Why a report can show 576p, 720p or 1080p
Read different dimensions without inventing a rule about the upload.
3 min read · Updated November 3, 2026For reports where the upload, browser result and mobile-web result show different sizes.
Start with the exact dimensions
Read width and height together. A portrait video at 576 × 1024 has 576 pixels across and 1024 down. Fulres can label its short side as 576p, but the dimensions underneath are the more precise description.
Labels in other tools may group that same frame into a nearby conventional quality tier such as 540p. A difference between those labels can be a naming choice rather than a different file. Compare exact dimensions before concluding that the measurements disagree.
| Frame dimensions | Short-side label | What the numbers say |
|---|---|---|
| 576 × 1024 | 576p | Portrait frame with a 576-pixel short side. |
| 720 × 1280 | 720p | Larger portrait frame; image detail still depends on the source. |
| 1920 × 1080 | 1080p | Landscape frame, despite sharing a tier with portrait 1080 × 1920. |
| 2160 × 3840 | 2160p | Portrait UHD-sized frame; no guarantee of delivered playback. |
Follow the file through each stage
A camera source, an editor timeline, an export and a published rendition can all have different dimensions. Find the earliest difference. If a 2160-pixel short-side capture was exported at 1080, preparation cannot restore the discarded samples.
A platform can also make alternative delivery files from one upload. Those files can have different dimensions, codecs and bitrates. Treat them as separate renditions instead of merging their best properties into one imaginary stream.
- 01Captured frame
- 02Timeline / crop
- 03Exported dimensions
- 04Prepared source
- 05Identified playback rendition
Available is different from inspected
A rendition list can advertise several options. That establishes that the response described those options, not that the browser fetched or decoded each one. Header inspection provides stronger evidence about an accessible file, provided the inspected file is the one named in the result.
If a fallback accesses a different file, its dimensions belong to that fallback. They cannot validate the inaccessible original. The report should keep that distinction even when the alternative produces more complete numbers.
The label differs, but width and height match
- Check
- Whether each tool uses a conventional tier or the exact short side.
- Next step
- Describe it as a labeling difference unless other evidence identifies different media.
Browser and mobile-web dimensions differ
- Check
- The source, URL identity and inspection evidence for each row.
- Next step
- Retain both source-specific observations; do not call mobile web the native app.
A larger option appears only in a rendition list
- Check
- Whether that exact option was successfully inspected.
- Next step
- Keep it advertised or metadata-only until verified.
Resolution does not measure all visible quality
An enlarged soft image can have more pixels than a detailed smaller image. Focus, source noise, crop, encoding and display scaling affect appearance. Codec and bitrate can help explain a difference, but they are not substitutes for viewing the matching scene.
For a fair comparison, use the same frame, display size and playback surface. Do not compare a full-screen desktop still with a small phone preview and attribute the entire difference to resolution.
Write a conclusion the evidence supports
A useful conclusion names the scope: “The browser stream inspected at this time was 576 × 1024.” That is more precise than “TikTok only supports 576p,” which would make a much broader claim.
If you cannot inspect another surface, leave its result unavailable. You can repeat a check later, but preserve the time of any cached observation so an old result does not look newly measured.
- Exact dimensions accompany the short label.
- Orientation is accounted for.
- Each rendition retains its source and evidence level.
- No advertised option is presented as played media.