Does converting to SRT lose my styling?

Yes, and silently. SRT has no styling model at all — bold, colours, fonts, positions and karaoke are not representable. Tools either drop the tags or copy them across as literal text, which then shows up on screen as backslash codes. If the styling is part of the deliverable, do not leave ASS or TTML; deliver those formats and let the receiver transcode.

Is one of these formats a standard and the others not?

Roughly. SRT has no published specification — it is a de-facto format defined by what players accept, which is why player behaviour differs at the edges. WebVTT and TTML are W3C specifications, IMSC is a TTML profile built for streaming deliveries, ASS/SSA comes from the SubStation Alpha lineage with no standards body, and iTunes iTT is Apple's TTML subset for HLS. The practical rule: the more formal the format, the less variance between tools — and the more setup it needs.

SRT vs VTT vs ASS vs TTML — what each format actually supports

Q: Can I just rename a .srt file to .vtt?

No. WebVTT requires a WEBVTT header line, turns the comma in the timestamp into a dot, and allows the cue number to be dropped (it becomes an optional identifier). A renamed SRT file is missing the header, so a spec-compliant player rejects the whole file — the failure is total, not partial. Our SRT to VTT converter does those three changes and keeps the timings byte-identical.

Q: Which format does a browser play natively?

Only WebVTT. An HTML5 track element points at a .vtt file; SRT, ASS/SSA, TTML and iTT all have to be converted to WebVTT before a browser will display them, regardless of what local desktop players accept.

Most "which subtitle format" answers list the file extensions and stop there. The differences that actually cost you a delivery are narrower: whether a browser will play it at all, whether the styling survives a conversion, and whether the file declares its own encoding. This page puts the five formats you are likely to meet side by side, then lists the four things that break when you move between them.

Compiled from the specifications and from what our own parsers implement — the studio reads and writes SRT, WebVTT, ASS/SSA, TTML and iTT, which is why the awkward cases below are specific rather than general.

The comparison

Property	SRT	WebVTT	ASS / SSA	TTML (IMSC)	iTunes iTT
Where it comes from	De-facto format, no published specification	W3C specification	SubStation Alpha lineage, community-maintained	W3C specification; IMSC is its streaming profile	Apple's TTML subset for HLS
Timestamp syntax	`00:00:01,000 --> 00:00:04,000` comma before milliseconds	`00:00:01.000 --> 00:00:04.000` dot; `01:00.000` short form allowed	`0:00:01.00` centiseconds, one-digit hour	`begin="00:00:01.000"` attributes on elements	`begin="00:00:01:00"` frame-based
Cue numbering	Required integer above each cue	Optional — the line is a cue identifier	Inside `[Events]` rows, no numbering	Not applicable (XML tree)	Not applicable (XML tree)
Styling	None in the format. Some players honour `<b>` / `<i>` / `{\an8}` conventions; others print them literally	Limited: `<b> <i> <u> <c>` plus `::cue` CSS, which player support varies on	Full: fonts, colours, outlines, shadows, per-event overrides	Full: styles and regions defined once, referenced by cues	Limited to what Apple's profile allows
Positioning	Not in the format; only through player-specific tags	Cue settings: `line`, `position`, `align`, `size`	`\pos`, `\an`, `\move` against a PlayRes canvas	`<region>` with explicit extents	Very limited
Karaoke / per-word timing	No	Timestamps inside a cue's text; playback support varies	Yes — `\k` and friends are core to the format	Possible, via timing on spans	No
Encoding	Nothing declared — legacy files are often GBK, CP1252 or UTF-16 with no marker	Must be UTF-8	Usually UTF-8, historically legacy encodings too	XML declaration states it	XML declaration states it
Native in a browser	No — convert to WebVTT first	Yes — what `<track>` loads	No	No	No
Typical use	Quick delivery, anything that accepts "subtitles"	Web video, HTML5 players	Fansub and styled releases; local players	Broadcast and OTT deliveries	Apple HLS / iTunes packages

The four things that break on conversion

What happens	Why	What to do instead
Renaming `.srt` to `.vtt` produces a file that will not load	WebVTT requires the `WEBVTT` header and dot decimals; a renamed file has neither	Run a real conversion: header, dot decimals, cue numbers become identifiers — SRT to VTT converter
Styling disappears, or appears as literal backslash codes on screen	SRT has no styling model; converters either drop tags or copy them through as text	If styling is part of the deliverable, keep ASS or TTML and let the receiver transcode
Timings drift after a frame-rate change	Retiming is arithmetic on every timestamp, not a copy step: a 25 → 23.976 change costs about 9.95 seconds a minute if the file is not recomputed	Frame-rate converter, or one fixed offset when the whole file is simply late
Text renders as mojibake on someone else's machine	An SRT with no encoding declaration is decoded with whatever the receiving tool guesses	Convert to UTF-8 before sending; the studio detects GBK / UTF-16 / BOM on the way in and writes UTF-8 on the way out

Which one should you deliver?

Three questions decide it, in this order:

Did the receiver name a format? Then that is the answer — and if they named a TTML or IMSC profile, deliver that, not an SRT converted at the last minute.
Will it play in a browser? Only WebVTT will. If the answer has to be "it just works in a web player", the target is VTT.
Does anyone need the styling, positioning or karaoke? If yes, ASS (local playback, fansub workflow) or TTML/IMSC (delivery to a platform) — SRT cannot carry it, and VTT carries very little of it.

Everything else — plain text, one language, correct timings — is what SRT is for, and it is still the format that opens everywhere.

Converting a batch, not a file?

The studio opens SRT, WebVTT, ASS, TTML and iTT in one workspace, converts between them, and exports a batch as a ZIP with a per-file report. Single-file conversion stays free; Pro is a $9 one-time licence — lifetime, up to 3 devices, 7-day refund while the key is unactivated.

Open the studio

Quick answers

Can I just rename a .srt file to .vtt?

No. WebVTT requires a WEBVTT header line, the timestamp separator becomes a dot, and the cue number becomes an optional identifier. A renamed file is missing the header, so a compliant player rejects the whole file — the failure is total, not partial.

Which format does a browser play natively?

Only WebVTT. SRT, ASS/SSA, TTML and iTT all have to become WebVTT before a <track> element can show them, whatever local desktop players accept.

Does converting to SRT lose styling?

Yes, silently: SRT has no styling model, so colours, fonts, positions and karaoke are not representable. Tools either drop the tags or pass them through as literal text that then renders on screen.

Which of these is a standard?

WebVTT and TTML are W3C specifications (IMSC is TTML's streaming profile); SRT has no published specification and is defined by what players accept; ASS/SSA is community lineage; iTT is Apple's TTML subset. The more formal the format, the less the tools disagree — and the more setup it needs.