In a message dated 7/25/06 2:27 AM, Phil Cruse wrote:
>> From: Marco Ugolini <email@hidden>
>> There are:
>> - matrix-based profiles (using the type of matrices you learn of in linear
>> algebra, usually 3x3): these are found in RGB working profiles like
>> AdobeRGB, ColorMatch RGB and ProPhoto RGB;
>> - and then there are profiles that are based on LUTs (Look-Up Tables, 2 per
>> rendering intent: one from source profile to the Profile Connection Space,
>> i.e. A2B; the other from the PCS to the target profile, i.e. B2A). CMYK
>> output profiles are LUT-based (e.g., US Web Coated (SWOP) v2, US Sheetfed
>> Coated v2, etc.).
>>
>> But it's worthy of mention to note that there is one RGB working space that
>> I know of, PhotoGamutRGB, that breaks the "rule" and is LUT-based.
>
> Marco
> I've been fascinated by PhotoGamutRGB for some time, as it breaks the "mould"
> of matrix-based RGB colour spaces. It appears to be technically a printer
> profile. I have found it useful as an INPUT profile for digital camera images
> which don't fit either Adobe 1998 or sRGB. It is also possible (if you have
> profile editing software) to edit it to get nice pleasing images from cameras
> which are un-profiled. Some of my clients use it most of the time. I advise
> them to check images with several RGB spaces and pick the best profile, but
> they take the easy option! I have ended up editing PhotoGamut (mainly
> reducing saturatin) and using it for my own camera. What are your, and
> others, opinions on PhotoGamutRGB?
> Cheers
Phil,
My knowledge of PhotoGamut RGB is very superficial, at this point in time.
In my professional imaging work, I mostly stick with the usual spaces
(AdobeRGB for RGB and US Web Coated SWOP or US Sheetfed Coated for CMYK).
I haven't found drawbacks in this way of working significant enough to force
me to reconsider these choices and try something else.
What I can see so far is that:
- PhotoGamut RGB has a 3D shape that, among the usual standard RGB spaces, I
would say is closest to ColorMatch RGB, which it still surpasses in quite a
few places. And that is good, in my opinion, because I consider ColorMatch
RGB too small for my taste (too geared towards those who are afraid to stray
from the safety of US Web Coated SWOP).
- Unlike sRGB, it encompasses both US Web Coated (SWOP) v2 and US Sheetfed
Coated v2 very nicely. AdobeRGB also encompasses these 2 CMYK spaces, but
with overall shape differences that appear more stark. ColorMatch RGB also
has the disadvantage that it clips a good portion of the color volume of US
Sheetfed Coated, not to mention what I consider a significant enough portion
of US Web Coated (SWOP) as well. None of that is the case with PhotoGamut
RGB.
- One slight disadvantage appears when I check how PhotoGamut RGB deals with
neutrals. Using the Lab Neutral Test file in ColorThink, then showing how
the neutral values are rendered in this profile, I see a strangely wavy
shape in luminosity values between 0 and roughly 50-60%, with peaks of up to
2 Delta E (I am not sure which flavor of Delta E is used by ColorThink 2.2).
This behavior is simply not found in ColorMatch RGB, where deviations are
small, smooth and constant (not at all "wavy"), with peaks below 1 Delta E.
(AdobeRGB does well too, except in the highlights, where it exhibits a
sudden and disconcerting deviation as high as 7 Delta E).
When all is said and done, these are observations that may not amount to
much in practice, making the use of PhotoGamut RGB more a matter of personal
preference than of proven advantage.
Regards.
--------------
Marco Ugolini
Mill Valley, CA
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Colorsync-users mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/colorsync-users/email@hidden
This email sent to email@hidden
