Colorimetry and software

jschlarb

Active member
Messages
68
Reaction score
0
Location
Duluth, GA, US
So I'll start by saying that my Dell U2410 monitor is now calibrated to sRGB, 6500k, 2.2 gamma and with dE

Unfortunately my joy was short-lived. I looked at test images for gamma, brightness and contrast using several programs. First I set the color management on all of these programs to point to my new ICC profile. Capture NX2 and Photoshop both look perfect. But Firefox and FastPictureViewer are showing a much lower gamma (around 1.5). The difference is obvious viewing side-by-side. I haven't tried rebooting yet, but I'll do that when I get home. Then I'll read old DPR threads to see how the camera settings fit into this.
 
Capture NX2 and Photoshop both look perfect. But Firefox and FastPictureViewer are showing a much lower gamma (around 1.5). The difference is obvious viewing side-by-side.
Just following up on this with a solution. The problem was that I had told Windows7 to scale my fonts by 50%. This also caused most programs to scale images by 150%. This distortion is why the gamma tests were failing. Turning off text scaling solved the problem.

BTW my note above was wrong. Photoshop and Firefox handle the 150% scaling request correctly (did not scale pictures), while CNX2 and FPV scale both images and text.
 
So I'll start by saying that my Dell U2410 monitor is now calibrated to sRGB, 6500k, 2.2 gamma and with dE
May I ask why limit yourself to sRGB? The U2410 is wide gamut which means that it will cover most of aRGB quite well. In order to profile it properly you should use Standard or Custom color mode, instead of the other presets which limit your gamut and your options. See here
http://www.tftcentral.co.uk/reviews/dell_u2410.htm

In order to get the best out of your hardware (video card LUT + Monitor), you should first try to get the color temperature and brightness right through the individual RGB gain settings (Custom mode only). Having done that, you can let the software do its profiling and generate video card LUT curves, which in my case are pretty close to being linear. I use the free DispcalGui/Argyll software with a Spyder 3 Express puck and the results are outstanding, also matching my inkjet colors quite closely.

The only thing left to do is to tell Win7 what profile to use and set up Capture NX2 to open raw files in aRGB.
Cheers,
Jack
 
Thanks for the tips Jack. I chose sRGB because I don't plan to print very much; I'm targeting displays. I was worried that if I adjust my photos using aRGB, they'll look washed-out on other monitors. Is that a valid concern, or can I use aRGB for my monitor, and sRGB for all my editing programs? I'm still getting the hang of color management. I use an x-rite eye-1 to calibrate; it was definitely worth the investment.

Thanks again.
 
I curious, (and certainly mean no offense), but why would you think your image would look the same on some other persons monitor when they may or may not calibrate it? I was under the impression that color management usually deals with getting what you see onscreen to match what you print? :)
Warm Regards
KarlT
--
Karl H.Timmerman M.A.J.D.
Website: http://www.karltimmerman.com
BLOG: http://www.karltimmerman.com/Ramblings/
'The best part of taking the moral highground ...... is the view'.
 
Thanks for the tips Jack. I chose sRGB because I don't plan to print very much; I'm targeting displays.
I presume that if you went to the expense of buying a wide gamut monitor it is because you wanted to be able to see colors more accurately than it is possible with standard monitors - most newer ones just cover sRGB, which is not good for displaying your parrot and canary collection :-) Therefore set up your software and excellent new monitor in aRGB to show off its stuff, as I outlined above.

In a color managed system, if properly tagged (i.e. you check the include ICC profile when saving) your images will look as close as possible to what you saw on your profiled, color managed monitor independently of what monitor the recipient is viewing them on. If they have a profiled sRGB monitor they will see an sRGB translation of your aRGB image (as if you had done the conversion in CNX2 yourself before saving - no loss of brilliance); if they have a profiled wide gamut monitor they will see a very close approximation of the aRGB image that you are seeing on yours.

The problem arises when you send a properly tagged, say aRGB, image to someone who is using non color managed software to view it (for instance Chrome), or if you send it untagged. In these cases the receiving software will ignore the tag and assume that your data is in the 'web's color space (sRGB) and display it as such. However, different numbers mean different things in different color spaces, so the image will look very different than the way you saw it on your monitor. It's a bit like receiving a set of dimensions to build a box and assuming that the values are in inches while in fact they were in centimeters :-)

To get around this issue, if you do not know whether the recipient of your image files is going to be using color managed software to see them (99% of cases), after post processing convert the image to the sRGB color space before saving and tag it (in CNX2 I have Adjust/ColorProfile/ConverttoProfile/ NikonsRGB/RelativeColorimetric/UseBlackPtComp set up as a pre-set saved adjustment). This way your viewer will see the image as close as possible to what you are seeing after this last step, independently of set up and monitor - but you will have taken advantage of the full power of your wide gamut monitor during PP.

Cheers,
Jack
 
Thanks for your very clear explanation. I just had 2 more quick questions:
  • Should I set the color mode on my D7000 to aRGB as well?
  • Do you save 2 versions of your files after PP, both aRBG and sRGB?
I assume I want to keep the RAW file in aRGB, and do the conversion to sRGB as the last step before saving to JPG for the web. It would probably also make sense to scale the image to about 4MP too; nobody needs to see a 16MP image on the web.

Thanks!
John
 
I resize to between 600 and 900 pixels, which is about a 350 Kb image in Srgb and atJpeg 80 unless I particularly want it to be larger for a specific reason.

No point in letting people steal your high res work.
 
Thanks for your very clear explanation. I just had 2 more quick questions:
  • Should I set the color mode on my D7000 to aRGB as well?
  • Do you save 2 versions of your files after PP, both aRBG and sRGB?
I assume I want to keep the RAW file in aRGB, and do the conversion to sRGB as the last step before saving to JPG for the web. It would probably also make sense to scale the image to about 4MP too; nobody needs to see a 16MP image on the web.
If you have Capture NX2, to get the most out of your images you should shoot NEF (+Jpeg). As the NEF contains raw data it does not matter what color space you choose in camera because you can change it later in CNX2 as if you had chosen that in-camera. I would suggest that you set the camera color space to sRGB, so that if you shoot NEF+Jpeg, the +Jpeg is ready to be emailed/shared as-is in the most universal format. This wil also ensure that the colors on the LCD screen of your Nikon camera are correct (most cameras do not color manage the LCD screen).

You should on the other hand set up CNX2 to open your NEFs in aRGB (Edit/Preferences/ColorManagement/Default color management=NikonAdobeRGB, check the 'Use this instead of embedded profile' tick box).

Once you have achieved your visual objective through adjustments in CNX2, save the NEF - that's your master file. If you then need a Jpeg to email or send to a printing service, simply open the NEF and add a step to convert to sRGB (Fit To the size you want and resharpen as needed) and Save it As a Jpeg. For web use, anything larger than about 1920x1200 will be too large for most screens today (I have a setting file called 'Email' that FitsTo 1600x1024, converts to sRGB and sharpens USM 25/3/2 in one go).

Cheers
Jack
 

Keyboard shortcuts

Back
Top