PRACTICAL GUIDES · Video and audio
How to make a video smaller without losing useful detail
A smaller video is useful only if people can still understand it. Start with the recipient’s limit and compare a real output rather than relying on a promised percentage reduction.
By Puffinmint · Updated
1. Define the limit and the details that matter
Note the current byte size, duration and destination limit. Small lettering matters in a lesson; moving details matter in a demonstration; clear audio matters in an interview.
Changing an extension or container does not guarantee fewer bytes. MOV and MP4 are containers, and their size also depends on the audio and video streams inside.
Trim video online2. Choose a representative test
Keep the original. Identify a section containing movement, small text and speech. A mostly static shot can look fine at settings that damage the rest.
Trim video can help create a sample, but trimming may itself re-encode the material. Use the sample to compare settings, then measure the complete output separately.
3. Adjust the compressor’s quality setting
Our compressor uses a quality parameter from 18 to 40, starting at 28. A higher number increases compression and accepts more visible differences. Change one setting at a time.
Output is MP4. The form has neither a target-size field nor a resolution selector. Its quality value is not a bitrate and cannot guarantee an exact number of megabytes.
TRY THIS STEP
Compress video
Use the full tool here. Load the sample or select your own file and adjust the settings.
A reproducible example
The sample is 4 seconds long at 640 × 360 pixels, with motion and a 440 Hz tone. Compress it and compare bytes, duration, motion and audio with the original. It is synthetic footage and does not represent every kind of video.
Synthetic sample · MP4 · 690,780 B
Our measured test
Test run on 3 November 2026 using the site’s conversion engine in an isolated Linux environment. These measurements apply to these samples; other files and engine versions can produce different results.
- Source
- 690,780 B
- Downloadable result
- 254,113 B
Settings: Compression CRF (18 = higher quality, 40 = smaller) 28
Downloadable result · MP4Result
Keep this page open. Larger files may take several minutes.
Choose files and options to prepare the download.
We use your files only for the result you request. No registration.
4. Example: estimate a budget before testing
Imagine a limit of 20 decimal MB for a 60-second video. The calculation gives a total average budget, some of which must cover audio and container overhead. It is a reference, not a tool setting.
One minute within a hypothetical limit
- Starting point
- 60 seconds; maximum 20 decimal MB
- Result
- An average total budget of about 2.67 Mbit/s
Calculation steps
Convert decimal megabytes to megabits.
Divide the available megabits by the duration in seconds.
Audio and container overhead use part of this budget. Measure actual bytes after compression instead of deriving a guaranteed size from the quality setting.
-
20 MB
Hypothetical total allowance.
-
60 seconds
Video duration.
-
About 2.67 Mbit/s
Combined average budget, subject to an output check.
5. Understand an unhelpful result
A source that is already efficiently encoded may not become smaller. Our tool reports when compression does not reduce size. Increasing compression further is useful only while the result remains fit for its purpose.
If the file remains too large, consider removing unnecessary footage. Do not assume the compressor offers pixel dimensions or guarantees unchanged visual quality.
| Operation | Effect |
|---|---|
| Compression | Re-encodes in an attempt to reduce bytes |
| Trimming | Reduces content and duration |
| Container change | Does not ensure a smaller file |
6. Verify the full exported video
Watch the beginning, middle, end and difficult scenes. Check duration, lip synchronization, lettering and rapid movement, then play the file on the recipient’s device or service.
Processing takes place on the server. Check the current limits and file handling information. Compare actual bytes with the requested limit, including whether it is expressed in MB or MiB.
Puffinmint created these samples on 3 November 2026. Source file data is measured; output is produced when you run the tool. Final size and accuracy are not guaranteed. Tick the checks after opening and comparing the result.
Sources and method
Frequently asked questions
Can I compress without any loss?
Lossy re-encoding changes the data. The practical goal is suitable quality, checked on the scenes that matter.
Why do identical settings produce different sizes?
Duration, motion, noise and the source affect the amount of data required. This quality parameter does not impose a constant bitrate.