[Author Prev][Author Next][Thread Prev][Thread Next][Author Index][Thread Index]

Re: [APML] Reflecting on Cygnus



Jerry,

> No, it's not.  That's not what you are doing.  You are actually  
> showing the difference with the correct profile applied, and with  
> an incorrect profile applied.
>
> Your test completely depends on what you have set as the working  
> space in Photoshop.
>
> If you have Photoshop's working space set to sRGB, you won't see  
> any difference at all.

After much experimentation I see that you're right.

> Well, there is, obviously, a ton of confusion about this entire  
> subject.  I don't think that not doing any color management at all  
> is any better of a solution.  In fact, I don't think you can even  
> "turn color management" off in Photoshop because the display is  
> always going to be output through the monitor's profile.
>
> I wonder if the brightness difference you are seeing is because you  
> are working on a Mac and have the gamma set to 1.8 instead of 2.2?

Well this was part of the experimentation. I used this chart to set  
the gamma as accurately as I could. Have you seen this?

http://www.aim-dtp.net/aim/download/mccp_v1.4.gif

The entire article is here:

http://www.aim-dtp.net/aim/calibration/monitor_calibration/ 
monitor_calibration.htm

The gamma needs to be set to 1.3 for this display for the best  
results on that gamma calibration chart. I suppose that's not a  
surprise since LCDs are supposed to be more linear than CRTs. They  
should require less gamma correction.

This highlights a problem between LCDs and CRTs for on-line viewing.  
Even though there were variations between CRTs, they weren't that  
great. The color gamuts weren't hugely different and the gammas  
weren't either. This probably accounts for why something done on one  
monitor looked passable on another, especially if both had been  
calibrated, even without color management.

Now along comes an LCD with a hugely different gamma correction. Non  
color managed images generated using a CRT look like crap on it when  
viewed in monitor space.


Color Sync provides some nice graphical data for each profile in the  
system. I screen shot a couple of these.

Colorant Tristimulus for the Dell 2405FPW, sRGB EIC61966-2.1, Adobe  
RGB 1998

http://images.aa6g.org/colorant tristimulus.jpg

You'll see right away that the 2405 color gamut is almost identical  
to sRGB.


Red Gamma Curve for the Dell 2405FPW, sRGB EIC61966-2.1, Adobe RGB  
1998 (green & blue are virtually identical to red)

http://images.aa6g.org/response curve.jpg

This is where the difference is.

So this is a new problem. Every unmanaged color image on the internet  
generated by looking at a CRT monitor looks washed out on this  
display in monitor space. So unless I use Safari and hope that images  
have a profile with them, any image I care about is going to look poor.

Now that I have the best calibration on my monitor that I know how to  
perform, I really need to view images with Photoshop in sRGB working  
space.... and save them with the embedded profile to remain  
compatible with the CRT world.

Seems like every time a new and more advanced piece of technology  
comes along it reveals existing but mostly unrecognized problems.

So although this display has only the color gamut of sRGB, it  
produces a brighter, higher contrast and sharper image.

Chuck

_______________________________________________
Astro-Photo mailing list
Astro-Photo@seds.org
http://seds.org/mailman/listinfo/astro-photo