JimKied
Offline
• • •
Upload & Sell: On
|
RustyBug wrote:
Great Question ... not provocative at all.
Basically, I just compare my R, G & B values for an area that is supposed to be a neutral color. This can be of a tonal value that is nearly white, gray or nearly black. If you have a known (or desired) neutral, then you can read your R, G & B values to asses which one(s) are high/low.
For instance, an RGB value of 127, 127, 127 would be a neutral gray. If the numbers were 127, 120, 127 then they would be low in green (opposite of magenta) and thus have a magenta cast for a neutral.
The same would apply to say 235, 245, 245 would be low in red (opposite of cyan) and thus a cyan cast, and a 220, 220, 245 would be high in blue and have a blue cast. Likewise, on the shadow end of the spectrum a value of 15, 20, 15 would be high in the green channel and be a green cast ... you get the gist.
Well I played around with the color
What I do is first study the light and the subject matter to assess what I think SHOULD, or I DESIRE for it to be neutral. Then I usually pick four points in the highlights that I want neutral. Often times they four points will not match, i.e. two might be high in green, one high in blue and one neutral or cyan or red or 
I use color balance tool in PS (has adjustments for highlights, mids & shadows) to bring the values for the known/desired neutral to match at say 220, 220, 220 or whatever range I'm working in. As I watch the four points, I typically find that I can't get them totally neutral, but three out of four I can usually get within five points of each other ... for me, that's my assessment of which light @ AI/AR is going to be the light I have as my WB.
Then, I repeat for my BB on the shadow ... then back to WB to refine what the BB may have tweaked on a bit.
It takes way longer to explain than to actually do, but once you get the hang of reading your numbers ... then you start to see the casts with your own eyes more as well, but you've always got the numbers to rely on.
Often times though, I can get a scene "neutral" ... and it can look a bit clinical. For those, I just dial back the opacity on the layer to taste (usually around 70%-80%) and kinda "split the diff".
BTW ... I make NO ADJUSTMENTS in ACR ... all in PS with the more refined color adjustment tools.
Anyway, I hope this helps with creating too much confusion.
...Show more →
Well, I researched Google and played around with the color balance tool, but I couldn't see how it was affecting the values - it seems more of a tool that you slide the sliders one way or another for taste - and I don't trust my taste.  I'm sure I must be doing something wrong, so I need to do more research in that area. Plus, I was not seeing any color tint in my psd file.
But I did notice this. When I converted to sRGB, there is a definite shift. Here is my process for converting to web. I convert to profile (sRGB). Change mode to 8 bits. These two moves created a very slight shift to being more warmer and maybe a little darker. Then I resized to 2/3rds of the way to my 850 pix wide final image. I oversharpen this particular size (150% this time - less than what I did before). This act of downsizing added more warmth (magenta?). After sharpening, I then resized to the 850 wide image shown here. The resulting web sized image is warmer and a little darker than the original. I sometimes translate this to being a little more saturated, but I am going to lean towards just warmer on this particular image. BTW, I used the perceptual rendering intent.
Bear in mind that this happens with every image I down size for web viewing. Does this happen for you guys also? And do you do anything about it before posting? And BTW, here is another version of the image which I want to post up to see how it looks on FM. Appreciate any input you have
|
, definitely not a turtle, but I guess a turtle looking rock.