Find where a video became blurry
Compare the source, export and published stream instead of guessing from one playback.
3 min read · Updated November 3, 2026For a video that looks softer than expected before or after posting.
Use one scene to compare three stages
Pick a moment containing small text, hair, fabric or movement. Compare the camera source, the editor export and the published playback at roughly the same display size. An enlarged screenshot can look softer simply because it stretches fewer pixels over more screen area.
Make a note of which file or player you are looking at. A timeline proxy, a low-resolution preview and a completed export are not interchangeable evidence. If the first degraded stage is the export, changing the upload route cannot recreate the original picture.
- 01Source looks right?
- 02Export looks right?
- 03Which published stream?
Choose the branch that matches the symptom
Use these checks before changing several settings at once. They are diagnostic paths, not promises about a platform’s internal decisions.
The source itself is soft
- Check
- Inspect focus, motion blur, low-light noise and any earlier compression.
- Next step
- Use a better source if available. Increasing output resolution will not reconstruct captured detail.
Source is sharp; export is soft
- Check
- Compare crop, output dimensions, quality setting and whether you reused a compressed export.
- Next step
- Return to the editable project and make a controlled export comparison.
Export is sharp; published playback is soft
- Check
- Identify the post, viewer context and the stream reported by Check video.
- Next step
- Compare exact dimensions and evidence labels. Keep unavailable fields unknown.
The picture is washed out rather than smaller
- Check
- Check color format and whether the player/display supports it.
- Next step
- Follow the color guide before treating it as a bitrate problem.
Test an export change in isolation
Choose a short section with the visible problem. Export it twice with the same dimensions, frame rate and codec, changing only the quality control. If the second file reduces blocking or smeared texture, the encoder’s data budget may have contributed. If both files are equally out of focus, more bitrate is not the missing fix.
For a framing problem, compare the crop instead. A 9:16 crop taken from a wide shot may use only part of its original pixels. A sharp full-width source is not proof that a tight portrait crop contains enough detail for the intended display.
Read the checker without mixing sources
Copy the published post link into Check video. Record the checked-at time and the row you are comparing: Browser, Embed or Mobile web. Exact dimensions describe a stream, while an advertised-renditions list describes available metadata.
For example, a Browser row at 576 × 1024 and a Mobile web row at 720 × 1280 are two different observations. They do not establish that every desktop viewer sees the first or every TikTok app user sees the second. Compare codec and bitrate only within each row.
If a stream is blocked or the media headers cannot be inspected, the checker may show metadata only or unavailable. That failure cannot be repaired by borrowing values from another stream.
Do not turn a small sample into an upload rule
A single successful-looking post does not prove a permanent compression bypass. Matching dimensions do not prove an unchanged encode. A smaller exposed rendition does not prove an account penalty. Keep the conclusion as narrow as the evidence supports.
For a useful posting comparison, keep the original export fixed, record the destination and repeat the check with the timestamp and source labels. If you change editor settings, posting route and account together, the result cannot isolate which change mattered.
- Same source scene and comparable display size.
- Known source and export filenames.
- Public post link and checked-at time recorded.
- Stream identities and evidence levels kept with the numbers.
- One variable changed per comparison.