Captions and markers

Every export pass can write timed sidecar files alongside the audio. You wrote the script as part of creating the project, so the app already knows what was said in each paragraph - and it knows when each paragraph starts and ends because it just stitched them. That is enough to produce caption tracks and marker files automatically, without any extra work from you.

Where to find these options

On the Merge and Export screen, the Export options block has four checkboxes:

• SRT
• WebVTT
• Premiere/Resolve CSV markers
• Audacity label track

Tick whichever ones you want. When you click Export, every ticked sidecar is written into the same folder as the audio file, named after the same stem.

The Export options block showing the SRT / WebVTT / Premiere CSV / Audacity labels checkboxes.

SRT subtitles (.srt)

The classic subtitle format. Plain text, numbered cues with HH:MM:SS,mmm timestamps. Compatible with practically every video player and every video editing tool ever made.

Each paragraph becomes one or more cues. Short paragraphs are one cue; long paragraphs are split into multiple sub-cues so no single line on screen is unreasonably long. The split is evenly spread across the duration of the paragraph - good enough for paragraph-level reading, not frame-accurate within a sentence.

When to use SRT

• YouTube uploads (subtitles can be loaded directly).
• Premiere, Resolve, Final Cut - all of them can import .srt as a caption track.
• Any web video player that supports embedded subtitles.

Filename

If your audio is saved as episode-01.wav, the SRT is saved as episode-01.srt in the same folder.

WebVTT subtitles (.vtt)

The newer subtitle format used on the web. Same cue structure as SRT but with a slightly different syntax (HH:MM:SS.mmm timestamps, a WEBVTT header, no cue numbers required). The sub-cue splitting rules match SRT.

When to use WebVTT

• HTML5 video on a website (<track kind="subtitles" srclang="en" src="...">).
• Any tool that requires VTT specifically.

Most modern video editing tools accept either SRT or VTT - pick whichever your downstream pipeline asks for.

Filename

If your audio is saved as episode-01.wav, the VTT is saved as episode-01.vtt in the same folder.

Premiere/Resolve CSV markers (.markers.csv)

A CSV file with one row per paragraph, each row a timed marker at the paragraph's start time. The full paragraph text goes in the marker description.

Import the CSV into Adobe Premiere Pro or DaVinci Resolve via their respective "Import markers from CSV" features. You will see one marker on the timeline at the start of every paragraph, labeled with the paragraph text. This is exactly the right shape for paragraph-level navigation while editing.

When to use Premiere/Resolve CSV markers

• Editing a long voice-over against video footage and you want to navigate by paragraph.
• Lining up B-roll cuts to specific paragraph beats.
• Building chapter timestamps for a YouTube description.

Filename

If your audio is saved as episode-01.wav, the markers are saved as episode-01.markers.csv in the same folder. The doubled extension is just a visual nudge ("this is a CSV of markers") - tools parse the contents, not the filename.

Audacity label track (.labels.txt)

A tab-separated text file in the format Audacity uses for label tracks: start_seconds <TAB> end_seconds <TAB> label_text. One label per paragraph, spanning the paragraph's full duration.

In Audacity, use File > Import > Labels... to load the file. You will get a labels track aligned exactly to the audio. This is invaluable for fine-grained per-paragraph editing in Audacity (apply effects to just one paragraph by selecting its label region).

When to use Audacity labels

• You are post-processing the merged audio in Audacity.
• You want to chop the merged file back into per-paragraph clips for separate handling.
• You are using Audacity for noise reduction and want labels for "where each line starts" so you can A/B specific takes.

Filename

If your audio is saved as episode-01.wav, the labels are saved as episode-01.labels.txt in the same folder.

How the timestamps are calculated

The app sums up each paragraph's effective duration in order. Effective duration means "after any crop has been applied". So:

• Paragraph 1 starts at 0.000 and ends at its (post-crop) length.
• Paragraph 2 starts where Paragraph 1 ended and ends one length later.
• And so on through the project.

The result is a clean timeline that matches the merged audio file frame for frame.

Behavior in split-export mode

When you turn OFF the "Merge all audio tracks into one" toggle, the audio is split into one numbered file per paragraph. The sidecars are still written, and they STILL use the merged-timeline timestamps (as if the paragraphs were going to be played back-to-back).

What happens if a sidecar fails to write

Sidecar failures are non-fatal. The audio export has already succeeded by the time the app writes the sidecars - you have your file. If something goes wrong with a specific sidecar (out of disk space, write permission, etc.), the status panel shows a message like (skipped: srt - write failed) and moves on. The other sidecars are still attempted.

Reference: sidecar formats