ambisonictoolkit/atk-matrices

A set of matrix files for matrix encoders and decoders

ambisonictoolkit

atk-matrices's Issues

Spherical sampling A/B matrices

Include spherical sampling A/B matrices.

Initially - FOA

E.g., T-designs:

https://en.wikipedia.org/wiki/Spherical_design
http://neilsloane.com/sphdesigns/
https://arxiv.org/abs/1502.03562
http://www.springer.com/gp/book/9783642259821

7_0 encoder, incorrect angles

The current 7_0 encoder should have these angles: [ 0, 30, 90, 135, -135, -90, -30 ]

Instead, the current encoder angles are: [ 0, 30, 90, 135, -130, -90, -30 ]

In the longer term, we may want to include alternative versions of 7.0 encoding. The larger issue is that there is no definitive standard for 7.0. (The usual story with "surround" formats!)

Add a 7.1 decoder

Migrate fixed matrices to atk-matrices

Will want to include various required weightings, e.g., A-to-B:

• 'canonical'
• 'diffuse'
• 'cardioid'
• &c.

T-design notice

Spherical Designs - R. H. Hardin and N. J. A. Sloane

McLaren's Improved Snub Cube and Other New Spherical Designs in Three
Dimensions, R. H. Hardin and N. J. A. Sloane, Discrete and
Computational Geometry, 15 (1996), pp. 429-441.

"A library of putatively optimal spherical t-designs." [Online].
Available: http://neilsloane.com/sphdesigns/index.html.
[Accessed: 04-Sept-2015].

REAPER text file reading is very primitive

I have done some experiments with reading matrices into REAPER today. It turns out that REAPER functionalities for this are fairly primitive. REAPER basically needs each line to contain one number only. Here's from the documentation of file I/O:

The format of each line in the text file can be either a floating point number, a binary number beginning with 'b', i.e. b0101010111, or a hexadecimal number beginning with 'x', i.e. xDEADF000.

This implies that the current formatting of matrices can not be used with Reaper.

The only suggestion that I can think of if we want to have a shared library of matrices, is that they have to be organised according to the rather cumbersome convention:

<number of rows>
<number of columns>

<value of cell 1, 1>
<value of cell 1, 2>
....
<value of cell 1, n>

<value of cell 2, 1>
<value of cell 2, 2>
....
<value of cell 2, n>

....

<value of cell m, 1>
<value of cell m, 2>
....
<value of cell m, n>

Any thoughts, @joslloand ?

Review fixed matrix requirements for atk-reaper

At this time atk-reaper actually uses 5_1 matrix.

• Check channel ordering for Quad.
• Check 5_0 vs 5_1 matrices

*.mosl.txt files - comma in comment lines cause read error in Reaper

I have discovered a weird shortcoming when reading *.mosl.txt files in Reaper. If a comment line contains a comma (,), the file will not be read by Reaper. This cause problem with all matrices in v.1.0 as a comma is used consistently in the // Dimensions: rows, columns line in the header section:

// set : FOA

// type : decoder

// op : matrix

// kind : hoa1

// note : A first order Ambisonic format exchange decoder.

// author : Ambisonic Toolkit

// Dimensions: rows, columns
4
4

@joslloand : The easiest solution is to just rid ourselves of that comma, changing the line to

// Dimensions: rows columns

Two questions:

1. Are these files only used in Reaper? If so, it should be straightforward to remove the comma
2. Are they autogenerated from the YAML file? If so the code for generating them also need to be updated

This issue relates to ambisonictoolkit/atk-reaper#92

FOA ambix1 (& hoa1) format exchange matrices

Remove redundant hoa1 matrices removed format exchange matrices to facility atk-sc3 development.

UPDATE: These matrices have been moved rather than removed. atk-sc3 uses these matrices in their new locations.

Two READMES!

Consolidate README.md & README.

update README

Make verbose. Should be similar to https://github.com/ambisonictoolkit/atk-kernels/blob/master/README.md

Include the following:

• list / summarize matrices
• credit 3rd parties
• installation instructions

remove duplicate hoa matrices

Review / Update License

License should be CC for matrices.