Muxsmith 1.2

This update adds local audio transcription, giving you a practical way to create, review, search, and export subtitles from your own media files.

Turn an audio track into subtitles

• Choose the exact audio track and spoken language before transcription.
• See language, codec, channel layout, channel count, and default-track metadata in the audio selector.
• Transcribe with Apple's on-device Speech framework. Muxsmith does not upload your audio or transcript.
• Review and edit the generated subtitle text while keeping its detected timing.

Review the result against the video

• Play the video beside a searchable transcript and jump directly to any cue.
• Adjust subtitle timing in small steps, then apply the same adjustment to other completed files.
• Use the vertical timeline to find conversations and long gaps in dialogue.
• Add, rename, move, or remove chapter markers before export.

Work through a batch

• Add multiple local or network files and follow inspection, extraction, and transcription progress for each one.
• Apply one spoken language to the batch while keeping the audio-track choice independent for every file.
• Search across completed transcripts and jump to a result in the relevant source.
• Cancel the active transcription or the remaining batch without discarding work that already finished.

Export subtitles or reviewed media

• Export a collision-safe SRT sidecar.
• Export a new MP4, M4V, MOV, MKV, or WebM file with the generated subtitles and reviewed chapters.
• Keep compatible video and audio streams without re-encoding.
• Batch-export selected completed files, with guarded source replacement available only after explicit confirmation.

Also improved

• English and unknown-language tracks now default to English (United States).
• Files added during transcription can be inspected immediately and added to the ordered queue.
• MKV sources can use a temporary compatible preview for subtitle-sync review while final export continues to read the original file.
• Transcription settings use a compact, consistently aligned macOS layout.

Muxsmith 1.2 requires macOS 14 or later. Transcription requires macOS 26 or later, an Apple-silicon Mac, and an available Apple Speech language model. macOS may download the selected language model from Apple, but Muxsmith does not send your media audio with that request.