Skin tone is the first thing a viewer subconsciously evaluates in a video. They don’t need to know anything about color spaces or grading — they can tell instantly whether a face on screen looks healthy or sick. The good news is that you don’t have to guess it by eye. You can simply measure it, and DaVinci Resolve gives you everything you need for free.
Watch the video: Stop Guessing Skin Tones — How to Use DaVinci Resolve Scopes
I’ve also covered the whole process in a video, where you can see every step happening on screen. This article is the written version, so you can come back to the numbers any time without scrubbing through the footage.
Why skin goes wrong even when you don’t notice
Plenty of situations push a shot toward one color cast or another. Mixed lighting, a bounce off a colored wall, the camera’s auto white balance, a fluorescent tube in the background. It always affects the whole image, but it does the most damage to skin, because skin is exactly where the eye has the most precise expectation of how things should look.
Sometimes you can see it at a glance. Far more often you can’t — especially after an hour of editing, when your eye has adapted to the error and started reading it as normal. That’s precisely what the scopes built into Resolve are for. Since we know how skin tone is supposed to behave, we can check it any time, in seconds.
What follows looks complicated and technical. That’s only because I’m explaining it. Once you’ve measured skin tone a few times, you’ll start to see it almost at a glance. Without measuring.
Setup: isolate the face
I’m working in the Rec. 709 color space. On the Color page, open up the video scopes.
If your shot is properly exposed straight out of camera, it should pass this test exactly. Correct white balance and correct exposure show up precisely as skin tone landing where it should, with no color shift at all.
To make this easy to see, start by isolating the area of the frame with the face you’re interested in:
- Switch to the Qualifier tab.
- Pick a bright, well-exposed part of the face.
- Click Invert.
- Turn on the highlight view.
- Pull the gain all the way down, so the whole image goes black and only the face is left.
Now, against that black background, you’re seeing nothing but correctly exposed skin. No speculars, no shadows. This is the step most tutorials skip, and it’s the most important one. Without it you’re measuring a blend of skin, background, hair and wardrobe, and the scopes give you an average that tells you nothing about skin at all.
Test 1: exposure
First, let’s see whether the skin is exposed correctly. The Waveform shows you how the shadows and highlights are distributed. Set the scope to the percentage scale.
- Light skin roughly 60 to 70 IRE
- Medium skin roughly 45 to 60 IRE
- Dark skin roughly 30 to 45 IRE
IRE = Institute of Radio Engineers, a historical unit of video signal level that survives today as a percentage scale.
It doesn’t always have to land perfectly. Depending on how the scene is lit, the face will sometimes read a little brighter, sometimes a little darker, and that’s fine. These are reference ranges, not a standard. What matters is that once you’ve measured, you know whether the exposure of the shot is fine — or whether you missed it by twenty IRE.
Test 2: hue
Now let’s check the hue of the skin. The scope for this is the Vectorscope. In its settings there’s a toggle for the skin tone indicator. If you can’t see it, switch it on.
Skin tone has to sit very close to that diagonal line. It’s fine if it spreads out on both sides of it. But watch out if it starts dropping below the line, into the green. Skin like that reads badly and looks unnatural.
Skin that’s too red plots above the line. Too green plots below it. Correct skin tone sits right along it.
The interesting part is what happens when you compare several faces at once. Light and dark skin have practically the same hue. They differ in brightness and saturation, not in color. Put three different faces on the Vectorscope and all three will stay around the same line. That’s exactly why this test works universally, with no need for lookup tables by skin type.
Test 3: saturation
Finally, saturation. This is where you have the most creative room, but it’s still worth knowing where you stand. The scope for it is the RGB Parade.
Skin saturation comes down to the separation between the red, green and blue channels. On normally rendered skin the order is red highest, green in the middle, blue lowest. The size of the gap between red and blue is exactly what we end up perceiving as skin saturation.
Try it yourself. Push the saturation up and the channels in the RGB Parade move apart. Pull the color back and the gaps close up. On a black-and-white shot, all three channels sit at exactly the same level.
On neutrally lit light skin you’ll be aiming for roughly these values:
- red around 65 IRE
- green around 55 IRE
- blue around 45 IRE
And here’s the most important number in this whole article. The difference between red and blue should be somewhere around 15 to 25 IRE. That’s healthy skin. If that difference is well over thirty, the skin is oversaturated. If it’s under ten, the skin goes grey and looks sickly.
Remember this one number even if you take nothing else away from this article. The red-minus-blue separation tells you more about skin than any other single reading.
What to watch out for
Always measure isolated skin, not the whole frame. Without the qualifier, everything else enters the measurement and the result is worthless.
Avoid speculars and shadows. What you want is the mid-range, well-exposed part of the face. A hotspot on the forehead is a specular highlight, not skin, and it will drag your readings up.
Treat these numbers as a reference, not a rule. The skin tone line on the Vectorscope is a guide, not a law of physics, and its exact position is a legacy of analogue broadcast. If your shot sits a few degrees off it and looks good, there’s nothing to fix.
Then go ahead and break it on purpose
Skin tone isn’t purely a technical standard. In the final shot you’re free to take the whole grade wherever your creative intent and the mood of the video call for. The difference is that once you know where zero is, you also know how far you’ve moved from it — and whether that was your decision or your mistake.
Nothing is stopping you from taking that technically perfect image and finishing it the way the big studios do. Shows like Game of Thrones or SILO show us where cinematography is heading these days. Into darkness, where you can only guess that something might be moving somewhere in there. If you can actually see the actors’ faces, the image clearly isn’t cinematic enough.
But even that darkness comes out better when you know what you sent into it.