Video is the heaviest file type most people handle. A few minutes of phone footage can be several hundred megabytes, which is too large for email, awkward for messaging apps and slow to upload. The good news is that most video can be made dramatically smaller without anyone noticing, provided you change the right things.

What is inside a video file

A video file is a container (MP4, MOV, MKV, WebM, AVI) holding at least one compressed video stream and usually one or more audio streams. The container decides how the pieces are packaged; the codec decides how the picture is compressed. Two files with the same .mp4 extension can use different codecs, which is why some play on a device and others do not.

The four numbers that decide file size

1. Bitrate

Bitrate is the amount of data spent on each second of video and is the single biggest influence on file size. File size is approximately bitrate multiplied by duration. At 8 Mbps a ten-minute clip is about 600 MB; at 2 Mbps it is about 150 MB. Bitrate is how much "budget" the encoder has to describe the picture.

2. Resolution

Full HD (1920 x 1080) has about four times the pixels of 960 x 540. For the same visual quality, more pixels need more bitrate. Dropping a 4K clip to 1080p often shrinks it to a quarter, and on a phone or laptop screen the difference is hard to see.

3. Frame rate

Most content looks natural at 24 or 30 frames per second. 60 fps footage takes noticeably more space and is worth keeping only for fast sport or gaming. Converting 60 fps to 30 fps typically saves 30 to 40 percent.

4. Codec

H.264 is the safe default: it plays almost everywhere. H.265 (HEVC) and VP9 reach similar quality in about half the bitrate, and AV1 is more efficient still, but support, especially for editing and older devices, varies. Efficiency has a cost in encoding time.

  • Sending by email or messaging app: 720p, 30 fps, H.264 MP4, around 1.5 to 2.5 Mbps, AAC audio at 96 to 128 kbps.
  • Uploading to a video site: keep the original resolution and frame rate, H.264 MP4 at a high bitrate (platforms re-encode anyway; do not pre-compress aggressively).
  • Sharing a lecture or screen recording: 1080p or 720p at 15 to 30 fps with a low bitrate. Screen content compresses extremely well because most of the picture does not change.
  • Storing a large library: VP9 in WebM or MKV at a moderate quality; keep important originals.
  • Embedding on a website: MP4 H.264 for compatibility, optionally WebM VP9 as an alternative source.
  • Short silent loops: animated GIFs are large and limited to 256 colours; short MP4 or WebM loops are smaller and look better.

Constant quality or target bitrate?

Encoders can aim for a constant quality, letting the bitrate vary with scene complexity, or for a fixed bitrate. Constant quality is best when you care about how it looks and do not have a hard size limit. A fixed bitrate is useful when you must hit a size, for example a 25 MB attachment limit. To estimate: size in megabytes times 8, divided by the duration in seconds, gives the total bitrate in megabits per second available (including audio).

Step by step

  1. Trim. Remove the dead start and end. It costs nothing and saves more than any setting.
  2. Lower the resolution to the largest size anyone will actually see.
  3. Choose H.264 MP4 unless you know the target supports something newer.
  4. Pick a quality level and look at the result on the device it is for.
  5. Check the audio. Speech is clear at 64 to 96 kbps; do not spend megabits on it.

Why a "compressed" video can come out bigger

If the source was already efficiently compressed, for example a video downloaded from a streaming site at a low bitrate, re-encoding at a higher quality setting makes it larger without improving it. Always choose a quality level that is no higher than the source. A good tool will also keep the original if it cannot make the file smaller.

Compatibility pitfalls

  • iPhone videos are often MOV with HEVC; some Windows computers need an extension to play them. Converting to H.264 MP4 solves it.
  • MKV is flexible but not accepted by many TVs, phones and editing apps.
  • AVI and WMV are legacy formats best converted to MP4.
  • Some websites reject files that are very long or very large regardless of format.

Using TinyFileHub

The Video Converter changes the container and codec (for example MOV to MP4, MKV to MP4 or MP4 to WebM), and the File Size Reducer shrinks a video with one slider. Both process your file once and hand it back; nothing is kept on the server.

In short: trim, reduce the resolution, use a sensible bitrate and a widely supported codec. Those four steps usually turn a bulky clip into one that sends in seconds and looks fine on every screen.

A worked example

You have a 12-minute screen recording of 1920 x 1080 at 30 fps that is 1.1 GB. That is roughly 12 Mbps, far more than screen content needs. Re-encoding as H.264 MP4 at about 2.5 Mbps video and 96 kbps audio gives about 12 x 60 x 2.6 / 8, or roughly 235 MB, with text still readable. Dropping to 1280 x 720 and 1.5 Mbps brings it near 140 MB. Cutting the first and last minute removes another 15 percent. Three small decisions took a file that could not be emailed to one that can be shared by link without anyone caring about the quality.

Estimating before you start

Multiply the total bitrate in megabits per second by the duration in seconds and divide by 8 to get megabytes. For a 25 MB limit and a 3-minute clip: 25 x 8 / 180 gives about 1.1 Mbps in total, so choose roughly 1 Mbps for video and 64 to 96 kbps for audio, at 480p or 360p if needed. Doing the sum first avoids repeated trial and error.

Things that change the result

  • Hardware encoders in phones and graphics cards are very fast but less efficient than software encoders, so they need a higher bitrate for the same quality.
  • Two-pass encoding spends bitrate more cleverly when you must hit an exact size.
  • Noise and grain are expensive to encode. Footage shot in low light compresses badly; denoising before encoding can save a lot.
  • Variable frame rate from phone screen recordings can confuse editors; converting to a constant frame rate avoids audio drift.