Free online USM to MP4 converter

Game cutscenes stored as .usm files usually will not open in standard players, editors, phones, or upload services. Converting a decrypted, decodable file to MP4 produces a format suitable for playback, editing, sharing, and archiving.

What the USM format is

USM is a multimedia container used by CRI Middleware's Sofdec video system. Game developers use it for prerendered cutscenes, opening movies, tutorials, trailers, and background video.

A USM file can interleave video, one or more audio tracks, timing information, and game-specific metadata. Codec support varies by game and Sofdec version; common streams include MPEG video or H.264 video with CRI ADX or HCA audio. Some game assets are encrypted, and no converter can decode them without the required game-specific key or an authorized decrypted copy.

Changing movie.usm to movie.mp4 changes only the filename extension. The streams must be demuxed or decoded and then written into an MP4 container.

What the MP4 format is

MP4 is a media container supported by current browsers, mobile devices, television players, editing applications, and sharing platforms. An MP4 with H.264 video and AAC audio provides the broadest practical compatibility.

MP4 does not itself define video quality or compression. Re-encoding older MPEG video to H.264 can reduce file size, but it creates a new lossy encode; stream copying avoids that loss only when the source codecs are suitable for MP4 playback.

Convert a decodable USM file

To convert without installing software, upload the .usm file to 101convert.com, choose MP4, start conversion, and save the output. Encrypted files and files containing unsupported proprietary streams must be decrypted or demuxed locally first.

For local conversion, use FFmpeg to check whether it recognizes the container and streams:

ffmpeg -i input.usm

If FFmpeg reports a video stream and usable audio, encode a broadly compatible MP4:

ffmpeg -i input.usm -map 0:v:0 -map 0:a? -c:v libx264 -crf 18 -preset medium -c:a aac -b:a 192k -movflags +faststart output.mp4

-map 0:a? includes every detected audio track without failing if no audio exists. Use -crf 18 for high quality, or approximately 20 to 23 for smaller files. -movflags +faststart moves MP4 indexing data to the beginning of the file for faster web playback.

If FFmpeg identifies H.264 video and AAC audio, remuxing can preserve the original streams:

ffmpeg -i input.usm -map 0:v:0 -map 0:a? -c copy -movflags +faststart output.mp4

Use stream copying only after testing the output in the intended player or editor. MPEG-2, ADX, HCA, and other nonstandard audio or video streams may be accepted by an MP4 muxer in limited cases but are not a reliable compatibility target.

Demux files that FFmpeg cannot read

If FFmpeg cannot open the USM file, extract its streams with a CRI USM demuxer. On Windows, VGMToolbox includes a CRI movie demuxing function that can produce separate video and audio files; the exact extensions depend on the source.

ADX audio is decoded by many FFmpeg builds. Convert it to WAV when needed:

ffmpeg -i audio.adx -c:a pcm_s16le audio.wav

For HCA audio that FFmpeg cannot decode, use vgmstream-cli when the file is not encrypted or its required key is available:

vgmstream-cli -o audio.wav audio.hca

Then encode the extracted video and decoded audio into MP4:

ffmpeg -i video.m2v -i audio.wav -map 0:v:0 -map 1:a:0 -c:v libx264 -crf 18 -preset medium -c:a aac -b:a 192k -movflags +faststart output.mp4

Replace video.m2v with the video file produced by the demuxer. If the extracted video is already H.264 and plays correctly after remuxing, use -c:v copy while encoding the audio to AAC.

Quality and compatibility limits

USM files may hold several language tracks or alternate audio streams. To select the second detected audio stream during direct conversion, use -map 0:a:1 instead of mapping every audio stream. MP4 can retain multiple tracks, but some players expose only the first track.

Demuxing can expose video with incomplete timestamps or separate streams whose timing depends on the original USM container. If audio drifts, use the streams produced by the same demuxing run and verify their durations before encoding.

Game-engine subtitles, branching playback data, and other presentation behavior are not automatically reproduced in MP4. Keep the original USM and extracted streams if archival fidelity matters, because H.264 conversion cannot restore detail lost in the source or preserve unsupported game-specific data.