TikTok safe zone for subtitles
Two regions loaded, one deliberately left out — and why.
What can overlap a subtitle
TikTok overlays a translucent tab row (Following / For You / search) near the top of the video and a vertical action rail (avatar, follow, like, comment, share) down the right edge. Author name, caption text, and the audio ticker sit lower on the left. A subtitle placed high can run into the tab row; one placed toward the right edge can catch the action rail.
What the tool loads for TikTok
In Reference · Organic mode, the active TikTok reference is a first-party capture: a physical Galaxy S20 Ultra, TikTok app 47.0.3, Korean UI, captured 2026-09-23. Of 19 screenshots reviewed, 5 qualified as a plain organic For You feed (standard tabs, no ads, no search or event badges, no carousel, collapsed 2-line caption) and became the base measurement. Confidence is recorded as medium-high.
- Top nav overlay — the tab row and search icon. Unlike the opaque bar above the video in the Reels capture, this row is translucent and drawn on top of the video — the video shows through the tab text — so it does cover part of the frame and is included as a region.
- Action rail — avatar through share icon, cut off where the video frame ends at the bottom.
No caption-row region is loaded, on purpose. In this capture, the author name / caption / audio-ticker block started only about 11 pixels above where the video frame actually ends — essentially all of it renders below the video, in the margin. That's not enough overlap to justify a region, so none was added. This is a finding scoped to this one device, app version, and 5-sample capture — it is not a claim that TikTok captions are always clear of the video.
Recorded regions
| Region | x | y | Width | Height |
|---|---|---|---|---|
| Top tabs and search icon (see-through, over the video) | 0 | 5 | 1080 | 120 |
| Right action rail (part over the video) | 936 | 1223 | 144 | 697 |
Source: first-party measurement · confidence medium-high · last verified 2026-09-23 (device capture date)
Values exactly as recorded, with no safety margin added and no averaging across sources. Only these regions are checked. Real UI changes with device, app version and caption length.
A note on measurement disagreement
An older, independent measurement of TikTok's action rail (from a different, web-player-based source) reported different coordinates than the first-party capture above. Neither replaces the other automatically — they're kept as separate evidence, and only the first-party one is active in production, because it meets a higher bar (physical device, repeated samples, documented method). The disagreement itself is a useful reminder that a single source, however careful, isn't the last word.
Using it
Switch Data to Reference · Organic, place your Subtitle, and check the TikTok column. Because there's no caption-row region, a subtitle placed where TikTok's own caption normally sits won't be flagged by this tool for that alone — that's a gap in current evidence, not a confirmation that space is clear.
Limitations
- Single device, single session, Korean UI, one app version.
- Caption was 2 lines, collapsed, in every sample; 1-line and expanded ("see more") captions weren't captured.
- The video frame's bottom edge was a soft gradient in this capture, not a hard line — a conservative boundary was chosen rather than an exact one.