Ambient light compensation?

Home Forums General Discussion Ambient light compensation?

Viewing 15 posts - 1 through 15 (of 18 total)
  • Author
    Posts
  • #4578

    Gregow
    Participant
    • Offline

    Hi,

    I’ve been trying to rework my calibration and profiling lately.  Previously I used the sRGB presets and just bumped up the quality and number of patches. This has worked pretty well, however I’ve noticed images that put on the web appearing more contrasty than intended.

    Now, this is from various different devices under different conditions. Like phones, tablets or different computers. None of them calibrated, but some pretty decent nonetheless.

    Anyhow, figured it might be because I’m sitting in a very dim environment. My ColorMunki Display reports around 10-20 lux.

    The question is, how do I best compensate for this (without changing the light)?

    Setting the ambient light adjustment in DisplayCAL to 20 lux gives a gamma that is way high, like 2,8-2,9.  I’ve set it to 2,47 with a sRGB-like function and it seems at least reasonable.

    For color managed applications I’ve set a CIECAM02 gamut map, like this:

    http://i65.tinypic.com/2rc4d4y.jpg

    Then in Photoshop I set up the profile for soft proofing with perceptual intent.

    Is this a good way to approach things, or am I just messing things up and getting a more inaccurate display?

    • This topic was modified 4 years, 11 months ago by Florian Höch. Reason: Fix broken image due to mixed content

    i1Display Studio on Amazon  
    Disclosure: As an Amazon Associate I earn from qualifying purchases.

    #4579

    Steve Smith
    Participant
    • Offline

    Hello…

    I would set the gamma to 2.2 to obtain the standard and to reduce contrast.

    Steve.

    #4587

    MW
    Participant
    • Offline

    I’d use the following settings that optimizes the gamma curve according to contrast ratio of the display device – gamma: 2.2-2.3, relative, black output offset: 0%.

    #4589

    Florian Höch
    Administrator
    • Offline

    All of these are valid approaches, but keep in mind that the calibration tone curve doesn’t impact the output of color managed applications.

    For color managed applications I’ve set a CIECAM02 gamut map, like this:

    That seems like a reasonable approach.

    #4591

    Steve Smith
    Participant
    • Offline

    Florian, can I ask you then, what tone curve will the color managed application use… I thought the calibration is placed in the video card affecting all output…(I know we’ve been through this before, but I find I am still confused.)

    Thanks.

    #4592

    Florian Höch
    Administrator
    • Offline

    Florian, can I ask you then, what tone curve will the color managed application use…

    It solely depends on the profiles used and the rendering intent. A relative colorimetric rendering will usually reproduce the source profile tone response in the destination (i.e. the display). A perceptual rendering may do something else, e.g. CIECAM02 gamut and appearance mapping (if a profile incorporates such a mapping).

    #4594

    Gregow
    Participant
    • Offline

    Hello…

    I would set the gamma to 2.2 to obtain the standard and to reduce contrast.

    Steve.

    Alas, but the problem is that contrast gets too low in dim environments.

    I don’t know if gamma 2.2 could be called a standard. It is perhaps the most common transfer curve, but it doesn’t quite adhere to any standard like sRGB. And of course, very few would strictly adhere to the sRGB standard (having non-zero blacks, brighter than 80cd/m2 displays, brighter or dimmer than 200lux ambient light that’s warmer or cooler than D50). Seems like the best bet would be to try and compensate for some of the discrepansies.

    I’d use the following settings that optimizes the gamma curve according to contrast ratio of the display device – gamma: 2.2-2.3, relative, black output offset: 0%.

    What is the reasoning behind this?

    All of these are valid approaches, but keep in mind that the calibration tone curve doesn’t impact the output of color managed applications.

    For color managed applications I’ve set a CIECAM02 gamut map, like this:

    That seems like a reasonable approach.

    Is it possible to get the CIECAM02 transfer curve to the graphics card LUT?

    Florian, can I ask you then, what tone curve will the color managed application use…

    It solely depends on the profiles used and the rendering intent. A relative colorimetric rendering will usually reproduce the source profile tone response in the destination (i.e. the display). A perceptual rendering may do something else, e.g. CIECAM02 gamut and appearance mapping (if a profile incorporates such a mapping).

    Do you know any good resources you can recommend for learning more about how CIECAM02 works? I’ve read a bit on Wikipedia, but it’s a bit too technical.

    #4596

    Steve Smith
    Participant
    • Offline

    It is my understanding that the sRGB standard is [email protected] Perhaps your monitor is an IPS with a poor contrast ratio of 1000:1 or so in which contrast will never approach the look of 3000:1 or 5000:1

    My room lighting is controlled at 15cd m2 and the contrast is perfect at 2.2. (6500K, 80cd m2)

    There is no true looking black on an IPS monitor… It’s gray. (At least compared to an HDTV calibrated the same.)

    #4597

    Steve Smith
    Participant
    • Offline

    Also, check to see if you are getting the maximum available dynamic rang of your monitor by doing a pre-calibration of the brightness and contrast controls using charts.  It’s critical to providing the proper light to dark range to the calibration program to work with…. One click in the wrong direction can cost you significant loss of ‘contrast'(or addition of crush) in the final calibration regardless of what your gamma is set to.

    #4598

    Gregow
    Participant
    • Offline

    It is my understanding that the sRGB standard is [email protected] Perhaps your monitor is an IPS with a poor contrast ratio of 1000:1 or so in which contrast will never approach the look of 3000:1 or 5000:1

    My room lighting is controlled at 15cd m2 and the contrast is perfect at 2.2. (6500K, 80cd m2)

    There is no true looking black on an IPS monitor… It’s gray. (At least compared to an HDTV calibrated the same.)

    Technically sRGB doesn’t have a gamma but tone response curve (if we define gamma as Vout = Vin^gamma). However, the sRGB response curve is very close to gamma 2,2. The difference is very slight, but still.

    Yes, I have an IPS with 880:1 calibrated contrast ratio.

    How do you know the contrast is perfect? Relative to what? Dim environments reduce the contrast and we no longer adhere to the sRGB standard.

    It’s not really a matter of gray blacks but darks appearing less compressed. Meaning, I see a wider range of tonality whereas the darks would appear crushed on other devices.

    #4599

    Gregow
    Participant
    • Offline

    Also, check to see if you are getting the maximum available dynamic rang of your monitor by doing a pre-calibration of the brightness and contrast controls using charts.  It’s critical to providing the proper light to dark range to the calibration program to work with…. One click in the wrong direction can cost you significant loss of ‘contrast'(or addition of crush) in the final calibration regardless of what your gamma is set to.

    There’s really only the brightness slider. The contrast control either clips darks or lights so that’s best left alone.

    I’ve checked the display at lagom.nl and it performs excellently. It’s a fairly new Eizo so I would have sent it back right away if it didn’t.

    #4600

    Steve Smith
    Participant
    • Offline

    I tried using that room light compensation function as well and I got the same results as you.  I was really surprised to see just how contrasty the calibration ended up being… I asked Florian about it and he said that I shouldn’t use that function.  (But he didn’t tell me why… I didn’t ask.)

    Since then I learned to use the gamut mapping feature, CIECAM02, set to ‘Darkened room’ and ‘perceptual’ using [email protected] with white point 80. I also did another calibration at 5000K for printing purposes.

    I’ve done enough calibrations to know when the gamma is within range. As I’ve watched shadow detail come and go in my photographs.  I have developed a real sense of where that range in on my device(s). As a Photographer using Photoshop quite regularly, I am able to see graphically what’s happening on my monitor, and what is suppose to be happening.

    Not entirely scientific, but proven consistent over time… My definition of perfect.  🙂

    • This reply was modified 4 years, 11 months ago by Steve Smith.
    #4602

    MW
    Participant
    • Offline

    What is the reasoning behind this?

    You’re using a IPS display in a dark environment, the displays contrast ratio could be a larger factor than the ambient lightning. A slightly higher gamma makes the displays output appear more contrasty which would help you hold back from using too much contrast when processing photos.

    #4623

    MW
    Participant
    • Offline

    @Gregow

    1. Why 80 cdm/2 when you’re not strictly following the sRGB standard?

    2. It begs the question, have you considered how display technology changes with time? As for the art, currently browsing flickr or 500px you will see many photos with raised blacks. For the sake of argument let’s assume it’s a stylistic trend. Can you rule out higher factory gammas of recent Apple devices playing a part? Could a photos with raised blacks look normal on displays with crushed blacks? Bottom line, are you concerned that your photos to look dated 5-10 years from now?

    #4624

    Steve Smith
    Participant
    • Offline

    Calibrating to 80 cd\m2 with a room brightness of 15 cd\m2 has a perceived proportionality to the standard.  (What the creator saw at standard settings, I will perceive as the same. (At my exponentially proportional settings.)

    I will see what was ‘intended’ (Intentional or not) No artist would edit photos on just any Apple device. Most have strictly calibrated wide-gamut monitors calibrated to standard. Therefore, when created properly, regardless of their standard…My standard will convert it to look the same to me, regardless. …  Kids on tablets, laptops and cellphones aren’t relevant to me.

Viewing 15 posts - 1 through 15 (of 18 total)

You must be logged in to reply to this topic.

Log in or Register

Display Calibration and Characterization powered by ArgyllCMS