Printing and the HSV versus HSL histograms

[Charles2]

This problem image has a dramatic disparity between its HSV and HSL range of values. It printed with much weaker colors than seen on the monitor.

http://www.flickr.com/photos/41790885@N ... otostream/

For comparison this image printed close to what is seen on the monitor. Same Epson R2880 printer, paper, and Epson profile.

http://www.flickr.com/photos/41790885@N ... otostream/

The result implies there is no big problem with monitor and printer calibration. There was no duplication of profile in the workflow. I printed many variations of the problem image:

Different profiles and paper specifications in the Epson 2880 print driver dialog
Extra saturation
Profile intent set at Relative Colorimetric or Saturation
A different paper
Expanded image to Full Range HSL, Composite | Filter at amount 15% onto original, saved for printing

Failure, until...

The last variation on the list happened after I noticed that the image has a nearly full HSV range but its HSL range is far below maximum at the highlight end (goes up to about the 71% point when you examine it in Levels and Color | HSL). By comparison, the image that prints without a problem has full range in both HSV and HSL. (Would I have noticed this difference if my processing software was not Picture Window Pro? Perhaps after more wasted paper, if ever.)

When I print, I save the image by itself or on a layout from PWP and open it in Fast Stone. The FS print dialog has a gamma adjustment. I eventually thought to reduce it from the default 1.00 to 0.80. Success! I could probably go to 0.85 gamma for somewhat better shadows, but at the moment resources and dedication are a bit low.

Questions:

• Is HSV-HSL disparity a sign of bad exposure and/or processing, or does it just happen sometimes, especially with flowers?

• Does color theory explain a relation between HSV-HSL disparity and print gamma? Or is this pair just an accidental contrast?

• When preparing a version of an image for the printer, is there a more or less standard procedure that will take care of this problem? Otherwise, it is print, try gamma adjustment, print, ...

[Dieter Mayr]

Charles

In HSV color model a value of V = 0% means pure black, and V = 100% the "purest " color, without any black.
In HSL L = 0% is also pure black, but L = 100% is pure white, so a "pure color is reached at L= 50%

In your first picture you have no bright white areas, but pure colors, so H goes high, but L does not, because of the lack of white.
In the second picture its different because of the white flowers and stones, so L goes up to high values too,
It has nothing to do with "bad" exposure, as long you do not consider every exposure as "bad" that follows the old rule in BW to have every tone between black and white in a good exposed picture.

Did you try to make a comparison between your printout and PWPs softproof ?
Did you try to print out from PWP itself with proper color management settings?
I normally get a quiet good comparison with softproof, but I print directly from PWP.