MultitudEnc is a free app by BrainCrash soft. It was born mainly to convert multi-channel audio of video files to stereo. There are many tools for video encoding, but I didn't find one that pleased me to convert just audio... so I made my own. With this tool, you can encode the audio to stereo while leaving the video untouched. This will make the process much faster than re-encoding the whole video. The app is built with WPF and C#, and uses FFmpeg under the hood for all media processing tasks.
MultitudEnc is basically a frontend for FFmpeg. You must have FFmpeg available for the app to work. Choose one of the two methods below.
bin\ folderPath under System VariablesC:\ffmpeg\binNo PATH changes needed. Place the executables next to the app:
Both ffmpeg.exe and ffprobe.exe are required.
ffmpeg -version.
If you see version info, you're ready.
Drag any number of video files (MKV, MP4, AVI, MOV…) onto the drop zone at the top of the app window. The zone highlights in amber when you hover over it. You can drop multiple files at once, each becomes its own collapsible File Card.
Each File Card auto-scans the file with ffprobe and lists all audio tracks.
Every track is checked by default, uncheck the ones you don't need.
✓ Include track in output[LANG] Language tag (ENG, SPA, FRE, UND…)[CODEC] Source codec (AC3, DTS, EAC3…)[BR] Bitrate or VBR[CH] Channel layout (5.1, 7.1, 2.0…)
The editable text box sets the title metadata tag
written into the output stream. ENG → "English" and other common languages are auto-filled.
Edit freely before encoding.
In the Audio panel, toggle encoding and set output quality:
192k, 256k).
-ac 2. Safe for all channel layouts.
In the Video panel, choose between stream-copying (fast) or full re-encoding:
libx264, libx265 or NVIDIA GPU: h264_nvenc, hevc_nvenc. NVENC requires an NVIDIA GPU (GTX 900+ / RTX).
ultrafast → placebo (slower = better quality). NVENC: p1 (fastest) → p7. The preset list rebuilds automatically when you switch encoder.
8M / 16M). Leave blank to skip.
yuv420p10le (CPU) or p010le (NVENC). Useful for HDR or high-fidelity encodes.
The Output Naming panel controls how output files are named. Outputs are always saved in the same folder as the source file.
.BluRay → Replace with nothing to strip a release tag.
_stereo or _enc to avoid overwriting the source.
-y. Always add a suffix when unsure.Hit PROCESS to start. Jobs run sequentially with real-time feedback:
The -ac 2 flag handles all channel layouts (5.1, 7.1, mono…) safely using FFmpeg's built-in matrix mixing. Already-stereo tracks are passed through with no quality loss.
Files with multiple audio tracks (e.g. English + Spanish) work great. Uncheck unwanted tracks, rename the ones you keep, they'll all be encoded and tagged in the output.
NVENC requires an NVIDIA GPU (GTX 900+ / RTX series). CQ controls quality like CRF. VBV fields map to -maxrate / -bufsize for NVENC, and to -x265-params / -x264-params for CPU encoders.
If both Encode Audio and Encode Video are disabled, PROCESS will show a warning and refuse to start. Enable at least one pass before processing.
When re-encoding audio to AAC, the bitrate shown in tools like MediaInfo may not match the one you selected. This is not a bug in this tool. It is a known limitation of how FFmpeg's AAC encoder works: it targets the requested bitrate but operates in a variable mode internally, and the container may carry over the original stream's declared bitrate rather than the re-encoded one.
In practice, the audio quality will reflect your chosen bitrate. If accurate bitrate reporting matters for your workflow, consider using AC3 instead, which encodes in true constant bitrate mode and reports correctly in MediaInfo.