Message
Re: [colorvision_group] Saturated blues rendered as purple - Did I miss something?
2008-03-26 by CDTobie@aol.com
In a message dated 3/25/08 5:18:00 PM, kirjoituskone@... writes:
I bought a Spyder3Print to make some custom profiles for our HP
DesignJets.
Used the 729 patch and extented greys targets.
The DesignJet printers build their grays from color inks, so are prone to 'tinted' grays under most viewing lighting. You would need a newer printer model with at least two gray inks to get stable grays in your images.
The profiles make saturated blues to render as quite purplish
and inspecting the generated ICC's on ColorSync Utility show that the
gamut is quite small.
Gamut size issues might best be checked by watching the 'how to' videos on our website. As for saturated blues turning purple, thats a well documented function of L*a*b* as a color reference space. If you take a saturated blue patch in an RGB file in Photoshop, and desaturate it, the result will be a purple. The line from that blue towards gray in Lab space moves through that purple. If you wish to control such hue shifts (blues to purple being the most common, and exaggerated one) you would need to use your profile to check what colors are out of gamut, and bring them in gamut yourself, to the tones you desire, instead of leaving it to the profile to do it mathematically. Or if you don't want to do that degree of correction effort, try the different logic in a different rendering intent. But don't expect out of gamut blues to be blue when printed to a printer that can't reach them.
Additionally, only after doing lengthy manual adjustments I'm able to
produce
truly neutral greys with overall pleasing colors.
Sounds like adjustment for your lighting conditions...
Before manual tweaking the neutral greys and colors has considerable
cast in the a* axis (green <--> red-magenta.)
The color of the cast varies with the media; Glossy Photo has ugly
green cast, Matte Canvas has reddish-purple cast.
May be more going on here, but thats the 'axis of metamerism' that unstable ink blends show under various lighting conditions.
I have also a XRite spectro at my disposal, and none of the
profiles generated with that piece of hardware/software caused
such weird color casts. Also XRite generated ICC's
show considerably larger and smoother gamut.
Again, may be other things going on, or you may just like the "automatic fixes" they have chosen to build into their profiles.
I like the versatility of Spyder3Print spectro and the controls/fine
tuning the Spyder3print software allows. Would like to continue using it.
I have:
- Always calibrated the spectro using the tile in the holder
- D65 complying working space for visual inspection/comparison
- Visually inspected the printed targets for each media (the blues
are in there)
- Measured A2-size targets for each media two times - in "Measured" mode
- Carefully looked for any inconsistencies in the measurements
Sounds like you have that part covered!
Anybody gotta clue why I'm getting such erroneous profiles?
And is it typical that even after 729 patch + grey patch targets the
neutral greys -- and skintones for example -- have quite considerable
colour cast?
No, but everyone's description of a considerable color cast is different, and you are working with a very volitile printing system... so that might or might not explain it.
Am I missing something obvious?
See above, and web videos. You might find an explanation in one or the other...
C. David Tobie
WW Product Technology Manager
Digital Imaging & Home Theater
Datacolor
CDTobie@...
www.datacolor.com/Spyder3
**************
Create a Home Theater Like the Pros. Watch the video on AOL Home.
(http://home.aol.com/diy/home-improvement-eric-stromer?video=15?ncid=aolhom00030000000001)