Color-grading for vertical video - what changes from 16:9

Most color grades are built for 16:9 viewing on a laptop. When you take that same grade and slam it into a 9:16 frame on a phone, three things change at once - the brightness target, the contrast curve, and the saturation ceiling. Ignoring any of them produces the muddy, grayed-out look that signals 'this was a horizontal video first.'

This is the working approach I use when I have to publish the same video in both aspect ratios.

The phone is a different device

A laptop display, in a typical viewing environment, sits at 80-120 nits and is calibrated for desk distance. A phone, watched on the train, can hit 600-800 nits - and the viewer's eyes are six inches from it. That's not a tweak, that's a different display.

For vertical, you grade for an eye that's much closer to a much brighter screen. Practically, that means:

• Lift the shadows by 5–8%. Crushed blacks on a laptop look cinematic. On a phone in daylight, they look broken - the viewer just sees a black panel and assumes the upload failed.
• Pull highlights down 3–5%. The phone is going to add its own headroom; you don't need to push to 100% to get punch.
• Soften the contrast curve. A snappy S-curve that flatters a wide shot turns vertical talking-heads into raccoon eyes. Aim for a gentler middle.

Saturation behaves differently in vertical

This is the part most graders miss. On a small, bright screen, the perceptual saturation of the same RGB values is higher. A 16:9 grade that reads as "warm and inviting" reads as "orange and aggressive" when you crop it to 9:16 and watch on a phone.

The rough rule: take your saturation back by 10-15% for vertical. If you graded to a +18 saturation lift on 16:9, the vertical sibling lives at +5 to +8. Skin tones especially - a slight push that looks healthy on a laptop will look sunburned on a phone.

Crop, then re-grade. Always.

If you do one thing differently between formats, do this: crop the timeline first, then build the vertical grade on top of the cropped frame. Do not export the 16:9 grade and reframe it.

The reason is that the area of the frame that's now showing in 9:16 has a different brightness distribution than the full 16:9. Your wide shot might have been balanced against the sky on the left - that sky is now cropped out, and the grade is responding to a frame that no longer exists.

The workflow:

1. Set the project to 9:16.
2. Reframe each shot (this is the part the Aspect Converter accelerates - it'll surface the Safe Area crop windows for each platform).
3. Apply a base grade - call it your "vertical LUT."
4. Then do per-clip adjustments on top.

Skipping the per-clip pass is the single biggest reason vertical re-cuts look muddy. The base LUT cannot read intent - you have to tell each shot what it's supposed to feel like.

The Safe Area reality

A 9:16 frame has a different safe area than its 16:9 source. On Instagram Reels, the top 14% is reserved for the username overlay, and the bottom 18% is reserved for caption and like buttons. That leaves roughly 68% of vertical real estate for your image.

When grading, watch the safe-area zone, not the full frame. A bright spot that's beautifully placed in the center of the 9:16 frame can fall directly behind the like-button overlay and disappear from the viewer's eye. The grade has to flatter the surviving zone, not the export zone.

Skin-tone math

A useful trick: in vertical, the face is a larger percentage of the visible frame than in 16:9. A medium shot that was 35% face on the laptop is 60% face on the phone. That means skin-tone errors are 1.7× more obvious.

Two adjustments that hold up:

• Hue: push skin tones a half-degree toward yellow, away from magenta. Phones aggressively warm faces; if your master is already on the warm side, the phone tips it over.
• Saturation: isolate the skin range (typically 7–35 on the vectorscope, depending on subject) and pull it back 5%. The rest of the frame can stay at full saturation.

A 9:16 base grade you can start from

Drop this on a vertical timeline as a starting point and tune from there.

Exposure       :  +0.0
Lift           :  +0.04   (R 0.04 G 0.04 B 0.05)   ← cool the shadows
Gamma          :  +0.0
Gain           :  -0.03   (-0.03 across the board) ← tame highlights
Contrast       :  0.85    (softer than 16:9)
Saturation     :  0.92    (lower than 16:9)
Skin Hue       :  -0.5°
Skin Saturation:  0.95

Treat those as starting numbers, not destinations. They get you in the right neighborhood; the eye finishes the grade.

Test on the actual device

I know - this is the advice everyone gives and everyone skips. But for vertical, you must. Export the cut, AirDrop it to a phone, watch it once in a bright room and once in a dark room. Things you'll catch on the phone that you'll miss on the laptop:

• Banding in the gradient sky (phones are aggressive about compression).
• Edge halos around high-contrast titles (the smaller frame magnifies them).
• Skin tones drifting magenta in shadow (phones cool low-light areas).

The 90 seconds you spend on the device save you the hour you'd otherwise spend trying to debug the grade on the laptop and wondering why it still looks wrong on the phone.

When to surrender the matched grade

There's a case for two genuinely different grades - one for 16:9 and one for 9:16 - when the formats are going to different audiences. A YouTube viewer expects cinematic; a Reels viewer expects punchy.

If the same video lives on both, build two grades that share a hue and split on contrast. The cinematic version sits with deeper shadows and more saturated mids; the social version sits with lifted shadows and slightly desaturated highlights. They feel like cousins, not strangers, but each one is calibrated for its destination.

Bring it to VideoCue

The Aspect Converter handles the reframe - you upload a 16:9 export, pick the destination ratio, and it returns the cropped frame plus the Safe Area overlays for each major platform. From there, the Color Grade entry walks through the underlying terminology (lift/gamma/gain, the vectorscope, the false-color tools) if you want to push past the base LUT and build your own from scratch.