This is my fork of the "Ambisonic Decoder Toolbox" found at https://bitbucket.org/ambidecodertoolbox/adt originally by Aaron J. Heller.
This fork is intended to be for practical audio use, not for academic use.
Aim is to make it more usable & compact.
- Space reductions: removing large uneeded files & compressing PDFs (The upshot is that it is now only takes up ~10M.)
- Auro & ITU decoder channels are now in SMPTE-ish order
- 7.0 ITU decoder completed, with implementable info for additional ITU configs
- Auro decoder is 9.0 instead of 10.1 to support AMBEO
- Removed some info during config generation
- Plots off by default for faster, error-free generation
If you're hacking/building your own decoders, if your vertical plane is irregular, say, Auro 3D 9.0, the height later takes from expected, but the ELL layer is partway between ELL & bottom to take the sound of ELL & the non-existent bottom layer. At least, this is true for 1oA. If you're making oA encoders instead, the coefficients can be very different to decoders, especially for irregular configs. Try to pad to regular shapes, make a decoder, then base your encoder off that.
- If you have >2 vertical speaker layers, you should use 2oA+ v
- If you have >4 horizontal speakers, you should use 2oA+ h
- If you have an irregular config, you should use 2oA +
- You should be using >1oA anyway, as 1oA is blurry.
- If you're playing back 2D 1oA or 1oA mixes with major content on the horizontal plane on >2 vertical speaker layers, the R channel from 2oA is important for Z resolution, otherwise ELL will be smeared over the height