Guide: convert a Premiere Pro XML to SRT

In Premiere Pro, a sequence typically starts at 01:00:00:00 (1 hour) — a professional convention. Your title clips are positioned from that point: a title 10 seconds in lives at 01:00:10:00. Checked (default): SRT timestamps reflect that exact timecode (01:00:10,000). Use this if your video file or broadcast system uses that timecode as the start — typical for a broadcast master or MXF. Unchecked: timestamps start at 00:00:00,000 regardless of the Premiere timecode. Right setting for an MP4/MKV exported for the web, YouTube or social media.

The SRT file is saved in UTF-8, the universal encoding that guarantees accents and special characters render correctly (é, è, à, ç, ñ, etc.). The BOM (Byte Order Mark) is a hidden 3-byte signature added at the very start of the file, which helps some software auto-detect this encoding. Unchecked (default): UTF-8 without BOM — compatible with YouTube, VLC, DaVinci Resolve, Premiere Pro and the vast majority of modern players. Checked: only enable if accents show up as garbled characters in a specific legacy Windows program.

No. The entire conversion runs in your browser, on your own machine. Neither the XML nor the generated SRT ever transit a server. Your internet connection is only used to load the page the first time.

This tool handles title clips placed directly on the timeline as if they were video clips, in both existing formats: legacy Titles (classic Text tool, plain text in the XML) AND Essential Graphics / modern Titles from Premiere Pro ≥ 2023 (text stored as a base64 FlatBuffer inside the filter parameters, decoded automatically). Subtitles managed via Premiere's Captions panel (native SRT, WebVTT, CEA-608…) use a different system — export those via File → Export → Captions.

Yes. The generated file is a standard UTF-8 SRT, recognized by YouTube Studio, VLC, DaVinci Resolve, Premiere Pro (SRT import), Final Cut Pro, connected TVs and all modern players on PC, Mac, iOS and Android.

This is almost always tied to the “Keep program TC” option. If your subtitles start at 01:00:00,000 while your video starts at 00:00:00 — the most common case — uncheck the option and re-run the conversion. Conversely, if subtitles appear too early, check it.

Drop several .xml files at once (multi-select or grouped drag & drop). Each file is converted independently with the same options, and a “Download all” button grabs every SRT in one click. Two equivalent standalone tools are also downloadable (single-file HTML and Python script) for offline or pipeline use — see the Downloads page.

Supported: every text line of a single title clip is read and joined inside a multi-line SRT block, in the order they appear in the XML. When several distinct title clips are truly simultaneous (full overlap), their texts are merged into a single SRT block — up to 4 clips. Above that, extra clips remain in separate SRT blocks and a warning states the exact count.

Select the sequence in the bin, then File > Export > Final Cut Pro XML to produce an .xml (xmeml) file. Drop that file here: the tool walks the video tracks, isolates title clips (legacy Titles and Essential Graphics), reads their text and timing, and produces a standard .srt. No audio clip, transition or video effect is needed — only titles are extracted. Handy when a project was subtitled with real graphic titles rather than with Premiere's Captions workflow.

Yes. Since Premiere Pro 2023, Essential Graphics text is stored as a base64-encoded FlatBuffer inside the XML. The tool automatically decodes those blocks to recover the readable text, without depending on Adobe. Legacy Titles (classic title designer) are supported too — detection is automatic, no Premiere-side setup required.