Why does your video still change after you export it? And why does nobody seem able to explain the problem without making it worse?
If you’re reading this, there’s a decent chance you don’t actually want to learn color science. You want to know which box to check. Your grade looks great in DaVinci Resolve. You export it. QuickTime looks a little washed out. Frame.io looks different. YouTube does something else. Maybe Chrome and Safari don’t even quite agree.
So you Google it…and enter hell.
“I FINALLY fixed the DaVinci Resolve gamma shift.”
“The ONLY export settings you need.”
“Why your colors look WRONG on Mac.”
We’ve watched a lot of these videos. Some chase the mythical perfect 1-1-1 tagging combination. Some tell you to change your project-level color management. Some recommend Rec.709-A. Some tell you absolutely not to use Rec.709-A. Others hand you a corrective LUT that deliberately darkens or alters your image so that, after another piece of software changes it, it hopefully lands back where you wanted.
And, maddeningly, some of these approaches work.
Sometimes.
That may actually be the most important thing to understand about the gamma-shift problem.
“Gamma Shift” Isn’t Really One Problem
The term has become shorthand for a whole pile of things happening at once. There’s the color space you worked in. The gamma curve you monitored. The metadata attached to the exported file. The way your operating system interprets that metadata. The way your video player interprets it. The display profile of the monitor. Browser color management. YouTube’s transcode. Frame.io’s player. Apple’s own display pipeline.
And sitting in the middle of all of it is you, looking at two images and saying:
Why the hell don’t these match?
That distinction matters, because a lot of tutorials quietly begin by assuming that whatever difference you’re seeing has one specific cause. It might not. Which is why two intelligent people can offer completely contradictory solutions and both provide screenshots proving that their solution “works.” They may simply be solving different problems.
Professional Tools Went Mainstream. Professional Knowledge Didn’t.
This is where the story gets more interesting to us. DaVinci Resolve is extraordinary software. It puts an absurd amount of professional color infrastructure in front of someone who may have started editing six months ago. That democratization is fantastic. but there’s an awkward side effect. The tool became accessible much faster than the knowledge underneath it did. A generation of filmmakers, creators, YouTubers and small production companies are suddenly working with controls that were historically the territory of colorists sitting in calibrated rooms, using reference displays, working inside tightly controlled delivery pipelines. Today, somebody can download Resolve for free, buy a beautiful MacBook Pro, shoot log footage on an iPhone and find themselves staring at menus containing:
Rec.709 Gamma 2.4. Rec.709-A. Display P3. DaVinci Wide Gamut. Scene referred. Display referred. Data levels. Color-space tags. Gamma tags.
And what they really wanted was:
Make the exported video look like the thing I just spent four hours making beautiful.
This is not a stupid expectation. In fact, it is an extraordinarily reasonable one.
The Missing Middle
The internet has made this problem stranger. At one end are genuine color professionals. These people often understand the subject extremely well. They know what reference monitoring means. They understand transfer functions, display transforms, tagging, calibration and broadcast standards. But expertise creates its own communication problem. Once you’ve internalized an entire technical system, it becomes difficult to remember which concepts are invisible to someone encountering it for the first time. Ask why QuickTime looks different and, five minutes later, you may find yourself receiving a lecture about scene-referred workflows. Technically correct. Completely useless to the person asking the question.
At the other end is the YouTube quick-fix ecosystem. This is where beginners increasingly teach other beginners. Someone discovers that changing one setting makes their particular MacBook, export format and delivery destination look better. They make a video called THE REAL FIX, and suddenly a workaround becomes doctrine. That’s dangerous—not because beginners shouldn’t share what they learn, but because a successful workaround and an explanation of the underlying system are not the same thing.
The people we need more of are in the middle. People far enough along the professional color path to understand what is actually happening, but recent enough students that they still remember which parts were baffling. Leading learners. Technical translators. People capable of saying: “Here’s the complicated truth. Here’s the part of it that actually matters to you. And here’s what I would do today.“ That middle layer feels surprisingly absent.
The Seduction of the Magic Export Setting
We understand why everybody wants one. We want one too. There is something deeply unsatisfying about being told that the solution to your finished video looking different is to go back into your project, change its color-management architecture and potentially rethink the grade.
Nobody wants to hear:
“First, undo everything.”
They want an export preset. A checkbox. A transform applied at the end. And there are absolutely situations where a carefully designed output transform or compensating LUT can produce the visual result you want. But philosophically, something feels wrong about deliberately making an image incorrect so that another piece of software can make it correct again. You’re no longer describing the image. You’re predicting the mistake of the next machine. That can be useful, but it’s not the same thing as solving the problem.
Maybe Perfect Matching Is the Wrong Goal
We’re still working on this ourselves. BRAEID does not currently possess the mythical gamma checkbox. What we’re increasingly interested in instead is predictability. If Resolve, QuickTime, Frame.io, YouTube, an iPhone and a calibrated reference display cannot be guaranteed to show mathematically identical images, then perhaps the practical professional goal is slightly different:
Build a pipeline you understand.
Know what you are monitoring.
Know what you are exporting.
Know how your primary delivery platforms tend to interpret that file.
Test deliberately.
And stop changing five variables every time something looks different.
The goal may not be eliminating every perceptual difference across every screen. The goal may be knowing why the difference is happening, how large it is likely to be, and whether it actually matters. That is a much less satisfying YouTube title. It is also probably closer to the truth.
So, What’s the State of Gamma in 2026?
Messy.
Professional color standards are colliding with consumer displays, modern operating systems, browser playback, streaming platforms and a rapidly expanding population of creators who were never trained as colorists. The software is getting easier. The underlying problem is not. And until the industry does a better job of building a bridge between those two realities, we’re probably going to keep seeing the same cycle: A washed-out export. A panicked Google search. A seventeen-minute tutorial promising the final answer. And another checkbox.
For now, our approach at BRAEID is less exciting:
Understand the pipeline. Change one variable at a time. Test on the platforms that actually matter. And be deeply suspicious of anyone who tells you they solved gamma shift forever.
Including, eventually, us.