann mobile-516-250-0909


From Michael Bittle March 20, 2007

<audio guy comments on color science - hide the women and children!>


Could it be that what we perceive as contrast is not purely function of black and white?

If I want to enhance the sound of a bass or kick drum, I can sometimes get good effect by playing around with 200hz, this sharpens the attack rather than the fundamental but the ear/brain extrapolates to give the sensation of 'more bass' or 'better bass definition'. Sort of like getting better 'black' by adjusting 'blue' (don't know if that works, just using a metaphor)? So maybe, just maybe, there is 'contrast' information from the eye/brain perspective, contained in color difference data as well.

Just wondering... Mike



From Joe Owens March 21, 2007

The other reason (for wanting to readjust contrast) might be (gamma response aside), that the dynamic contrast between hues of similar luminance is lost. Anecdotally, some publications in the print industry suffer tremendously coping with this. An excellent example can be found in the "Air Combat" types of magazines. Obviously they don't attract a lot of commercial interest and are forced to print most of their product in black-and-white. However, most of the beauty shots of the aircraft are submitted in colour. And most of the aircraft are in "low-visibility" camouflage, usually framed against a lovely blue sky -- beautiful if you admire these things. When printed in black and white of course, the camo does its job and the aircraft winks out of sight, no need to come diving out of the sun-- its gone.

It is disturbing how casually most black-and-white from colour is achieved by simply turning down the saturation dial. That is just wrong, wrong, wrong on almost every level.


Joe Owens Presto!Digital Colourgrade 302-9664 106 Avenue Edmonton, Alberta T5H0N4 +1 780 421-9980 jpo@prestodigital.ca


From Rob Lingelbach March 20, 2007

On Mar 19, 2007, at 11:55 AM, Michael Bittle wrote:

> > Could it be that what we perceive as contrast is not purely > function of black and white?

yes, I think you're beginning to describe what I was at a loss to explain, that color adds something that makes up for a lack of contrast, like your example that a 200Hz emphasis will improve the perception of bass. Description of this by example may be our only recourse, as the technical/scientific explanation is elusive.

Joe Owens wrote:

> It is disturbing how casually most black-and-white from colour is  

achieved by simply

> turning down the saturation dial.

and this depends on what you mean by disturbing, but I agree. There are some other factors of which Richard Kirk's post hinted, for example the interesting halation produced by vidicon cameras, which is very difficult to replicate with modern equipment; the interesting grayscale (with low contrast!) of kinescope images. As a child I too was prevented from watching TV with the room lights down; the patriarch said "it will ruin your eyes!" of course he also thought reading in bed would ruin my eyes, when actually it just ruined my sleep.

Instead of removing color from an image and leaving a low-contrast black and white image, I think for a moment of adding color to a medium-or-high-contrast black and white image, which is done with toning and tinting. Toning and tinting, done to photographs, (forgetting colorization of moving picture film) can be pleasing to the eye, when it is done with low or very low saturation. There is a connection between color and contrast that seems to be in some kind of inverse ratio, depending somewhat on the artistic taste of the viewer.

--Rob

-- Rob Lingelbach http://www.colorist.org/robhome.html


From Joe Owens, March 20 2007

Yes, and colour and texture are also interchangeable. Consider a smooth-plastered surface vs a stippled one.


Joe Owens Presto!Digital Colourgrade 302-9664 106 Avenue Edmonton, Alberta T5H0N4 +1 780 421-9980 jpo@prestodigital.ca


From Michael Bittle March 20, 2007

absolute values (luma ~ x IRE, chroma ~ y IRE) but rather some sort of relativistic calculation;

contrast = {red with respect to green} transformed by {green with respect to blue} transformed by {blue with respect to red}

 (I'll leave the real equations to real mathematicians)

Mike


From Tom Nottingham March 20, 2007


I believe the description (by others below) is exactly what I have been saying:

Chromatic values consist of three components: Saturation, Hue, and Luminance. All three interact to make up a color stock's contrast, with colors like Red and Yellow normally having high luminance values, and others such as Blue and Cyan normally having much lower luminance contents. This can be seen when RGB outputs from Telecine are viewed separately on a black and white monitor. The Blue record normally has the lowest contrast, and the Red record has the highest. Green is considered "normal" which is why the original Rank Encoders derived luminance from the Green channel. This phenomena can be see in the Lab as well as in various TV systems - it is the nature of color as it occurs in the environment. Of course, there are always exceptions. Blue can have a high luminance content in night-club scenes with strong color lighting, but the in the normal scene, the Blue record has fairly low gamma.

Tom Nottingham


From Joe Owens March 20, 2007

Quote: Green is considered "normal" which is why the original Rank Encoders derived luminance from the Green channel.

I always thought it had more to do with the old Y derivation for composite Y,I,Q which was (roughly) 70/20/10 G,R,B or something like that. Which took advantage of flesh tone being roughly on the I/-I vector and reducing bandwidth on the Q/-Q vector (magenta/green) while building resolution back up in that area because Y is mostly green.

Joe Owens Presto!Digital Colourgrade 302-9664 106 Avenue Edmonton, Alberta T5H0N4 +1 780 421-9980 jpo@prestodigital.ca


From Rob Lingelbach

On Mar 19, 2007, at 12:06 PM, Joe Owens wrote:

> aircraft are in "low-visibility" camouflage, usually framed against > a lovely blue sky -- beautiful if you admire these things. When > printed in black and white of course, the camo does its job and the > aircraft winks out of sight, no need to come diving out of the > sun-- its gone. > > It is disturbing how casually most black-and-white from colour is > achieved by simply turning down the saturation dial. That is just > wrong, wrong, wrong on almost every level.

Does there exist some kind of simple optical device (a form of eyeglasses) that would desaturate our vision and make it black and white? It would be the opposite of looking at the world through rose-colored glasses. But it would be really interesting, and if I'm not mistaken, the brain might eventually compensate by adding in the color? So that when the glasses are removed after a few days the wearer now sees everything with double the saturation??? I'm not sure about that.. but I do know that if inverting glasses are worn, so that everything looks upside-down, and they are worn for a few days, the brain of the user compensates and reverts the image to right-side- up. So then she takes the glasses off, and the brain takes another few days to compensate, which in the meantime creates a human with upside-down-vision, no glasses required.

The film I saw explaining this, shown in a junior-highschool science class, showed a POV of the guy with inverted vision riding a motorcycle in traffic. Impressive feat.

-- Rob Lingelbach http://www.colorist.org/robhome.html


From Joe Owens, March 20 2007

In one of those "extra materials" DVDs from the Golden Era, there was a feature done on Edith Head, who is often pictured wearing tinted glasses --pink, I think, or maybe yellow.... hmmm.... It was not strictly to be chic, as she maintained that it helped her visualize how the costumes would look in black&white. And then of course, she was a master of colour and her work leaped off the screen in Technicolor. And it can be said with some degree of certainty that there are some physical differences between males and females. I am quite happy about some of them... but the word on the street is that females have a richer retinal rod/cone population, and for that reason, they see more colours than males -- or at least perceive a more subtle palette, and can detect differences that males cannot. This probably has some individual variation and it is possible that some richly endowed males can see better colour than a deprived female. Just like some guys could use "a little support" up top....  ;)

Joe Owens Presto!Digital Colourgrade 302-9664 106 Avenue Edmonton, Alberta T5H0N4 +1 780 421-9980 jpo@prestodigital.ca


From Jeremy Pollard March 20, 2007

> Does there exist some kind of simple optical device (a form of
> eyeglasses) that would desaturate our vision and make it
> black and white?

There are various accounts from 'achromats', people that are unable to perceive colour at all, that may inform how the world appears in "black and white". Oliver Sacks describes a

case where an 

artist loses his colour vision following head trauma (to his occipital lobe) in his book, 'An Anthropologist on Mars'. The story goes on to describe the artist's perception of colour and surmise how the brain interprets chromatic and tonal information.

"Everything appears to me as [if I were] viewing a black-and-white television sc reen," reports an achromat in Oliver Sacks' book, An Anthropologist on Mars. "My brown dog is dark

gray. Tomato juice 

is black. Color TV is a hodge-podge."

Another personal account written by a complete achromat may be found here: http://consc.net/misc/achromat.html

Interestingly, the author notes (about halfway through the article) that, "Colou red and B & W pictures are usually indistinguishable to me."

Finally, the following article is an interesting (and fairly brief) discussion a bout colour interpretation and its role in motion detection:

http://hno.harvard.edu/gazette/1998/08.06/BrainsColorProc.html


-- Jeremy Pollard VP Sales --- jeremy.pollard@risingsunresearch.com --- Rising Sun Research p. +61 8 8400 6494 - f. +61 8 8400 6401 http://cinespace.risingsunresearch.com



From Rob Lingelbach, March 20, 2007

On Mon, March 19, 2007 10:41 pm, Jeremy Pollard wrote: > > Another personal account written by a complete achromat may be found here: > http://consc.net/misc/achromat.html

Here is a telling paragraph, written by the achromat's extremely detailed biography as Jeremy notes above:

"Coloured and B & W pictures are usually indistinguishable to me. But sometimes I can, often quite easily, tell coloured and noncoloured pictures apart. Coloured pictures may look less crisp, or slightly less in focus, and often have less contrast than comparable monochrom pictures."

Proof that images originally in color suffer from lack of contrast when desaturated.

((Another line from the biography: "I have one sister (born 8.10.1943) and one brother (born 5.10.1945) who are both, like myself, typical, complete achromats, confirmed by the anomaloscope." How many anomalies the anomaloscope can check today might be an interesting question.))

-- Rob Lingelbach http://www.colorist.org/robhome.html


From Jack Kelly, March 20, 2007


Tim Sassoon wrote:

> color lives in the midtones, but in the > absence of color, we want more dramatic > tonal separation.

Apologies for hijacking this thread and asking what is probably quite a beginners... But this question has been nagging me for ages and this seems like a good time to ask!

I've often wondered why tonal manipulations can affect colour saturation. For example, take a picture of a sunset... subtle tweaks to the RGB curve can have quite a profound effect on the saturation of the sky. For example, I often find myself using curves to try to make the clouds a little more contrasty and this can result in the sky changing from a light, watery cyan to a really saturated cyan. Why is this? I guess that saturation and contrast are linked in as much as desaturation and reduction in contrast both result in making the image more gray but I can't quite wrap my head around the relationship between tonality and saturation - any hints would be most welcome!

Many thanks, Jack Dir/Prod/Post London


From Rob Lingelbach March 20, 2007


On Tue, March 20, 2007 9:39 am, Jack Kelly wrote:

> Apologies for hijacking this thread and asking what is probably quite a > beginners... But this question has been nagging me for ages and this seems > like a good time to ask!

"no question is ever too basic or simple" -- a TIG creed. The group exists to help answer any questions, particularly those that may seem "simple", as those can be hard to find answers to.

> I've often wondered why tonal manipulations can affect colour saturation. > For example, take a picture of a sunset... subtle tweaks to the RGB curve > can have quite a profound effect on the saturation of the sky. For example, > I often find myself using curves to try to make the clouds a little more > contrasty and this can result in the sky changing from a light, watery cyan > to a really saturated cyan. Why is this? I guess that saturation and > contrast are linked in as much as desaturation and reduction in contrast > both result in making the image more gray but I can't quite wrap my head > around the relationship between tonality and saturation - any hints would be > most welcome!

yes, you're relating the other side of the effect that was under discussion, that if a color scene suffers low contrast when the chroma is subtracted, then if contrast (in your case RGB curve) is increased while there is chroma in the scene, the chroma will have to be reduced. Experienced colorists know this instinctively (hence the original thread title - "instinctive contrast") and have a hand on th e chroma knob (an argument for using a dedicated panel- not a mouse or qwerty keyboard) for an instant adjustment to the color when the contrast (including gamma, black level, white level) is increased or reduced. It avoids the client saying, when you adjust the contrast, "but now the color's changed." ....which can open up a can of worms; it's much nicer to "couple" the color and contrast changes by using two hands.

Now that doesn't really answer your question, it only verifies the condition, an d I think the answer can be found in the recent posts about saturation and contrast, if a more scientific explanation is what you need.

regards Rob -- Rob Lingelbach http://www.colorist.org/robhome.html TIG admin


From Tim Sassoon March 20, 2007

In a message dated 3/20/07 5:40:40 AM, jack-cml@focuspointfilms.co.uk writes:

> I've often wondered why tonal manipulations can affect colour saturation. >

Saturation results from differences between color channels. If all channels are the same, you have a shade of gray. At the other extreme, if one channel is full on, and the others off, then you have a primary color at full brightness and saturation. If one is full on, and the others half on, then you have a primary color at full brightness and half saturation.

When you, for instance compress levels, you're also increasing the contrast between channels, and thus increasing saturation.


Tim Sassoon SFD vfx & creative post Santa Monica, CA


From Rob Lingelbach March 20, 2007

> >> I've often wondered why tonal manipulations can affect colour >> saturation.

I'd like to know what you mean Jack by 'tonal manipulation' - does it mean manipulations of luminance, or of some component of color?

On Mar 20, 2007, at 2:36 PM, TSassoon@aol.com wrote:

> Saturation results from differences between color channels. If all > channels > are the same, you have a shade of gray. At the other extreme, if > one channel is > full on, and the others off, then you have a primary color at full > brightness > and saturation. If one is full on, and the others half on, then you > have a > primary color at full brightness and half saturation. > > When you, for instance compress levels, you're also increasing the > contrast > between channels, and thus increasing saturation.

on reflection, this makes a lot of sense and accounts I believe for the contrast phenomenon we've been discussing *in the electronic realm* (video). I don't know however if you can apply this explanation to non-electronic contrast differences when you extract color from an image observed for example by the achromat, where the effect takes place entirely in the brain/eye system.

I'm wondering if there's a way one can remove "only color" from a painting (subtractive color as opposed to additive color in video) and see that the resulting grayscale image suffers a lack of contrast? Because if the achromat sees a lack of contrast when looking at color images due to his lack of color sense, then we are not referring to 'channels' any longer, correct?

Rob -- Rob Lingelbach http://www.colorist.org/robhome.html


From Arnell Patscott March 19 2007

Along the lines of this thread, I wanted to point out that in the March issue of

Scientific American there is a paper on Illusory Color and the Brain. They make
the point that images that start out as color, and are turned B&W loose quite a
bit of the image, more than just color. They also state that color is not separ

ate from shape recognition, but very much a part of it.

The article goes from the rods and cones to the brain centers responsible for in terpreting them. There is a small preview of the article at http://www.sciam.com/article.cfm?chanID=sa006&colID=1&articleID=CBDAEF3D-E7F2-99 DF-3BA6019D1B7D1E08

but to see the whole thing, including diagrams of false color and brightness, yo u either have to buy the online version, or get the March 07 edition. I highly recommend it as ground breaking work on human color perception.

Arnell Patscott

(No I do not work for Scientific American)


From Jim H. March 19, 2007

On Mar 19, 2007, at 2:29 PM, Rob Lingelbach wrote: > > This isn't the explanation I was looking for regarding the need for > extra > contrast after subtracting color from an image...

The vision explanation for this effect is that there are three types of cones involved in color vision and they combine in different ways to produce an internal luminance channel, plus a red-green, and a blue-yellow opponent color channels. The perception of contrast is the combined signal from the luminance channel *and* the red-green and blue-yellow channels. In practice, the blue-yellow has so little influence that it is ignored. (think about how either blue or yellow text of equal luminance on a white background are equally hard to read.)

So contrast mostly comes from the perception of the luminance and the red-green channel. If you remove the red-green channel information, then you are only left with a portion (say about 70%) of the contrast information the brain depends on, thus the image looks like it has less contrast. Boosting the gradient of the luminance of the image by 30% or so restores much of the appearance of the image under the same viewing conditions.

These are human visual system effects that are independent of whether it is on film projected in a dark room, or on a bright screen.

Jim H. Starwatcher Digital


From Rob Lingelbach March 20, 2007


On Mon, March 19, 2007 10:44 pm, Arnell Patscott wrote: > Along the lines of this thread, I wanted to point out that in the March issue > of Scientific American there is a paper on Illusory Color and the Brain. They > make the point that images that start out as color, and are turned B&W loose > .....

Going to have to try to find that Arnold. That's one magazine almost always worth buying (if you're not in Brasil, where it costs about 25 dollars).

There have been a number of studies done on the artistic differences between fil m and SD/HD video, where among the more thoughtful, and evocative cinematographers , there was a consensus that film's "detachment" from the "reality of the image" contributes so much to our sense of art, that film will always be these artists' first choice as a recording medium. (caveat: the day may be arriving where the best video combined with the best post-production will create a similar evocative removal from reality).

I think the same principle can be applied to black and white (grayscale) images versus color images: since we (other than achromats-- which I will now read up on, thank you Jeremy) normally see in color, black and white is a further removal from reality, that can contribute to a sense of suggestion, expression, poignancy.

Rob

rob@calarts.edu TIG admin http://www.colorist.org/robhome.html



From Aaron Shaw, March 20, 2007

This is a subject that fascinates me to no end. Might I suggest the following website?

http://www.handprint.com/HP/WCL/wcolor.html

I don't know the authors credentials but it meshes perfectly with my own reading and discussions with my father who has a PHD in optics. Definitely worth the read though it is exceptionally long.

Aaron Shaw Weet Media Montana


From Richard Kirk, March 20, 2007


Hi.

A bit more on black & white .vs. color...

We can see color differences even when the two objects we are comparing are separated by a large angle. We are unable to do the same with luminance information. If we plot our sensitivity to frequencies of chroma and luminance contrast, then the luminance sensitivity peaks at about 10 cycles ber degree, and falls off slowly for higher angles.

I used to work with printed images, which typically had a contrast range of 100:1. It was necessary to flatten out the contrast range if you were to represent a typical transparency on print. The simplest non-spatial way of doing this is to compress the luminance in some approximately perceptual space. However, this often made the images lack edge contrast, so, I attempted to filter the images, removing the long-period luminance contrasts where possible to preserve the short-range contrasts. This is very similar to the high contrast composite images people produce today. Given the right subject, you get an image with something with what looks like an unnaturally high tonal range on a color reflection picture. This, in many ways, is what I had expected my original approach of enlarging the luminance contrast to look like. However, on unsuitable subjects, like black and white text against a grey background, this can look absurd.

We prefer pictures with a boosted contrast, but if we simply boost the luminance contrast on a color image, then we are not getting the right color and luminance contrasts at high and low spatial frequencies. If, however, we drop the boost on the luminance low frequencies, then things seem to work again. This suggests that there is something in the low spatial frequencies of the luminance component of color contrast detection that limits how much tone boost we can accept.

As for Rob's suggestion of reducing saturation - there is an easy way of doing this. We can reduce the luminance. Not all the way to scotopic vision (though that is a solution of a different sort) but so the Hunt effect limits our perception of colorfulness as the light levels go down and our color-difference signals start to get swamped by noise. Some of you may not have met 'colorfulness' - it isn't an easy concept, so here's an example. I used to make 8mm stop-motion animations as a child. I used to project these on a silver screen in the sitting-room when I could get it. When this was in use, I used a pad of paper on the floor of my bedroom as the screen. I noticed that the image was apparently much more colorful and somehow 'crisp' looking than on the large screen. Yet, as you move the projector towards the screen and the whites get brighter, our senses tell us that all the individual colors are staying the same. So, we need separate numbers for the saturation with respect to white (relative colorfulness) and the amount of color we are seeing (absolute colorfulness).

You can do experiments to estimate how colorfulness varies with light level by trying to match saturation in your two eyes when presented with a target with different light levels - you will want a lot more saturating in the dimmer image to get a match. Okay - this is not really a realistic viewing model, but colorfulness is a hard thing to measure.

PS: This is also a good argument for not making digital projectors as saturated as you can. If you filter the light really hard to get saturated primaries, you will lose brightness, and your sensation of

You can do experiments to estimate how colorfulness varies with light level by trying to match saturation in your two eyes when presented with a target with different light levels - you will want a lot more saturating in the dimmer image to get a match. Okay - this is not really a realistic viewing model, but colorfulness is a hard thing to measure.

PS: This is also a good argument for not making digital projectors as saturated as you can. If you filter the light really hard to get saturated primaries, you will lose brightness, and your sensation of colorfulness will actually go down. More light can be as good as more color, particularly down at 50 nits and below.

Cheers. Richard Kirk

-- FilmLight Ltd. Tel: +44 (0)20 7292 0400 or 0409 224 (direct) Artists House, Fax: +44 (0)20 7292 0401 14-15 Manette Street London W1D 4AP


--Rob Lingelbach 08:10, 23 March 2007 (PDT)