Re: [APML] Reflecting on Cygnus

[Chuck Vaughn <aa6g@aa6g.org>]

Jerry,

I did some playing around here. Maybe you can help identify the  
source of the difference.

> First, let me ask you what version of Photoshop you are using, and  
> what you have set as the working color space, and what OS you are  
> using?

CS2, Dell 2405FPW Calibrated with ColorSync (not Monitor RGB), and  
Mac OS 10.4.1.

> My point was, his intentions don't matter because most people will  
> view the image in a web browser that does not use color management,  
> so they are going to be viewing it without the profile applied.

And my point is it looks like he adjusted it to be viewed in sRGB. If  
viewed in a browser without that profile applied it may look far  
different. Here's what I see between the two:

http://images.aa6g.org/RG_NGC6914NMM.jpg

You need to view this in PS with profiles applied. At a minimum  
you'll see the relative difference. The top is sRGB and the bottom in  
no color management.

> As I pointed out, this test is not valid because it depends on what  
> the working color space is set to in Photoshop. The image will look  
> radically different depending on the size of the gamut of the color  
> space chosen.

Maybe the main difference in this monitor is brightness and contrast  
as contrary to what I thought the color space of this monitor must be  
similar to CRTs because I don't really see much difference when I  
change the between the various common working spaces. If I change to  
something unusual I can see a difference.

> But, note that this is not because, as you say, there are  
> differences between color spaces.  It is because the wrong profile  
> is being applied to the data.

I think I have this straight as it's what I observed above.

> A more valid test would be to set Photoshop's working color space  
> to the monitor profile, and then see what the image looks like with  
> and without the sRGB color tag applied. (Note that you should never  
> set Photoshop's working space to the monitor profile for real).

Why? If you're only viewing on monitors what difference does it make?  
Or is it one of the standard working spaces is just to normalize  
between monitors?

> However, if he has an LCD display and he adjusted the image for  
> correct brightness on his, and you view it on your LCD display,  
> there shouldn't be that much difference, if both are set up correctly.

Have you noticed that the brightness control on LCDs doesn't act the  
same as it does on CRTs? Turning down the brightness doesn't compress  
the blacks like it did on CRTs. Instead lower contrast does that.

> If you save an image as an sRGB jpeg and then open it in a browser,  
> and then tile both the browser and Photoshop so you can see the  
> image in both at the same time on the same monitor, be it a CRT or  
> an LCD, there should not be that much difference between the two  
> images. IF the monitor is correctly calibrated and has a correct  
> monitor profile and if Photoshop and the OS are all set up  
> correctly for color space management on both your end and on the  
> image creator's end... which, believe me, a very difficult set of  
> criteria to satisfy.

The little image in the link above does just that. You can see the  
big difference.

> But still, your test in photoshop is not valid as you are doing it.

The test is exactly valid for showing the difference being the  
profile applied and not applied when viewed in a given working space.  
That's all the intent ever was. It was not to say one is right and  
one is wrong.

It seems to me that the bottom line is don't do any color management  
if you know the images will be viewed mostly in browsers that don't  
manage color. Embedding profiles and then ignoring them is worse than  
ignoring profiles altogether. ;-)

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