Reduce video resolution online

Drop a file in below and it opens with 720p already selected — one step down from Full HD, and the step most files want. Change it to 1080p, 480p or 360p in the resolution row. The encode runs inside this page, so the video is never uploaded anywhere.

Drop your video here

or click to browse — MP4, MKV, MOV, WebM, AVI and more

The four steps, and what each one costs you

Every step down roughly halves the pixel count, and pixels are what the encoder has to spend bits on. The right-hand column is the video bitrate band each step actually suits — the same thresholds the tool uses when it suggests a resolution for a size target, so the advice here and the advice in the compressor cannot drift apart.

Step Pixels (16:9) Of 1080p Suits a video bitrate of
1080p 1920 × 1080 100% 2,000–4,999 kbps
720p 1280 × 720 44% 800–1,999 kbps
480p 854 × 480 20% 400–799 kbps
360p 640 × 360 11% 150–399 kbps

Read it in whichever direction you have the number for. If a size limit leaves your video stream 600 kbps, 1080p is not on the table and 480p is; if you are starting from a resolution and want to know what to budget, the band is the range to aim at. Below about 150 kbps no step rescues the clip — trim it instead.

4K down to 1080p

A 3840×2160 frame holds 4× the pixels of 1080p, which is why phone and drone 4K files are so large for their length. Dropping to 1080p is the least painful edit available on that footage: on a phone screen or in a chat client the difference is close to invisible, and the encoder suddenly has four times the bits per pixel to work with. The tool has no 4K step because it never needs one — leaving the resolution on Original keeps 4K, and picking 1080p is the downscale.

Resolution is the second lever

Reach for the bitrate first. The quality slider and target-size mode both change how many bits the video stream gets, and for most files that alone is enough. Resolution is what you add when the bitrate you need is too low for the frame you have — when the arithmetic for a hard size limit leaves 1080p with less than it can carry, and the encode would come back soft and blocky.

The two combine well. Set the target size, look at what the tool solved for the video stream, and match the resolution to the band in the table above. That is the whole method: decide the budget, then pick the frame the budget can afford.

What the scaling does to your frame

Each step fixes the height and lets the width follow from the source's aspect ratio, rounded to an even number because H.264 and H.265 encode in even-sized blocks. Nothing is cropped, letterboxed or stretched: a 4:3 clip at 720p comes out 960×720. Worth knowing for portrait video — the step always names the height, so a 1080×1920 phone clip at 720p becomes 404×720, a much harder cut than the same step applied to landscape footage. Go one step gentler on vertical clips than you would on a widescreen one. There is no upscaling worth doing either: picking a step larger than the source only spends bitrate on pixels that were interpolated, never recorded.

Questions

Does lowering the resolution make the file smaller on its own?

Not by itself — file size is bitrate times duration, and the resolution is not the bitrate. What a lower resolution does is let a lower bitrate look clean, because the same bits are spread over fewer pixels. In the tool the two levers work together: pick a step down, and the quality slider or the target size then has an easier job to do.

Can I scale a video up to 1080p or 4K?

You can pick a step above your source, but there is no reason to. Upscaling invents no detail, and the extra pixels cost bitrate, so the file gets bigger and looks no better. Only pick a step below whatever the source already is.

What happens to the aspect ratio?

It is preserved. Each step sets the height and lets the width follow, rounded to an even number because H.264 and H.265 require even dimensions. A 4:3 clip scaled to 720p becomes 960x720, not 1280x720 with bars.

Is downscaling slower than compressing at full resolution?

The scaling itself costs a little time, but the encode that follows is working on far fewer pixels, so a downscaled encode almost always finishes sooner overall than the same encode at the source resolution.

Keep reading