How to choose between Gamma 2.2 and Gamma 2.4 when color correcting?

Gamma 2.2 vs Gamma 2.4 – How, Why and When (in DaVinci Resolve)?

August 23, 2017

When you start delving into the finer points of color correction you'll eventually learn about Gamma 2.2 vs Gamma 2.4. Learn how to choose between them.


How do you set up DaVinci Resolve for different Gamma settings (and why)?

We recently had a question come in about a common source of confusion:

What color & gamma settings should I use in FCPX & Resolve when it’s going either to the internet or onto DVD’s?

It’s a great question – but opens a can of worms when it relates to setting your gamma on your reference display and the various options in DaVinci Resolve. Why a can of worms? Because you have to consider not one but two different scenarios for setting your gamma:

  • Scenario 1: Your color correction / mastering environment matches the delivery / viewing environment. For example, your mastering environment is set for broadcast and you’re delivering for broadcast.
  • Scenario 2: Your color correction / mastering environment does NOT match the delivery / viewing environment. For example, you’re set for broadcast but delivery is for the Internet.

Each viewing environment has a different recommended gamma setting. And when it comes to setting gamma for color correction, what’s the rule for setting gamma?

1st Rule of Setting Gamma: It’s Based on Your Viewing Environment

Most discussions of setting gamma start with the display you’re using or your final delivery – but that misses the point of Rule #1: When mastering / color correcting your images, you set your gamma based on your viewing environment.

In fact, the differences in the standard gamma settings used by Mac and Windows versus the standard gamma settings for television or cinema are all based on overall brightness levels in the room!

Why?

It’s due to the Simultaneous Contrast Effect of human perception.

The Simultaneous Contrast Effect and How it Relates to Gamma Settings

Wikipedia has a good explanation of how Simultaneous Contrast works. But a single pictures really is worth a thousand words and in the image below, the gray square in the middle of both boxes have precisely the same brightness values (if you don’t believe me, just pull out a color sampler and check for yourself:

The Simultaneous Contrast Effect
The middle gray boxes have precisely the same RGB brightness values – yet they look different. Why? That’s the Simultaneous Contrast in action.

The reason the two middle boxes look like they have different brightness values comes down to human perception. The human brain make brightness assumptions comparatively, not absolutely. None of us has a built-in eyedropper tool that gives us absolute RGB readouts. Our perception of the brightness of an object in a room is effected by room itself.

And if those middle gray boxes are a television or computer screen, and the surround fields is the overall brightness of your room… what implication does that have while color correcting??

The exact same image, in two different rooms—one bright and one dark—will look different! Particularly the brightness values.

Using Gamma to Solve the Simultaneous Contrast Effect

How would you fix the problem of Simultaneous Contrast on the two boxes above? Let’s call the top box our ‘reference environment’; it’s the one with the dark surround field. To get the bottom box to match, we can apply a brightness adjustment to push the middle gray box brighter until its ‘perceived brightness’ matches our reference environment.

And that’s what precisely we’re doing when we move our images between Gamma 2.2, 2.4 and 2.6. We adjust those middle boxes to make them brighter or darker so their apparent brightness in the different viewing environments is equalized.

For most of us, the question comes down to choosing: Gamma 2.2 vs Gamma 2.4?

In this Insight, I’m not going to dive any deeper into the technical explanation of gamma (you can read that explanation on Charles Poynton’s site). But if you’re going to make an informed decision about choosing your gamma settings then you need to understand the standard gamma settings for different viewing environments:

  • Gamma 2.2 – This gamma setting is generally used for brighter viewing conditions. Think: Office setting with bright overheads and lots of indirect daytime light from exterior windows. Under these conditions, the viewing screen needs to pop the shadows or the image becomes a mushy mess and ‘lowering’ the gamma setting helps the image fight glare. Hence, gamma 2.2 is the standard gamma setting for Windows and has been the standard gamma setting on Apple computers since Mac OS 10.6 (prior versions of the Mac OS used gamma 1.8) since computer displays assume a bright, daytime viewing condition.
  • Gamma 2.35 & 2.4 – This gamma setting is generally used for darker viewing conditions. Think: Watching Prime Time television at night in a living room. Under these conditions the danger of the image washing out is much less and the darker parts of images are easier to see. Hence, Gamma 2.35 and 2.4 are used in dim (but not blacked-out) viewing conditions.

Gamma 2.35 & 2.4 have been the unofficial gamma settings for television since its inception. I say ‘unofficial’ because for the first 40 years of TV, the gamma setting was determined by the cathode ray gun on phosphor-based tubes. And the gamma setting of the CRT was fixed and unchangeable. When the gamma of CRTs was measured by the BBC in the early 2000’s, they determined the effective gamma of professional CRTs was between 2.3 and 2.4.

  • Gamma 2.35 – This was the BBC’s initial recommendation for plasma & LCD televisions, to most closely match professional CRTs. This gamma setting has largely been abandoned.
  • Gamma 2.4 – Many professional post-production facilities resisted gamma 2.35 and opted for gamma 2.4 – because they could match their LCDs and plasma displays to their CRTs more easily if they used gamma 2.4. This trend became so strong that the BBC now recommends gamma 2.4. And this is the current standard if you’re mastering for broadcast.
  • Gamma 2.6 – This is the gamma setting is used for ‘blacked out’ viewing conditions, typical of a movie theater. Gamma 2.2 and gamma 2.4 generally look too flat and dim in black box conditions. Hence, for DCI delivery to cinema a Gamma 2.6 is specified.
  • Resources for further research: If you want to dig deeper into this yourself then check out these links below (which open to a new window)

Gamma 2.2 vs Gamma 2.4: How do you choose?

Notice the common thread in my description of each gamma setting… gamma is based on the expected viewing environment. So…

In my grading suite, I have my room’s viewing conditions set for broadcast mastering. All exterior sources of light are blocked, and I have dim (but not completely dark) and controlled sources of light. Therefore, I always set my entire workflow for gamma 2.4.

The delivery destination makes no difference when I’m color correcting! I set my gamma appropriate to my viewing conditions, not my delivery specification.

IF I wanted to master at gamma 2.2 then I need to brighten up the overall levels in my grading suite to more closely match that of a not-too-well-lit office – for the gamma 2.2 setting on my reference display to make any sense.

IF I was grading to the viewer on my computer display and didn’t have a professional external reference display, what gamma would I choose? Well, if the computer display is at its native gamma then I’d set my viewing conditions to match the gamma setting appropriate for that display (usually 2.2, unless you’ve calibrated it differently).

But what if you’re delivering for a different viewing environment?

There are two common scenarios here:

  1. You’re mastering in a broadcast environment but delivering to digital cinema.
  2. You’re mastering in a broadcast environment but delivering to the internet.

In both scenarios, you color correct using the gamma appropriate for your mastering environment and then do a gamma adjustment when rendering – to transform the image so it looks similar to your image, in those darker or brighter viewing conditions.

Colorists often get more detailed – and after applying the overall Gamma adjustment they make scene-by-scene or shot-by-shot adjustments.

When you start getting really detailed like that, colorists call this making a Trim Pass.

If you’re delivering to multiple viewing conditions, Prime Time, internet and cinema – then you might consider doing several trim passes (with each trim being a separate export), each with slightly different gamma settings for those viewing conditions.

How to make overall Gamma adjustments to your timelines in DaVinci Resolve

This video Insight is all about using DaVinci Resolve to transform your images from your viewing / mastering environment and make a gamma adjustment to prepare the images for the delivery viewing environment.

Note: I wrote this post after recording and editing the video – so I don’t quite put my explanation in the video as neatly as I have done here… but you’ll see the workflow for executing this theory.

Have questions?

That’s what the Comments section is for! And don’t forget, if you want email notifications to the comments on this Insight, press the ‘Subcribe’ button at the bottom left of this page, just above the footer.

Enjoy!

-pi

UPDATE – March 2018: Revised an incorrect statement about gamma 2.4 / 2.2 viewed in a black box setting. HT: Steve Shaw.


Comments

25 thoughts on “Gamma 2.2 vs Gamma 2.4 – How, Why and When (in DaVinci Resolve)?”

  1. I guess where I have the most trouble is trying to ensure sure my non professional displays are actually at the 2.2 or 2.4 gamma in the first place. I had a ton of trouble calibrating my LG OLED, I kinda of gave up trying to use calman on it, everytime I restarted the tv and checked the gamma it seemed off.

    With apple computer displays can an icc accurately set a gamma or is the graphics card reading some meta from the icc and handling it?

    I’m also a bit confused on when to use viewer lut, should the computer monitors be sRGB 2.2 and the viewer lut convert that to 2.4, or does Davinci adjust the gamma in the viewers somehow?

    1. The only reliable way to calibrate the consumer OLEDs is with a LUT box. In that case, powering down won’t lose your settings – and you’ll get far more accurate calibrations as a result.

      I’m not really the person to ask about how to color manage a GUI display… because I’m always using a calibratable professional external display and rely on that. If you’re grading to a computer display then you’ll always be at a disadvantage. But I’ll reach out to some people we know who might be able to get you a better answer. I’m about to start a calibration series – and as a follow-up to that series I’ll tackle calibrating GUI displays.

  2. Hi Patrick,

    I’m delivering a job for the Kazakistan Expo 2017. It’s a double video installation in a not properly black environment because several displays in the room. I asked for the specific gamma of video projectors, but they don’t know.I color graded the footage on my reference FSI CM 250 gamma 2.4 So I delivered at gamma 2.4, because I think that the video projection should adhere to the Itu-R BT.709 specifications.
    I did wrong?

    1. Nope. You did right. You graded to the spec. And they couldn’t tell you otherwise. To my eyes, the difference between 2.2 and 2.4 isn’t all that big. But it sounds like what you were delivering to had controlled light sources with the main source of light coming from other displays – so 2.4 is definitely appropriate (lacking further guidance from the venue).

  3. A topic that has haunted my learning for the last few months. I feel this has answered all of my questions and it was all laid out very simply. Thank you very much. This article and video alone is worth the months sub fee, in my opinion.

    Now if only I hadn’t forgotten to bring my notepad to the library today, I’d be able to make notes…Haha.

  4. Hi Patrick,

    Can you link to the managed color workflow insight you mention in the video? I couldn’t find it.

    Do platforms on the internet adjust gamma at all? For instance, I want a Gamma 2.2 workflow for bright office/computer display viewers, but if I upload my video to YouTube, do they do anything to my footage to make adjustments – or is this solely based on the display the viewer is watching on.

    Do you guys/can you guys make a video on how Premiere handles color? Is it entirely REC709? Can you adjust to another rec? What should I do if I want to master for REC2020 but need to round-trip back to Premiere after Resolve?

    Thank you!

    1. Hi Daniel – In DaVinci Resolve, it’s called Resolve Color Management. Here are two Insights on the subject:

      First Look: https://mixinglight.com/color-tutorial/davinci-resolve-12-first-look-resolve-color-management/

      RCM with RAW: https://mixinglight.com/color-tutorial/resolve-color-management-raw-footage/

      Also – ACES is a color managed workflow and Robbie’s ACES series will get you started if you’re working to that specification. Here’s the first Insight in that series:

      Getting to know ACES Part 1: https://mixinglight.com/color-tutorial/getting-know-aces/

      RE: Platforms adjusting gamma – Gamma adjustments happen all over the place; on your computer, in your browser, when uploading to a video sharing service, when displaying from a video sharing service. You can, and should, expect your gamma will get bounced around once you’re outside a carefully color-managed room, like in digital cinema or a grading suite. That’s why we grade to a standard… the standard is a center-point around which all our devices will orbit.

  5. Hi Patrick, another great insight! You’ve addressed an issue I’ve ignored for about the entire time I’ve been grading.
    I have just been focused on 2.4.

    Did you always convert your web-destined videos to gamma 2.2 Even when there were no Resolve Fx and no Rcm? Are all the Mixinglight.com videos gamma 2.2?

    I will definitely start experimenting with this.

    Thanks.

  6. I’ve just had this insite flagged up via a post on LGG, and there I’ve had to disagree with the suggestion that you can alter your gamma for your grading environment.

    While the theory does make some sense, the reality is that any grading environment should be managed to match the specifications for a ‘grade-1’ monitoring environment, which means having a surround illumination of 10% of peak monitor brightness for SDR, and 5 nits for PQ based HDR – as per EBU specifications.

    Gamma should be set to 2.4 for Rec709 deliverables, and 2.2 for sRGB (the internet).

    The concept of altering display gamma is really only valid for home viewing, where you often cannot manage the viewing environment (unless you have a real Home Cinema setup).

    Professional grading rooms should always have the environment managed to match the above stated specifications, and not alter gamma.

    Edited to add: SMPTE now specify a Reference Viewing Environment to have a surround illumination of 5 nits, +/- 0.5 nits, for both HDR and UHD.

    And gamma is specified as per the colour space/video standard – Rec709 at 2.4 (or bt1886 – but we do not recommend that… as the concept is flawed), sRGB at 2.2, and PQ based HDR at ST2084, while HLG has it’s own gamma standard (interestingly WITH a variable component for surround illumination, but as surround illumination is fixed at 5 nits for a Reference Viewing Environment, that too is essentially fixed…)

    1. Hi Steve. Thanks for leaving a comment.

      My contention is that that you set your gamma for your actual grading conditions and then alter to match the deliverable. And I make this contention based on the well documented and often referenced ‘Simultaneous Contrast Effect’. IOW, if I’m color grading for 2.2 in a setting appropriate for 2.4, I’m deluding myself in thinking the results I produce would be the same as if I were actually in the intended viewing setting. As you know, gamma 2.2 isn’t magic. It’s designed for the lighting conditions it assumes.

      I do teach that If you deliver for broadcast then you need to follow SMPTE standards. And I do teach that a professional colorist should strive to meet SMPTE standards for… standardization and interchange between professional post houses.

      But SMPTE gets a lot of things wrong (as evidenced by the 5 nit surround illumination, which is nuts and seems to completely ignore the long-term eye strain issues of looking at these direct view displays in near black conditions for thousands of hours).

      If you want to point me to that LGG thread I’ll be happy to chime in if this Insight is being interpreted differently than I intend.

      1. You actually need to remember that the concept for home viewing is (loosely – as the historical reality is actually different) based on using 2.2 gamma within a home viewing environment to perceptually match the image when graded 2.4 gamma in a reference viewing environment.

        Based on that theory, when grading for the internet, you should also grade 2.4 in a reference viewing environment, so the PC display at 2.2 gamma does exactly the same as a home TV, and appears to give a perceptual match.

        Make sense?

        And while I’m not necessarily in agreement with a 5 nit surround for SDR garding (I do prefer the older 10% of screen max luminance), for PQ HDR it is correct, due to the EOTF used (which makes shadows very dark).
        But, standard are standards and we all need to adhere to them, or there will be no consistency at all.

        And for a reference viewing environment you should always manage the environment, not alter your display gamma.
        (Ignoring the the variation between 2.2 and 2,4 is very small indeed…)

        As for the idea that home TVs were based on 2.2 gamma to generate this ‘perceptual match’, the truth is very different.

        In the ‘old days’ we all uses CRTs, in the studio, and at home, with the ame inherent gamma.
        When flat panel displays came along, there was a need to be backwards compatible.
        And so the manufacturers (incorrectly) added a 2.2 fixed gamma to the screen manufacturing (nit the display processing electronics, to the actual screen driving electronics).
        This means all flat panels have an in-built 2.2 gamma at the screen.
        (Even HDR displays).

        So, the concept of a home TV having a 2.2 gamma to look perceptually the same as a grading display in a reference viewing environment is just happenstance, not reality.

  7. Very entertaining.
    I am moving from tv as a CCU op to Resolve!
    I am experiencing something different with 15b4 and RCM using the ramp as an input.
    I have been working in Rec. 709 2.4.
    When I apply the Colour Space Transform Node, I see the expected result when applying a 2.2 gamma on the output.
    However, when I go to RCM and apply Rec.709 to TimeLine and Output, the curve is very different – a sharp upward curve in the blacks.
    Am I doing something wrong? https://uploads.disquscdn.com/images/019049f2b0cea2e490fcaebcac0793cf33aa59e1bb3aee40482de44afc128250.jpg https://uploads.disquscdn.com/images/5c105bd1af92e6c65de0dee970a6adf186caeb80524ce232008acd7df40a1668.jpg

Leave a Reply

Hundreds of Free Tutorials

Get full access to our entire library of 900+ color tutorials for an entire week!


Start Your Free Trial
Loading...