Digital BW, The Print

As the result of recent experiences, I am curious if anyone uses the 
ProPhotoRGB colorspace in photoshop when correcting/printing photos.

I had some photos I wanted to correct into B&W, and someone suggested 
trying the prophotoRGB instead of my usual adobeRGB. The expanded 
space was of course beautiful and seemed to capture subtle highlights 
much better (at least on the histogram: I only have an 8-bit monitor 
so I'm sure I'm not seeing everything). 

I went to print on my Epson 2400 with UT3D inkset, and found that 
much of my nuances and subtleties were completely lost on the printed 
output (using museo sr paper). I expected some of this, since the 
printer is a much lower resolution device. 

It took some effort to essentialy compress the colorspace into 
something the printer could better output, and at the end of it I 
thought I would have been better off just sticking with AdobeRGB 
colorspace, which is already smaller and has a shorter path to get to 
printed output.

musings include:

- do folks use prophotoRGB colorspace in general?
- If so, how do you think about/manage output to lower res devices?
- If it creates more headaches than not, is it even worth considering 
for bw desktop output?

>>>>
The benefits of a
larger space are that as output devices get better, they'll be able to
produce a larger gamut of color.....<<<<

Would you say this is true with the current product offerings in printers? 
Is there any increase in color gamut resolution between say the Epson 2200 
and 4000?


Take care,
Gregory david Stempel
www.fireframeimaging.com
www.soundexposure.org

>
> - do folks use prophotoRGB colorspace in general?
> - If so, how do you think about/manage output to lower res devices?
> - If it creates more headaches than not, is it even worth considering
> for bw desktop output?
My understanding is that there's a trade-off with larger color spaces: 
namely, the large the space, the more colors it can represent; but the 
larger the steps are between colors when the space is represented by a 
limited number of samples (bit depth). So, if you use a space which is 
considerable larger than your output device, you won't see a benefit 
from the parts of the color space that the output device can't produce, 
and you might lose subtle transitions of colors due to the large spaces 
between colors in the file in the larger space.  The benefits of a 
larger space are that as output devices get better, they'll be able to 
produce a larger gamut of color.  Whether there's any gain or trade off 
in practice will depend on a lot of things. It should be too hard to 
test this with your system.  If you're working with raw files from a 
digital camera, convert the raw file to two different files, each with a 
different color space. Then do your photoshop stuff on both of them. 
Print with preview using a color profile for your output device, and see 
what, if any, differences there are.

> different color space. Then do your photoshop stuff on both of them. 
> Print with preview using a color profile for your output device, and
see 
> what, if any, differences there are.
>

here's an interesting test I just conducted:

created two identical 16-bit gray scale ramps, one in adobeRGB and the
other in proPhotoRGB colorspace. histogram and numerical details look
identical.

applied an identical 's' type curve to both images, bumping the
highlights a bit and darkening the shadows. just to create some
representative processing.

converted both profiles to Epson 2400 museo sr color profile (I think
this was supplied by crane and downloaded from their website).

The prophotoRGB registered the mean color as 154.11 with Median 168

the adobeRGB registered the mean color as 143.23 with median 172

Don't know what this means other than in this simple experiment they
don't translate the same, even when the resulting color profile is
identical. the prophoto histogram appears to have more luminance. 

interpretations welcome :)

Adobe RGB is a gamma 2.2 colorspace.  I don't know what the gamma is  
for proPhoto RGB, but it is probably not 2.2 and that may be why you  
are seeing this difference.

On Mar 20, 2007, at 4:12 PM, bvalente wrote:

>
>> different color space. Then do your photoshop stuff on both of them.
>> Print with preview using a color profile for your output device, and
> see
>> what, if any, differences there are.
>>
>
> here's an interesting test I just conducted:
>
> created two identical 16-bit gray scale ramps, one in adobeRGB and the
> other in proPhotoRGB colorspace. histogram and numerical details look
> identical.
>
> applied an identical 's' type curve to both images, bumping the
> highlights a bit and darkening the shadows. just to create some
> representative processing.
>
> converted both profiles to Epson 2400 museo sr color profile (I think
> this was supplied by crane and downloaded from their website).
>
> The prophotoRGB registered the mean color as 154.11 with Median 168
>
> the adobeRGB registered the mean color as 143.23 with median 172
>
> Don't know what this means other than in this simple experiment they
> don't translate the same, even when the resulting color profile is
> identical. the prophoto histogram appears to have more luminance.
>
> interpretations welcome :)
>

Hi,
The differences 168 vs 172 are due to gamma differences.  Pro Photo is 1.8 and Adobe 
RGB is 2.2.
Paul

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "bvalente" <bvalente@...> wrote:

>
> 
> > different color space. Then do your photoshop stuff on both of them. 
> > Print with preview using a color profile for your output device, and
> see 
> > what, if any, differences there are.
> >
> 
> here's an interesting test I just conducted:
> 
> created two identical 16-bit gray scale ramps, one in adobeRGB and the
> other in proPhotoRGB colorspace. histogram and numerical details look
> identical.
> 
> applied an identical 's' type curve to both images, bumping the
> highlights a bit and darkening the shadows. just to create some
> representative processing.
> 
> converted both profiles to Epson 2400 museo sr color profile (I think
> this was supplied by crane and downloaded from their website).
> 
> The prophotoRGB registered the mean color as 154.11 with Median 168
> 
> the adobeRGB registered the mean color as 143.23 with median 172
> 
> Don't know what this means other than in this simple experiment they
> don't translate the same, even when the resulting color profile is
> identical. the prophoto histogram appears to have more luminance. 
> 
> interpretations welcome :)
>

And don't forget to play with different rendering methods (perceptual, 
saturation, colormetric...) when moving from you editing space to your 
printer profile.  Use soft proof to play around with this.