Mathematical relationship where pixel values correspond directly to light intensity — opposite of gamma-encoded space. Essential for physically accurate compositing and color grading.
You are working in a color space that maps physical reality: a pixel value of 50 means exactly half the brightness of a pixel value of 100. No tricks, no curves — pure mathematics. This is linear color space, and it's the reason your VFX actually work. While most monitors and cameras compress their signals with gamma encoding (to exploit human perception), you must compute in linear space because additive and multiplicative operations follow physical laws there.
The consequence on set: Your RAW files from RED or ALEXA arrive linearly on the hard drive — or you linearize them immediately in editing. Why? Because every color correction, every luma adjustment, every layer compositing must happen in linear space. Try adding two lights in sRGB (gamma-encoded) — you'll get garbage because the brightness doesn't add up as it should. In linear space, 0.5 + 0.3 = 0.8. Period. In gamma encoding, you lose information and get banding and color distortions.
Practical Workflow
You import your footage into DaVinci, Nuke, or After Effects — and immediately set the Project Settings to linear. In Nuke, this is the non-destructive standard scenario: Input Node reads RAW, works linearly, Output Node converts back to display gamma. In Fusion, you need the so-called Linear Color Processing Pipeline. Even if your monitor displays sRGB (gamma-corrected), the computation runs linearly in the background.
The tricky part: If you accidentally composite in gamma space — perhaps because the project settings are wrong — you won't notice until the final output. Your layer blending modes will work incorrectly, your grading tools will yield mathematically incorrect results, your keys will be muddy. You'll see this immediately when you view linearly. That's why switching from gamma to linear is one of the most common mistakes for junior colorists. The adjustment isn't an option — it's standard in every professional pipeline since around 2008, when the DCI specification and 32-bit floating-point color became standard.