FPS subtitle converter
Subtitles that start almost right and finish several seconds off were timed for a different frame rate. Multiplying every cue by the ratio between the two rates (23.976 ⇄ 25, 25 → 29.97, 29.97 ⇄ 59.94, PAL speed-up and back) puts it back on the picture — inside this browser tab.
Check which problem you have — 60 seconds, no upload
At 0:00
Already wrong on the very first line? A constant error from second one is a fixed offset, not drift — use shift timestamps. Within ~100 ms? Keep going.
At 30:00
Note how far off it is and in which direction. Cues running ahead of the picture: the video rate is lower than the subtitle rate. Behind: higher.
At 60:00
Roughly double the 30-minute error → linear drift, this page fixes it. About the same → still an offset (or a bad mux), this page will not help.
Drift per minute ≈ error at 60 min ÷ 60. Match that number against the table below: the row it lands on tells you exactly which two frame rates to pick, without guessing.
Drop an .srt / .vtt / .ass file here
or click to choose one · nothing leaves this page
Before / after
Nothing is uploaded — not even to “have a look”. The file is parsed and exported in this tab, so there is no sample for us to receive; the 0 / 30 / 60 check above is the whole diagnostic. Some online converters ask you to send your input and output files so they can investigate — this one has nothing to send.
Frame durations
| Frame rate | One frame | Where it shows up |
|---|---|---|
| 23.976 fps | 41.708 ms | Film on Blu-ray / streaming (24p with pulldown) |
| 24 fps | 41.667 ms | Film, cinema, most festivals |
| 25 fps | 40.000 ms | PAL regions, European broadcast, DVD/Blu-ray in PAL |
| 29.97 fps | 33.367 ms | NTSC broadcast, US cable, many deliveries |
| 30 fps | 33.333 ms | Screen recordings, some web delivery |
| 59.94 fps | 16.683 ms | High frame-rate sport and gameplay capture |
How far off does each combination end up?
| Subtitles timed for → video plays at | Timing factor (source ÷ target) | Drift per minute | After 60 min |
|---|---|---|---|
| 23.976 → 25 | 0.959040 | −2.46 s | −147 s |
| 25 → 23.976 | 1.042709 | +2.56 s | +154 s |
| 24 → 25 | 0.960000 | −2.40 s | −144 s |
| 25 → 24 | 1.041667 | +2.50 s | +150 s |
| 23.976 → 24 | 0.999000 | −60 ms | −3.6 s |
| 29.97 → 30 | 0.999000 | −60 ms | −3.6 s |
| 30 → 29.97 | 1.001001 | +60 ms | +3.6 s |
| 25 → 29.97 | 0.834168 | −9.95 s | −597 s |
| 23.976 → 29.97 | 0.800000 | −12.00 s | −720 s |
| 29.97 → 59.94 | 0.500000 | −30.00 s | −1800 s |
| 30 → 59.94 | 0.500501 | −29.97 s | −1798 s |
| 59.94 → 29.97 | 2.000000 | +60.00 s | +3600 s |
| 59.94 → 30 | 1.998002 | +59.88 s | +3593 s |
Negative drift = the cues have to move earlier (the video plays faster than the file assumes). Positive = they have to move later. Anything above roughly 100 ms per minute cannot be hidden with a fixed offset, which is why this page multiplies instead of shifting.
About 59.94: it is double 29.97, so a 59.94 ↔ 29.97 mix is a factor of two — the cues end up at half or double time, not subtly off. In practice a file “timed for 59.94” played on a 29.97 timeline is the rare case; the 23.976 ↔ 24 and 29.97 ↔ 30 rows are the ones people actually need.
Retiming a whole series?
The studio re-times every open file at once, exports them as one ZIP and can hand you a CSV/JSON timing report to prove the result. Pro is a $9 one-time licence — lifetime, up to 3 devices — with a 7-day refund while the key is unactivated. Single-file retiming stays free, including export.
Questions people ask before using it
Why do subtitles drift more the longer the video plays?
The file was timed against a different frame rate than the video you are watching. Every cue is multiplied by the ratio between the two rates, so the error accumulates — 25 fps subtitles on 29.97 fps footage drift about 9.9 seconds per minute of content, and every cue has to move earlier by that much.
Which frame rate do my subtitles belong to?
Whatever the cut they were authored against used. Watch the direction of the error: cues that run early and get worse mean the video rate is lower than the subtitle rate; cues that fall behind mean it is higher. The table above gives the expected drift for each combination, so you can match it against what you see.
Is this the same as a PAL speed-up fix?
Yes. 25 → 23.976 (or the reverse) is the classic PAL/NTSC case: about 4 percent fast or slow, which is 2.5 seconds of drift per minute — far too much for a fixed offset to fix.
How do I tell drift from a fixed offset before re-timing the whole file?
Check three points in the player you actually use: the first line (0:00), around 30 minutes, and around 60 minutes. If the error at 60 is roughly twice the error at 30, it is linear drift and this page is the right tool — divide the 60-minute error by 60 to get the drift per minute and match it against the table above. If the error is the same at all three points, it is a fixed offset: shift the whole file instead. If the error only appears after the player seeks or buffers, it is the player's own rendering, not the file.
My file is 59.94 fps — which conversion do I need?
Pick by direction, not by the label. 59.94 is double 29.97, so 59.94 ↔ 29.97 changes every cue's timing by a factor of two (drift 30–60 seconds per minute) — that is a whole-file speed change, not a subtle offset. Most “59.94” problems are really a 29.97 ↔ 30 file on the wrong timeline; start with the 30 → 29.97 row (−/+60 ms per minute) and check again at the 30-minute mark.