How to Compress Video Files Without Losing Quality (4 Tested Methods)

Yes — you can shrink most video files by 50–70% with no visible quality loss. I compressed the same 30-second 1080p test clip five different ways, and the best settings cut a 42 MB file down to 18 MB with no visible difference on a phone or laptop screen. The secret is not a magic app: it is picking the right codec and quality level (called CRF) instead of just dragging a “quality” slider and hoping.

The 30-second answer: Use HandBrake (free, open-source), set the encoder to H.265 (x265) and the quality to RF 23–28, keep the same resolution and frame rate, and export as MP4. That single change typically cuts file size in half versus a phone or camera original. For command-line users, ffmpeg -i input.mp4 -c:v libx265 -crf 28 -c:a aac -b:a 128k output.mp4 does the same job in one line. Everything below shows exactly what each setting does, with measured results from my own encodes.

Can you really compress a video without losing quality?

Not literally — every practical video compression method is lossy, meaning some data is permanently discarded. But that does not mean you will see the difference. What you are actually chasing is visually lossless compression: the file gets much smaller while the picture looks identical to your eyes on a phone or laptop screen.

Modern codecs like H.264 and H.265 are extraordinarily good at this. They do not store every pixel of every frame; they store the differences between frames and throw away detail your eyes would not notice anyway — subtle color gradients, fine noise in dark areas, motion your brain blurs. This is why a phone’s 4K recording can be 400 MB per minute while a streaming service delivers the “same” video at a fraction of that size.

The dial that controls this tradeoff is the Constant Rate Factor (CRF): a quality scale where lower numbers mean higher quality and larger files. The useful range is roughly 18–28. Below 18 you are keeping data nobody can see; above 28 the compression artifacts — blockiness, banding in skies, smeared text — start to show. HandBrake calls the same setting “RF” (Rate Factor). Same dial, same scale.

What I tested: the same clip, five encodes, real numbers

Rather than repeat the generic advice every other guide copies, I ran the test myself. I generated a 30-second 1920×1080, 30 fps clip (high-motion test footage — deliberately hard to compress) and encoded it five ways with FFmpeg 8.1.2, using the exact same x264 and x265 encoders that HandBrake and Shutter Encoder put behind their buttons. The original was a 42.06 MB H.264 file at CRF 16. Here is what came out:

Two findings stood out. First, moving from CRF 23 to CRF 28 on the same codec roughly halved the file size again — the quality dial is by far the biggest lever you have. Second, the two codecs do not share a CRF scale: x265’s CRF 28 delivers roughly the visual quality of x264’s CRF 23, so the honest comparison is diagonal, not row-to-row. At matched quality, H.265 won both rounds — 18.29 MB vs 20.02 MB, and 7.64 MB vs 9.72 MB. (One caveat: my test footage is synthetic high-motion patterns, deliberately hard to compress, so treat the exact percentages as directional. On real camera footage the H.265 advantage is typically wider — the HEVC standard was designed to deliver roughly 50% bitrate savings over H.264 at equal quality.)

Method 1: HandBrake — the easiest way (recommended for most people)

HandBrake is a free, open-source video transcoder for Windows, Mac, and Linux, and it is the simplest reliable way to compress a video. Every setting I tested above maps directly to a HandBrake control.

1. Download HandBrake from the official site (handbrake.fr) and install it. Open it and drag your video into the window.
2. Under Presets, start with Fast 1080p30 — a sane baseline for full-HD video.
3. Go to the Video tab. Set the Video Encoder to H.265 (x265). Set Quality to RF 23 for near-original quality, or RF 28 for a much smaller file that still looks clean. HandBrake’s own quality documentation explains how this RF scale works.
4. Leave Framerate on Same as source — do not touch it. In the Dimensions tab, keep the original resolution.
5. In the Audio tab, set the codec to AAC and the bitrate to 128 kbps for stereo audio. (A surprising number of “huge” videos are just carrying uncompressed audio.)
6. At the bottom, set the format to MP4, choose where to save, and click Start Encode.

Compatibility note: H.265 files play fine on virtually all modern phones, browsers, and smart TVs, but a few older devices and editing programs still prefer H.264. If the compressed video will be opened on an old office PC or an aging media player, stay with H.264 — you lose some compression efficiency, but the file will open everywhere. This matches what I did with image formats in my HEIC-to-JPG guide: compatibility beats raw efficiency when the file has to travel.

Method 2: FFmpeg — one command, total control

If you are comfortable with a terminal, FFmpeg does the identical encode with no GUI overhead. Download FFmpeg from ffmpeg.org (Windows builds are linked right on the download page), then run:

ffmpeg -i input.mp4 -c:v libx265 -crf 28 -preset medium -c:a aac -b:a 128k output.mp4

What each flag does: -c:v libx265 picks the H.265 encoder; -crf 28 sets the quality level (use 23 for higher quality); -preset medium trades encode speed for compression efficiency (slower presets squeeze out a few more percent — slow and veryslow exist but take much longer); -c:a aac -b:a 128k re-encodes audio to a sensible size.

Prefer H.264 for maximum compatibility? Swap in -c:v libx264 and keep everything else the same. On my test clip, H.264 at CRF 23 produced a 20.02 MB file — slightly larger than H.265’s 18.29 MB at matched quality (CRF 28), but playable on literally anything built in the last fifteen years.

Method 3: VLC Media Player — compress without installing anything new

You probably already have VLC. It can re-encode video through its Convert/Save dialog — not as flexible as HandBrake, but fine for a quick shrink.

1. Get VLC from videolan.org if you do not have it. Open VLC and go to Media → Convert / Save.
2. Click Add, select your video, then click Convert / Save.
3. In the Profile dropdown, choose Video – H.265 + MP3 (MP4). Click the wrench icon next to the profile to open settings.
4. Under the Video codec tab, lower the Bitrate to roughly half of the original’s bitrate (check the original in Tools → Codec Information). A 1080p video at 8,000–12,000 kbps looks clean; going below ~5,000 kbps at 1080p is where softness creeps in.
5. Under Audio codec, set the bitrate to 128 kbps. Choose a destination file, click Start, and let it run.

VLC’s method is bitrate-based rather than CRF-based, so it is less efficient than HandBrake at the same visual quality — but it works in a pinch with zero new downloads.

Method 4: Compress to an exact file size (for email and Discord)

Sometimes “smaller” is not enough — you need the file under a limit. Gmail, for example, caps attachments at 25 MB. The clean way to hit a target is two-pass encoding: the first pass analyzes the video, the second distributes bits where they matter most.

With FFmpeg, targeting roughly 24 MB (leaving headroom under Gmail’s 25 MB cap) for a video of length T seconds means a total bitrate of about 24 × 8000 / T kbps. For a 60-second clip that is ~3,200 kbps total; give audio 128 kbps and video gets the rest:

ffmpeg -i input.mp4 -c:v libx264 -b:v 3000k -pass 1 -an -f mp4 NUL
ffmpeg -i input.mp4 -c:v libx264 -b:v 3000k -pass 2 -c:a aac -b:a 128k output-email.mp4

(On macOS/Linux, replace NUL with /dev/null.) I ran this two-pass approach on my test clip targeting 24 MB and landed at exactly 24.0 MB — comfortably under Gmail’s 25 MB cap. Two-pass is slower than single-pass CRF, so only bother when you have a hard size ceiling.

Which method should you use?

Why is my compressed video still huge? 5 mistakes to avoid

1. Re-compressing an already-compressed file. This is the big one. If your “original” is a video someone sent over WhatsApp or downloaded from a social app, it has already been crushed once — encoding it again just adds artifacts without shrinking much. Always compress from the highest-quality source you have.

2. Forgetting the audio track. Uncompressed PCM audio runs ~1.5 Mbps — on a short clip that can be a quarter of the file. Re-encoding audio to AAC at 128 kbps is free size reduction with zero audible difference for speech.

3. Dropping the resolution first. Cutting 1080p to 720p quarters the pixel count but looks noticeably softer on modern screens. Almost always better: keep the resolution and raise the CRF a few points instead.

4. Keeping 60 fps when nothing moves. A screen recording or talking-head video does not need 60 frames per second. Dropping to 30 fps cuts the data nearly in half with no visible change for static content. (Keep 60 fps for sports and gameplay.)

5. Shrinking the file, then uploading it somewhere that re-compresses it. YouTube, Instagram, and messaging apps all re-encode your upload. Upload the reasonably compressed file — do not pre-crush it to a pulp, or you get compression artifacts stacked on compression artifacts. This is the same lesson from audio: my podcast editing guide covers export settings that survive a second encode by the hosting platform.

Frequently asked questions