Subtitle Statistics Tool

Subtitle Statistics Tool reads a SubRip file pasted into the SRT Content box and returns a five-line readout: total cues, total duration in seconds, the time text is actually on screen with a coverage percentage, the average reading speed in characters per second, and the total character count. The report is plain text, easy to paste into an email or QC sheet and downloadable as a .txt file.

These numbers answer the questions that surround every delivery. A vendor pricing a translation needs a character count. A reviewer wants to know whether pacing is comfortable or relentless. A compliance check asks whether captions actually cover the programme or whether a reel is missing. Judging that by eye across a thousand cues is impractical; one readout settles it.

The calculation works from the parsed cue list. Total duration spans the first cue's start to the last cue's end, which is the subtitled portion rather than the video runtime. On-screen time sums every cue's individual duration, and coverage expresses that as a percentage of the span, so a low figure means long silent stretches and a figure near 100 percent means near-continuous captions. Average reading speed divides characters by on-screen seconds, and the character count excludes whitespace so it stays comparable with the 17 to 21 cps thresholds style guides use.

The analysis runs in your browser on the pasted text with no upload, so a long feature file is measured instantly and confidential material stays put.

• Five metrics in one pass: A single run of the SRT Content box gives cue count, total duration, on-screen time, coverage percentage and average reading speed, replacing several separate checks.
• Coverage percentage: On-screen caption time is compared against the span from the first cue to the last, which immediately exposes gaps where a section of the programme was never subtitled.
• Whitespace-free character count: Spaces and line breaks are excluded from the character total, so the figure reflects actual readable content and matches how reading-speed limits are normally defined.
• Average reading speed in cps: Total characters divided by total on-screen seconds gives a file-wide pacing figure you can compare directly against the 17 cps comfort target and the 21 cps upper limit common in style guides.
• Duration measured from the cues: Total duration spans the first cue's start to the last cue's end, giving an honest picture of the subtitled portion even when you do not have the video to hand.
• Plain-text, shareable output: The readout is a short block of labelled lines that pastes cleanly into a delivery note, a ticket or a spreadsheet cell without reformatting.
• Clear handling of empty input: If nothing is pasted, or the text contains no parseable cues, the tool says so rather than reporting misleading zeros.
• Instant and private: Analysis happens in the page with no server round trip, so nothing about the file is uploaded, logged or retained.

1. Open the subtitle file you want to measure, select all of its contents and paste them into the SRT Content box.
2. If you would rather see the format of the report first, click Load sample to analyse a short built-in example.
3. Click Run and read the five reported lines in the result panel.
4. Check the coverage percentage first, since an unexpectedly low figure usually points to a missing section or a long untranslated stretch rather than a stylistic choice.
5. Compare the average reading speed against your delivery specification, treating anything much above 21 cps as a sign that cues need lengthening or text needs condensing.
6. Use the total character count for translation quoting or for comparing the source file against a completed target-language version.
7. Copy the report into your QC notes, or use the download button to save it as a .txt file attached to the delivery.

Example 1 - a healthy interview file. Captions run almost continuously and pacing sits comfortably below the usual limit.

Total cues: 412
Total duration: 1784.6s
Text shown: 1602.3s (90% coverage)
Avg reading speed: 15.8 cps
Total characters: 25316

Example 2 - a file with a problem. The same runtime reports much lower coverage and a fast average, which together suggest a missing section and cues that are too short for their text.

Total cues: 301
Total duration: 1781.2s
Text shown: 1015.9s (57% coverage)
Avg reading speed: 23.4 cps
Total characters: 23772

Comparing the two makes the diagnosis obvious: roughly the same amount of text is being squeezed into far less on-screen time, so the file needs retiming before it can be delivered.

• Spot missing subtitles instantly: A coverage percentage well below what the content warrants flags an untranslated reel or a dropped section long before a viewer or a client finds it.
• Price translation accurately: A whitespace-free character count gives vendors and freelancers a solid basis for per-character quoting instead of an estimate from runtime alone.
• Evidence for QC sign-off: A short labelled report can be attached to a delivery as documented proof that pacing and coverage were checked, which is often required by broadcasters.
• Catch pacing problems early: Seeing a file-wide average above the comfortable reading range tells you to fix timing while the file is still in your hands rather than after rejection.
• Compare source and target files: Running the same analysis on both language versions quickly reveals when a translation has expanded far beyond what the original timings can carry.
• Confidential by default: The file is analysed in your browser and never uploaded, so scripts under embargo can be measured without any transfer.