Bitrate, resolution and frame rate: which one to cut to shrink a video

A video’s file size is its bitrate multiplied by its length, and nothing else. Resolution and frame rate matter only because they change how much bitrate you need for the picture to look good. So the order is: lower the bitrate first; drop the resolution when the bitrate gets too low for the number of pixels you are asking it to cover; and only reduce frame rate to go from 60 to 30, because cutting frames saves far less than you would expect.

The one equation that matters

Bitrate is how much data the video is allowed per second, usually written in megabits per second (Mbps). Multiply that by the duration in seconds and divide by eight to get megabytes.

A three-minute clip at 8 Mbps works out to 8 × 180 ÷ 8 = 180 MB. The same clip at 4 Mbps is 90 MB. That is the whole story of file size; everything else is about what that bitrate buys you.

It follows that the most direct way to hit a target size is to set the bitrate that gets you there. A tool that asks you for a target size, like the Shrink mode in Manatee, is doing exactly this arithmetic for you: you name the megabytes, and it finds the quality that fits.

Bitrate: the lever that actually changes the size

Modern codecs like H.264 and HEVC do not store every pixel of every frame. They store a full frame occasionally, and for the frames in between they store only what changed. The bitrate is the budget the encoder has for describing those changes.

Lower the budget and the encoder starts economizing: coarser detail in busy areas, smoothed textures, and, when it runs out of room, the familiar blocks and banding in flat regions like skies and walls. Fast motion, confetti, rain, water and grass are expensive; a talking head against a plain wall is cheap.

This means the “right” bitrate depends on the content. A lecture recording can look good at 2 Mbps in 1080p. A drone shot over trees will look bad at four times that.

Halving the bitrate halves the file size. That is a clean, predictable trade, and for moderate cuts (say, a phone’s 4K clip at 50 Mbps down to 15 Mbps) the quality loss is genuinely hard to see on a phone or laptop screen.

Resolution: cut it when the bitrate can’t keep up

Resolution is the number of pixels in each frame. 4K (3840 by 2160) has four times the pixels of 1080p (1920 by 1080), which has about two and a quarter times the pixels of 720p (1280 by 720).

Resolution does not by itself determine file size: a 4K file at 5 Mbps is exactly the same size as a 1080p file at 5 Mbps. What it determines is how thinly the bitrate is spread. At 5 Mbps, the 4K version has a quarter of the data per pixel, so the encoder smears detail to cope, and you end up with a big frame full of mush. The 1080p version at the same bitrate looks cleaner, and a phone or laptop screen cannot show the extra 4K pixels anyway.

So the rule is: when the bitrate you need for the target size is too low for the resolution, drop the resolution. Rough bands where 1080p stops looking good are somewhere around 3 to 4 Mbps for typical footage; below that, 720p at the same bitrate usually looks better than 1080p. For 4K, keep it only if you can afford upwards of 15 to 20 Mbps, and only if it will be watched on a 4K screen.

Dropping resolution on its own, without changing the bitrate, does not reduce size. Most tools tie the two together (QuickTime Player’s Export As 1080p picks a bitrate appropriate for 1080p), which is why resolution changes appear to shrink files. It is the bitrate that moved.

Frame rate: the lever that gives back less than you’d think

Frame rate is how many frames are shown per second: 24 or 25 for film-style footage, 30 for most phone video, 60 for sports and screen recordings of games.

Intuition says 30 fps should be half the size of 60 fps. It is not, because of the way codecs store only the differences between frames. At 60 fps, consecutive frames are very similar, so each one is cheap to describe. Remove every second frame and the remaining frames are twice as far apart in time, so each one now has more change to describe and costs more. In practice, going from 60 to 30 saves somewhere between a fifth and a third of the bitrate for the same visual quality, not half.

Going lower than 30 is usually a mistake for anything with motion: 15 fps looks like a slideshow, and the saving is modest. The exception is a slide-style screen recording, where almost nothing changes between frames and frame rate barely affects size.

The one good reason to cut frame rate is 60 to 30 for content that does not benefit from 60: a family video, a product demo, a talking head. For a screen recording, our guide to making it smaller covers the specifics.

Codec: the fourth lever nobody mentions

Bitrate, resolution and frame rate all trade size for quality. Changing the codec can improve the trade itself.

HEVC (H.265) reaches roughly the same quality as H.264 at something like 60 to 70 percent of the bitrate. An iPhone records HEVC by default for exactly this reason. The catch is compatibility: H.264 plays everywhere, on old laptops, in every browser, inside every chat app. HEVC plays on Apple devices and most modern hardware but still trips up older machines and some web uploaders. If the recipient is a person on an Apple device, HEVC is a free win. If the destination is “anyone”, stick with H.264. We go through the trade-offs in H.264 vs HEVC.

Audio is the other lever people reach for, and it is almost never worth it. A 128 kbps AAC track is about 1 MB per minute, a rounding error next to a video track at 5 Mbps, and cutting it makes music sound worse.

A decision order you can use

Faced with a file that has to get smaller, work through these in order and stop when you hit the size.

1. Trim. Cut what nobody needs to see. Seconds removed cost nothing in quality.
2. Lower the bitrate to whatever gets you to the target, as long as it stays above the comfortable band for the resolution (roughly 3 to 4 Mbps for 1080p, 8 Mbps and up for 4K).
3. Drop the resolution one step (4K to 1080p, 1080p to 720p) if the bitrate from step 2 would be below that band, then set the bitrate again.
4. Cut 60 fps to 30 if the footage is 60 and does not need to be.
5. Switch to HEVC if every recipient is on a modern device and you still need more room.

If you are working to a number rather than a feel, the Shrink mode in Manatee takes the target in megabytes and finds the quality that fits, so you are not doing the arithmetic by hand; it never touches the original and writes the result alongside. It is not an editor, so step 1 still happens in QuickTime Player or iMovie. For the manual route with specific size caps, see how to shrink a video to a specific file size.

Questions

Why does my 1080p export look worse than the 4K original even though it is still large?

Almost always because it was exported from an already-compressed file rather than the camera original, or because the tool used a low bitrate by default. Check the bitrate in QuickTime Player’s Window, then Show Movie Inspector. If it is under 4 Mbps for 1080p, the bitrate is the problem, not the resolution.

Is a constant bitrate or a variable bitrate better?

Variable. It lets the encoder spend more on the hard scenes and less on the easy ones, which gives better quality for the same file size. Most Mac tools use variable bitrate by default; you only see the choice in advanced encoders.