cgardner
Offline
Dedicated FM
Upload & Sell: Off
|
 p.1 #16 · highlighting in layers question?? | |
Look at the left side of the color picker where there is no color. Now look down. What happens? The tone value gets darker (more gray) until it becomes black. Now look in the upper right corner where the saturation and brightness are both 100%. Now look down. What happens? The value (gray component) increases, just as it would if you started with a can of red ink and added black ink to it to tint it darker.
The horizontal axis saturation. The vertical axis is value (grayness / brightness). All the colors in the middle of the color picker are combinations of less than 100% saturation and less than 100% value.
Colorspace is described in three dimensions as in the wireframe below:
The vertical scale, like the earth's axis, is the value or grayness component just as shown in the color picker: the north pole is white, the south pole black. Using the analogy of a globe the most saturated colors are on the surface. As you move closer to the core the colors become less saturated. In the 3D wireframe above in fact that the most saturated hues of SWOP CYMK inks are less saturated than the outer boundaries of saturation in the AbobeRGB editing space. The inks are real measurable entities, Adobe RGB is an arbitrary "working space" boundary for editing color, which was based in part on the need for an editing space large enough to fit the different shapes of RGB monitor and CYMK offset inks used in commercial reproduction at the time.
In the mid-1970s I worked at National Geographic creating color separations for the relief on the NGS maps and illustrations and from there I worked in web offset magazine printing as production manager during the timeframe when SWOP standards for proofing first came into use. The company were I worked was on the committee which developed the standards. I learned about 3D color space mapping during the same timeframe in the 1970s from a the color scientist at DuPont who had invented its Cromalin pre-press proofing system and I taught printing technology for five years at the college level. So do know what I'm talking about here. 
Apart from 100% primary RGB and secondary CYMK colors all colors are overlapping combinations, either literally in the case of printing or perception of a RGB matrix pattern in the case if a monitor. If we add equal parts of RGB lights we get neutral tone (gray balance). If for example the red channel gets a bit brighter than the G and B, say 160, 120, 120 what we actually have is a 120,120, 120 gray value with a +40 red bias. If we have 250,120,10 there is a still 10,10,10 gray component in the color.
WIth RGB color the dominant color carries very little detail. Look at channels in a bright red object and you'll see the red channel has relatively little detail. The detail, expressed by darker values of that hue , are created by the other two colors, green and blue. Combined they act to add a gray component to the bright red carried in the red channel. To the extent there is more green or blue in the darker detail the darker hues will shift from neutral towards yellow (R+G) or violet (R+B). There is no black in RGB to decrease the value (darkness) of a hue -- the gray component comes from changing the non-dominant colors in the triad.
In a highly saturated secondary color, like a bright yellow daffodil both the R and the G channels will be maxed out to create the saturated yellow hue leaving only one subordinate channel, the blue, to add the graying component which will change the value of the yellow darker. That's why its difficult to capture the same detail seen by eye in yellow flowers in a digital capture.
CYMK printing is a bit more difficult to wrap the head around. First off, the inks are transparent. They must be because the illuminant must pass through the overlapping CYM dots, bounce off the white paper, back through the CYM ink to meet the eye. Don't think so? Try printing on black paper and see how your color looks
Secondly 100% C+Y+M doesn't produce neutral gray. Set your CYMK preferences to SWOP v2 and then enter R-128, G=128, B=128 into the color picker then look at the corresponding CYMK % C=45% M=37% Y=38% K=2%. More cyan than magenta or yellow are needed to achieve a neutral gray because the ink pigments are cross-contaminated. Change the color preferences to GRACOL and the CYMK % in the color picker will change to C=44% M=35% Y=35% K=8% because that standard set of inks is formulated differently.
As with RGB the lowest common % in a CYM combination represents the baseline gray component in the color. Years ago a clever scientist discovered that darker values of colors such as the C=80% M=60% Y=30% below, which have a significant gray component...
... can be printed mostly black ink screened to the % dot matching the gray component, with lesser percentages of CYM used to tint it. So that blue above might be printed with 30% black, 50% Cyan, 30% magenta and minimal yellow. Yellow ink in blue mostly adds value (grayness).
The viewer, looking at both images printed with the full CYM% vs the one with the color reduced can't detect the difference when it technique is applied correctly.
That process called "gray component replacement" is commonly used in offset printing and saves a huge amount of money on ink because black is much less expensive than CYM. Reducing the amount of color also makes it much easier to maintain consistent color over a long pressrun.
The same color theory principles apply when editing photos in Photoshop. Skin tones are a combination of RGB and underlying the color are equal parts of RGB, the "gray component". Red is the dominant color. Red in skin highlights when correctly exposed is at about 240 on a scale of 0- 255 so there isn't much multiply will do to it saturate significantly more. Green the second most dominant and blue the least. Any move in Photoshop to darken a skin tone, regardless of what control is used will primarily affect the balance of green and blue to red.
As for multiply, I took the illustration used above:
And then applied 10-100% of a multiply layer to it in 10% increments using the opacity slider. I saved each iteration and then assembled them together to show exactly how various amounts of multiply affect a hue with varying degrees of saturation (the horizontal color blocks) and varying degrees of value (brightness / grayness):
This is the top row with 100% brightness and varying levels of saturation from 60 = 100%
Not a huge change....
This is the column on the right where the saturation was kept constant at 100% and the brightness was decreased from 100% to 60%
Multiply has more of a perceptible effect on colors with increased values.
Like any tool sucess in using multiply on a photo is a matter of skill and experience. I first learned of the technique in a magazine article written by Dan Marguis back in the day before PS had adjustment layers. I rarely use the multiply adjustment layer to darken and saturated normally exposed skin. I use it mostly for toning down distractions in clothing and background and vignetting the edges of an overall dark photo.
|
