FFmpeg Audio Encoder¶

A cross-platform desktop front end for FFmpeg and the audio encoders that surround it. Drag files in, inspect their audio streams, configure an encoder, and queue the work.

Version 5 is a rewrite. The Tkinter interface of v4 has been replaced by a PySide6 one built on a new codebase, with a durable encoding queue, crash-safe outputs, and adapters for external and licensed encoders alongside FFmpeg's own.

Install

Download a bundle for your platform and get past the first-launch warnings.

Set up your tools

FFmpeg is required. opusenc, qaac, fdkaac, and DeeZy are optional and discovered automatically.

Encode something

From dropped file to finished output.

Encoder reference

Every encoder and every option, generated from the source.

What it does¶

• Per-stream encoding. Inputs are probed with ffprobe and MediaInfo, every audio stream is listed with its codec, layout, and duration, and you pick which one to encode.
• Sixteen encoders in three families. Eight built into FFmpeg, three standalone command-line encoders, and five DeeZy adapters for licensed Dolby formats.
• A durable queue. Jobs survive a restart. Start selected jobs, stop after the current one, retry a failure, or cancel outright.
• Crash-safe outputs. Every encode writes to a job-specific .part file that is published only after the encoder exits successfully.
• Delay handling. Container delays are detected per track; filename markers such as [DELAY -21ms] are a fallback for single-track audio-only inputs.
• Presets, plus automatic restoration of the last encoder configuration you used.

What it does not do¶

FFmpeg is not bundled. Neither are the optional encoders. The application detects what is installed and disables what is missing rather than failing at encode time; see Required and optional tools.

The application does not transcode video, mux into video containers, or edit metadata. It encodes one audio stream to one audio file.

Getting help¶

Help ▸ Copy diagnostics puts a summary of the detected toolchain on the clipboard, and ffmpeg-audio-encoder --diagnostics prints the same information as JSON. Attach one of those to any issue you open.