Script that will take a frame-sequential mk3d 3D video and convert it to Apple Spatial Video.
The mk3d file can be created with MakeMKV or similar.
You will need an Apple Silicon-based Mac running macOS Sonoma (see caveats).
You must also have the following installed and configured on your system:
ffmpeg, most easily installed with homebrewldecod(part of the JM H.264/AVC reference software), Mac binaries available on GitHubspatial-media-kit-toolmp4box, part of the GPAC framework
You will also need sufficient disk space to store the intermediate files and the final Spatial video.
| Option | Example use | Description |
|---|---|---|
-x--software |
--software |
Use libx265 software encoder (slower, higher quality). This is the default encoder.Since the libx265 conversion is lossless, there is no quality to specify. |
-t--hardware |
--hardware |
Use hevc_videotoolbox hardware encoder (faster, lower quality).Specify quality with --hevc-quality. |
-q--hevc-quality |
--hevc-quality 50 |
Quality to use with hevc_videotoolbox (1 to 100, 100 is highest quality/bitrate; default 50)Larger target quality (especially above 50) will use lots of disk space. |
-s--spatial-quality |
--spatial-quality 50 |
Quality to use with spatial-media-kit-tool (1 to 100, 100 is highest quality/bitrate; default 50) This will have the largest impact on the bitrate/size of the final spatial video. |
-a--audio-bitrate |
--audio-bitrate 512k |
Compress audio with AAC at this bitrate using aac_at to save space.If omitted, lossless LPCM audio will be used instead. |
-k--keep-intermediate-files |
--keep-intermediate-files |
Instead of removing intermediate files when they are no longer needed, leave them on disk. Good to use when you are testing so you can run a later step without redoing all previous steps. Makes the process require more total disk space. |
-v--verbose |
--verbose |
Enable verbose output. Not too noisy, good to enable to see what's going on. |
This script has been tested on these configurations:
- Apple Silicon M2 Mac running macOS 14.3.1
It might on Intel-based Macs, as well as other versions of macOS Sonoma, but your mileage may vary. If you get it working on your machine, please submit a PR to update this list!
This will not work on Linux or any other non-Apple OS, as the Spatial video encoding relies on APIs only available in macOS.
Big thanks to:
- sturmen on the Doom9 forums, for a an encoding guide using
FRIM Decoderas well as creating the spatial-media-kit-tool - Vargol on GitHub, for making the JM reference software build properly on macOS as well as an example script that was a useful reference