Guide
How to reduce video file size without losing quality
You can usually cut a video's size by 50–90% with no visible difference by re-encoding it with H.264 or H.265 at a quality-based setting (CRF 20–24) instead of the camera's fixed high bitrate. Lower the resolution only if the video will be watched on small screens. Trimming and removing unneeded audio save space with zero quality cost.
The video compressor uses quality-based encoding with sensible presets and runs on your device. Start with Balanced and compare the preview before going smaller.
How to shrink a video with Videozify
- Trim off unneeded footage with the video trimmer. It's free size reduction with no quality cost.
- Open the video compressor and drop in the trimmed file.
- Choose Balanced (visually lossless for most footage) or set a target size if you have a hard limit.
- Keep the original resolution unless the video is only for phones; then choose 1080p or 720p.
- Press Compress, compare the preview with the original, and download.
Why camera files are bigger than they need to be
Phones and cameras encode in real time with a hardware encoder and a fixed, generous bitrate, often 15–50 Mbps for 1080p or 4K, so that the worst-case scene (confetti, water, foliage) still looks good. A calm indoor scene gets the same bitrate it doesn't need. A software encoder with a quality-based setting gives each scene only the bits it needs, which is where the large savings come from.
CRF: quality-based encoding
CRF (constant rate factor) tells the encoder to hold a quality level rather than a bitrate. For H.264, CRF 18 is close to visually lossless, 23 is the default and still looks very good, and 28 is visibly softer. Each step of +6 roughly halves the file size. Simple scenes come out small and complex scenes larger, while quality looks consistent throughout.
Use a target bitrate only when you have a hard size limit, such as an email or Discord cap. Then size equals bitrate times duration, and quality depends on how much fits.
- CRF 18–20: archive quality, about 30–50% of camera size
- CRF 22–24: sharing quality, about 10–25% of camera size
- CRF 26–28: small files for chat, visible softening
Resolution: the biggest lever
Pixel count drives file size. 4K has four times the pixels of 1080p, and 1080p has 2.25 times the pixels of 720p. If the video will be watched on phones or in a chat app, 1080p or 720p looks identical to 4K on that screen and is far smaller. Keep 4K only for big screens or for footage you'll crop or edit later. Never upscale: it adds size with no detail.
Frame rate and audio
60 fps needs noticeably more data than 30 fps. For talking, tutorials and most everyday clips, converting 60 to 30 fps saves space with little visible effect; for sport and games, keep 60.
Audio is small, but not nothing. AAC at 128 kbps is transparent for most listeners; camera audio is often 256–320 kbps. For speech, 96 kbps is plenty. If the sound isn't needed at all, remove it with the mute tool.
H.264 vs H.265 (HEVC) vs AV1
H.265 produces files about 40–50% smaller than H.264 at the same quality, and AV1 slightly smaller again. The catch is compatibility: H.264 plays everywhere, H.265 needs recent hardware or paid codecs on some Windows PCs, and AV1 encoding is slow and not supported by all devices. For sharing, H.264 is the safe choice; for archiving on devices you control, H.265 saves space.
Native options
Mac: QuickTime Player's File > Export As offers 4K, 1080p and 720p, using Apple's hardware encoder, which is fast but less efficient than software encoding. The Finder's Encode Selected Video Files quick action offers HEVC for smaller files.
Windows: Clipchamp exports at 480p, 720p or 1080p. It's fine for social video, but you can't set CRF or bitrate.
Handbrake (free, Windows/Mac/Linux) exposes CRF ("Constant Quality"), presets and codecs. It's the best desktop option if you're comfortable with settings.
Phones: iPhone's Settings > Camera > Record Video lets you record at 1080p30 instead of 4K60 to avoid huge files in the first place.
Example: a 1 GB phone video
A 3-minute iPhone video at 4K30 HEVC is around 1 GB. Re-encoding to 1080p H.264 at CRF 23 typically gives 80–150 MB, a 7–12× reduction, and on a laptop or phone screen most people can't tell them apart. Keeping 4K at CRF 23 in H.264 gives perhaps 300–400 MB. Converting to 720p at CRF 24 brings it down to 40–70 MB, still clean on a phone.
The exact numbers depend on content. Static interviews compress far more than handheld footage of a beach or a concert with flashing lights.
How to judge quality
Compare the original and compressed versions on the screen where the video will actually be watched. Pause on scenes with fine detail (hair, foliage, text) and fast motion. Look for blocky patches in dark areas and smearing on motion. If you see none at normal viewing distance, the compression is visually lossless for that use. Zooming in to 400% on a still frame will always reveal differences; that's not how people watch videos.
Two-pass and presets
When you need a precise file size, two-pass encoding analyses the video first and then distributes bits where they're needed most, giving better quality at a fixed size than a single pass. Encoder presets (from ultrafast to veryslow) trade encoding time for efficiency: slower presets produce smaller files at the same quality. In a browser, medium or faster presets keep processing times reasonable.
Keep your originals
Compression is one-way. Once detail is discarded it can't be restored, so keep the original files on a drive or in cloud storage, and treat compressed copies as sharing versions. If you need a different size later, compress again from the original, never from a compressed copy.
What doesn't work
Zipping a video saves almost nothing, because it's already compressed. Converting to GIF makes files much larger. Repeatedly compressing the same file degrades quality each time, so always compress from the original. And no tool can restore detail once it's been compressed away.
Frequently asked questions
Can you really compress without losing quality?
- Without visible loss, yes. Mathematically every re-encode loses something, but at CRF 18–23 it's invisible on normal screens.
What's the best format for small files?
- MP4 with H.264 for compatibility, or H.265 if all your devices support it.
Does lowering resolution always help?
- Yes for size. Only do it if viewers won't watch on large screens.
Is my video uploaded?
- No. Compression runs in your browser.
Why is my compressed file bigger?
- The source was already efficiently compressed. Choose a smaller preset or lower resolution.
Should I use 30 or 60 fps?
- 30 fps for most content; 60 fps for fast action.
Is H.264 or H.265 smaller?
- H.265 is about 40–50% smaller at the same quality, but H.264 plays on more devices.
What CRF should I use?
- CRF 20–23 for H.264 is a good balance of size and quality.
Is HEVC worth it for sharing?
- Only if recipients' devices support it; otherwise use H.264.
Does removing audio help much?
- A little. Audio is usually 5–10% of a video file, so muting helps most on very tight size limits.