 |
|
|
|
|
|
 |
| ||||||||
 |
 
|
|
|
|
|
|
|
|
|
| ||||||||
|
|
| 
| User list | Rules | You are not logged in.
>As far as USM, I recommend doing in the the Luminosity channel mode in RGB mode without conversion to LAB and then back.
I am not familar w/ this technique, can you elaborate?
FYI I am a big believer in color profiling... I have a color calibrated monitor, ICC profiles for the labs that I use for printing, etc. I was just saying for my purposes I sharpen @ final size, then make all of the necessary color adjustments just as a last step.
Offline
Sean T.McHugh wrote:
Another complication of using an unsharp mask is that it can introduce subtle color shifts.
http://www.cambridgeincolour.com/tutori … p-mask.htm
As usual, this cambridgeincolour.com tutorial is concise and very well illustrated: I could not add any useful matter...
Last edited by GURL (2006-03-05 19:26:45)
Offline
> Another complication of using an unsharp mask is that it can introduce subtle color shifts.
If simply run on RGB, I agree. Using Lightness in LAB avoids this...
I am now playing around w/ that other suggestion, not sure I am getting the hang of it yet.
Offline
ckiely: you can acheive better results by:
initially upressing image to a desirable size
make all color and tonal adjustments with adjustment layers
save layered file in PSD or PSB format - this is your MASTER file.
flatten image
sharpen (there are MANY ways to do this)
save as tif
to apply USM in Luminosity mode: apply USM, press Cntrl-Z (undo), press Cntrl-Shift-F (Edit/Fade) and then choose mode Luminosity. You can also modify the opaciry of the applied USM at this point.
Offline
Michael Ezra wrote:
Try this, open image saved in sRGB color space (gamma 2.2) and assign (do not convert) ProPhoto ICC profile (gamma 1.0) to it. You will see dramatic difference.
Because of its very wide gamut, don't forget that ProPhoto colorspace must be used in 16bits mode. If not, you won't have enough precision, and poor results.
Offline
Another method is to duplicate the existing layer, set the blending mode to "Luminosity" and the opacity to say, 50%, then "over" sharpen this new layer. Adjusting opacity adjusts the strength of the effect.
You can also adjust the Blending Options settings so that neither the very darkest nor the very lightest areas of the sharpened layer are used. This should prevent noise in the dark areas from being sharpened and prevent near white pixels from being "sharpened" to pure white.
Search the web on "mid tone sharpening" for variations on this approach.
Offline
What type of sharpening are you currently using? You might want to try the edge sharpening action that I have. I got it from my art school instructor who ran Rowell's business for several years.
I'm sure you're already doing it right, but anyway print profiling should come after sizing then sharpening that way the LAB conversion is a non-issue. Why would anyone save the up-rezzed file as the master file? Uprezzing should be for the print file only for that specific size of print output. There's no other purpose for uprezzing other than to take up harddrive space.
Offline
