On this page 10 sections
A two-minute clip from a modern phone can easily be 500 MB. A one-hour screen recording can be several gigabytes. These files are too big for email, slow to upload and expensive to store. The good news: most of that size is unnecessary. With the right settings you can often cut a video to a quarter or a tenth of its size and nobody watching will see the difference.
This guide explains what actually makes video files large, then walks through the settings and tools that shrink them while keeping them sharp.
Why video files are so big
Three things decide file size:
- Bitrate: how much data is used per second. File size is simply bitrate multiplied by duration. See what is video bitrate.
- Resolution and frame rate: more pixels and more frames need more data to look good.
- Codec: the compression method. Newer codecs describe the same picture with fewer bits.
Cameras and phones record at very high bitrates because they cannot re-think their choices later, and storage is cheap. A phone might record 1080p at 17 Mbps and 4K at 50 Mbps or more. For viewing, 1080p looks excellent at 5 to 8 Mbps with a good encoder. That gap is where your savings come from.
The golden rules of compressing well
1. Use a quality-based setting, not a fixed bitrate
Most encoders offer a constant quality mode: CRF in x264 and x265, RF in HandBrake, CQ in hardware encoders. You pick a quality level, and the encoder spends exactly as many bits as each scene needs. Simple scenes become tiny, complex scenes get what they need. It is the most efficient way to shrink a file without visible loss.
Good starting points:
- H.264 (x264): CRF 20 to 23. Lower is higher quality. 18 is often visually lossless.
- HEVC (x265): CRF 22 to 28. The scale differs; 24 in x265 is roughly equivalent to 20 in x264.
- AV1 (SVT-AV1): CRF 28 to 35, depending on preset.
2. Choose an efficient codec
- H.264 plays everywhere. Choose it when you are sending files to others or uploading to platforms.
- HEVC (H.265) saves roughly 30 to 50 percent at the same quality. Excellent for Apple devices, modern TVs and archives, but some websites and older PCs do not accept it.
- AV1 is the most efficient mainstream codec, with savings of up to 50 percent or more over H.264, and royalty-free. Encoding is slower, and older devices cannot decode it.
Our comparison of H.264 vs H.265 vs AV1 covers the trade-offs in detail.
3. Use a slower preset if you have time
Encoders have speed presets, from “ultrafast” to “veryslow”. Slower presets search harder for efficient ways to describe each frame. At the same quality, “slow” can produce files 10 to 20 percent smaller than “fast”. For one-off exports, “slow” or “medium” is a good balance. “Veryslow” has diminishing returns.
4. Lower the resolution when it makes sense
If your video will be watched on phones or in a small window, 4K is wasted. Dropping from 4K to 1080p cuts pixels by 75 percent. From 1080p to 720p cuts them by more than half. For screen recordings with small text, keep the resolution and use a quality setting instead.
5. Match the frame rate to the content
Talking heads, lectures and screen recordings do not need 60 fps. Converting 60 fps to 30 fps saves bits. Keep 50 or 60 fps for sports, gaming and fast motion.
6. Do not forget audio
Audio is small, but a WAV or PCM track in a MOV file can be surprisingly large. Convert to AAC at 128 kbps for speech or 160 to 192 kbps for music.
7. Never compress an already-compressed file twice if you can avoid it
Each re-encode loses a little more detail. Always compress from the best original you have.
Method 1: HandBrake (free, Windows, Mac, Linux)
HandBrake is the easiest free desktop tool for high-quality compression.
- Download HandBrake from its official site and open your video.
- Choose a preset from the Preset menu. “Fast 1080p30” or “HQ 1080p30 Surround” are good starting points.
- On the Video tab:
- Video Encoder: H.264 (x264) for compatibility, or H.265 (x265) for smaller files.
- Framerate: “Same as source”, or 30 if you want to reduce 60 fps footage.
- Quality: choose Constant Quality and set RF to 20–22 for H.264 or 22–26 for H.265.
- Encoder Preset: Medium or Slow.
- On the Dimensions tab, lower the resolution if needed, for example to 1920×1080 or 1280×720.
- On the Audio tab, choose AAC at 128 or 160 kbps.
- Pick a destination and click Start Encode.
Compare the result with the original on the screen where it will be watched. If you see blockiness, lower the RF by 2 and encode again. If it looks identical and you want it smaller, raise RF by 2.
Method 2: FFmpeg (free, command line)
FFmpeg gives you full control and is perfect for batches.
Good-quality H.264, widely compatible:
ffmpeg -i input.mov -c:v libx264 -preset slow -crf 21 -pix_fmt yuv420p \
-c:a aac -b:a 128k -movflags +faststart output.mp4
Smaller HEVC for Apple devices and archives:
ffmpeg -i input.mov -c:v libx265 -preset slow -crf 24 -tag:v hvc1 \
-c:a aac -b:a 128k -movflags +faststart output.mp4
Resize to 720p and 30 fps at the same time:
ffmpeg -i input.mp4 -vf "scale=-2:720,fps=30" -c:v libx264 -preset slow -crf 22 \
-c:a aac -b:a 128k -movflags +faststart output-720p.mp4
Hit a specific file size using two-pass encoding. First calculate the bitrate with our bitrate calculator, then:
ffmpeg -y -i input.mp4 -c:v libx264 -preset slow -b:v 2400k -pass 1 -an -f null /dev/null
ffmpeg -i input.mp4 -c:v libx264 -preset slow -b:v 2400k -pass 2 -c:a aac -b:a 128k output.mp4
On Windows, replace /dev/null with NUL.
A few notes on the flags: -pix_fmt yuv420p ensures the widest playback compatibility, -tag:v hvc1 makes HEVC files play in Apple’s QuickTime and Photos, and -movflags +faststart moves index data to the start of the file so web playback begins immediately.
Method 3: in your browser, with no upload
For quick jobs, our free online video compressor works entirely inside your browser. You choose a file, a target resolution and a bitrate, and it re-encodes the video on your own device, so nothing is uploaded. It runs in real time, so a three-minute clip takes about three minutes, and it offers fewer options than HandBrake or FFmpeg. It is ideal for shrinking a clip to share in a chat app or email.
Method 4: hardware encoding for speed
If you have a lot of footage, GPU encoders are far faster than software:
- NVIDIA NVENC (
h264_nvenc,hevc_nvenc,av1_nvencon newer cards) - Intel Quick Sync (
h264_qsv,hevc_qsv,av1_qsv) - Apple VideoToolbox (
h264_videotoolbox,hevc_videotoolbox) - AMD AMF (
h264_amf,hevc_amf)
Hardware encoders are typically 5 to 20 times faster but produce slightly larger files at the same quality. For archives and uploads where you have time, software encoding wins on size. For daily batches, hardware wins on speed.
Settings for common goals
| Goal | Codec | Resolution | Setting |
|---|---|---|---|
| Email or chat attachment (under 25 MB) | H.264 | 720p or 480p | Two-pass to target size |
| YouTube or Vimeo upload | H.264 | Source resolution | CRF 18–20, high quality (platform re-encodes) |
| Website background or hero video | H.264 or AV1 | 720p–1080p | CRF 24–28, no audio, short loop |
| Online course lesson | H.264 | 1080p | CRF 21–23, 30 fps |
| Personal archive | HEVC or AV1 | Source | CRF 22–24 (HEVC) |
| Streaming platform (ABR) | H.264 plus HEVC/AV1 | Full ladder | Capped CRF, see our encoding settings guide |
How to check you have not lost quality
- Watch it where it will be watched. Phone, laptop or TV, at normal viewing distance.
- Look at the hard parts: dark scenes, gradients like skies, fast motion, and fine textures like hair and grass. Artefacts appear there first.
- Compare side by side by pausing both versions on the same frame.
- Use objective metrics if you process many files. VMAF, developed by Netflix, scores perceived quality from 0 to 100. Scores above about 93 to 95 are usually indistinguishable from the source for most viewers. FFmpeg can compute it with the
libvmaffilter.
Mistakes that ruin quality
- Using a very low fixed bitrate for a high resolution. The encoder cannot keep up.
- Choosing the fastest preset for final exports.
- Compressing a file that was already heavily compressed, such as a video downloaded from a social platform.
- Upscaling. Making a 720p video into 1080p adds pixels but no detail, and increases file size.
- Stripping too much audio quality. Muffled speech is more noticeable than a slightly softer picture.
Summary
To compress a video without visible loss, use constant-quality encoding (CRF or RF), pick an efficient codec, use a slower preset, and lower resolution and frame rate only where the viewer will not notice. HandBrake is the friendliest tool, FFmpeg the most powerful, and our in-browser video compressor the quickest for small jobs. Start with the settings above, check the hard scenes, and adjust by a couple of steps until the file is as small as it can be while still looking right.
Frequently asked questions
Can you compress a video with zero quality loss?
Only with lossless compression, which produces very large files. In practice, 'without losing quality' means without visible loss: modern encoders at sensible settings produce files that look the same as the original at a fraction of the size.
Why is my phone video so large?
Phones record at high bitrates, often 15 to 50 Mbps for 1080p and 4K, so the camera never runs short of data. Re-encoding at 5 to 8 Mbps for 1080p usually shrinks the file by 70 to 90 percent with no visible difference.
Is HEVC or H.264 better for compression?
HEVC gives similar quality at roughly 30 to 50 percent smaller size, but not every device and website accepts it. H.264 is the safe choice for sharing. AV1 is even more efficient but slower to encode.
Does zipping a video make it smaller?
Hardly at all. Video files are already compressed, so ZIP or RAR saves only a few percent at most.