Free online DFXP converter. Convert DFXP to SRT, VTT, ASS, SSA, SUB and more online — no limits, no watermark.
.dfxp caption file or click "Upload .dfxp". Each file is read as XML timed text, so the timing and cue text are parsed before anything is rewritten.<track> captions on the open webDFXP stands for Distribution Format Exchange Profile. It is not a separate format so much as the original name for TTML (Timed Text Markup Language), the W3C's XML caption standard — the spec was first published as a W3C Recommendation on 18 November 2010, and the "DFXP" label was kept for the profile designators inside it. A .dfxp file is the same XML you would find in a .ttml file, just carrying the legacy extension that Adobe Flash players, Premiere exports, and some broadcast and OTT captioning pipelines emit.
The problem is that almost nothing outside those legacy workflows wants a raw .dfxp. Flash is gone, and most modern players, editors, and upload forms expect a simpler subtitle file. Converting away from DFXP is usually about getting the captions into something the destination actually reads:
<track> element reads, so if you are embedding a <video> on a page, VTT is the file the browser expects..ttml simply puts the file under the name modern tools recognize, without throwing away the styling and positioning the markup already carries.One thing to know: DFXP can store styling that simpler formats cannot. Color, top/bottom positioning, italics, and frame-based timecodes (HH:MM:SS:FF) all live in the XML. When you convert to a plain-text target like SRT or TXT, that styling is flattened to text and timing — the words and sync survive, but bespoke fonts and positions do not carry over.
| Format | Type | Styling support | Best for |
|---|---|---|---|
| DFXP / TTML | XML (W3C timed text) | Color, position, italics, frame timecodes | Flash legacy, broadcast/OTT exchange |
| SRT (SubRip) | Plain text | None (italics only, ad hoc) | Universal playback, uploads, editors |
| VTT (WebVTT) | Plain text (W3C) | Cue settings, basic styling | HTML5 <track>, web video |
| ASS / SSA | Plain text (styled) | Full fonts, colors, positioning | Media players, fansubs, karaoke |
| SBV (SubViewer) | Plain text | None | YouTube Studio captions |
| TXT | Plain text | None (no timing) | Transcripts, proofreading |
| Property | Value |
|---|---|
| Full name | Distribution Format Exchange Profile |
| Modern name | TTML — Timed Text Markup Language |
| Standards body | W3C |
| First Recommendation | TTML 1.0, 18 November 2010 |
| Syntax | XML — a <tt> root wrapping <head> (styling and regions) and <body> → <div> → <p> cues |
| Timing attributes | begin, end, and dur on cue elements |
| Timecode forms | Clock time (00:01:23.400) and frame-based media time (00:01:23:12) |
| Styling | tts: attributes — color, background, font family and size, text alignment |
| Positioning | <region> elements declared in the head and referenced by cues |
| Language tagging | xml:lang, which any element can carry |
| Common extensions | .dfxp, .ttml, .xml |
| Downstream profiles | IMSC (streaming), SMPTE-TT and EBU-TT (broadcast) |
| Best for | Caption exchange between production systems — not direct playback |
Because it is XML, a .dfxp is verbose compared with a plain-text caption file: a track that is a few kilobytes as SRT routinely runs several times that as DFXP, since every cue carries its own element, attributes, and region reference.
Effectively, yes. DFXP — Distribution Format Exchange Profile — was the original name for what the W3C now calls TTML (Timed Text Markup Language). The W3C spec itself notes that earlier drafts referred to TTML as DFXP, and the name was retained for the profile designators. So a .dfxp and a .ttml file are the same XML caption format under two different extensions; converting DFXP to TTML mostly just relabels the file into the name modern tools expect.
You lose the visual styling, but not the captions. DFXP's XML can carry colors, on-screen positioning, italics, and frame-based timecodes; SRT is plain text and stores only the cue text and start/end times. The conversion keeps every line and every timestamp intact, but flattens custom fonts, colors, and positions to ordinary text. If preserving styling matters, convert to ASS/SSA or stay in TTML instead.
Because DFXP is just XML, any text editor will open and show the raw markup. To actually display it as captions you generally need a player or tool that speaks timed text — legacy Flash players, some broadcast and OTT subtitling systems, or editors that import TTML. For everyday use it is far easier to convert the DFXP to SRT or VTT, which open in VLC, YouTube, web players, and standard subtitle editors without special support.
SRT is the most reliable choice for YouTube, and SBV is the platform's own native plain-text caption format — YouTube Studio accepts both. Either keeps your timing and wording from the DFXP while dropping the XML that the upload form does not need. If you also publish the video on your own website with an HTML5 player, convert a copy to VTT for the <track> element.
Yes — timing is the part that always carries across. In our testing, a DFXP caption file using HH:MM:SS:FF frame timecodes converted to SRT preserved every cue's start and end time after the converter translated frames to the millisecond timestamps SRT uses, with cue order and line breaks intact. The text and sync are the priority of the conversion; only format-specific styling is dropped when the target cannot represent it.
Yes. The conversion runs entirely in your browser. It is never shared, published, or made public, and there is no sign-up or watermark. Subtitle files are small, so conversions finish almost instantly, whether you convert one file or several.
IMSC — Internet Media Subtitles and Captions — is a W3C profile of TTML: a constrained subset with a published feature list, defined so that streaming platforms can exchange captions without every player having to implement the whole of TTML. So an IMSC document is a TTML document, and a .dfxp file is TTML under an older extension; what differs is how much of the spec each is permitted to use. In practice a modern OTT or streaming delivery spec will ask for IMSC, while an older Flash-era workflow or a Premiere export hands you DFXP — and both parse here as the same XML timed text.
For anything new, write the .ttml extension, or whichever profile your distributor specifies. The .dfxp name belongs to the era when Adobe Flash players and Premiere exports were the main consumers of timed text, and Flash reached end of life on 31 December 2020. Nothing about the XML itself changed — the markup you already have is still valid TTML — so converting DFXP to TTML is usually just a relabel that stops modern tooling from rejecting the file on its extension alone. Keep the .dfxp name only where a specific legacy captioning system explicitly demands it.