# Talk:Ambisonic data exchange formats

## Math needs verification

Someone please look over this. I have taken Nachbar et al. and Chapman as the basis for notation, which means that some other sources had to be converted to this notation. I'm positive that something has gone horribly wrong at least once despite my best efforts. Just a humble sound engineer who's picking all this stuff up as he goes. Please validate. Nettings (talk) 22:38, 19 December 2013 (UTC)

The 4π should be removed from the N3D / SN3D definition to make consistent with table listing of individual harmonics. This will also match the original definition of N3D harmonics by Daniel. Otherwise the harmonics listed should be divided by Rdmgrdmg (talk) 14:34, 24 February 2016 (UTC)
Yep, having just gone through plotting spherical harmonics, all mine were out by a factor of as mentioned above. This needs removing (and seems to be present in much literature too). Also, the Condon-Shortley phase is slightly incorrect to. Condon-Shortley is in the Matlab function legendre (and needs removing). It looks like it's in the S3D normalisation as it's in again, perhaps to remove it from legendre in this case? DrWig (talk) 17:59, 24 February 2016 (UTC)
As to CS phase, I removed the quote from Daniel, which either was a mistranslation on my part (although I checked it again and believe I got it right) or reflected a much older version of MATLAB. Instead, I tried to find out how current versions of Mathematica, MATLAB, and GNU Octave handle it.Nettings (talk) 21:13, 25 March 2016 (UTC)
Here's a link to a page giving the different normalisation schemes for Spherical harmonics. Note the missing 4π in the SN3D equation. https://shtools.oca.eu/shtools/www/conventions.html --DrWig (talk) 11:43, 12 April 2016 (UTC)

## Format table needs filling in

My idea is to add a table that lists all components of all formats with their weighting, spherical harmonic formula, and visualisation. If someone beats me to it, that's fine. :) I'll start by chopping up Zotter's wonderful Matlab renderings into single harmonics and uploading them. Nettings (talk) 22:38, 19 December 2013 (UTC)

## Or, Another Approach?

Although the talk about normalisation, ordering and so on is interesting on the theoretical side, it has no relevance to someone wanting a format to exchange ambisonic data. In practice, only FuMa and SN3D/ACN are in wide use. Could we just include the encoding formulae for these and be done? This would be much easier to do, and a lot less confusing for a reader. At the moment, the page reads as if users have to be familiar with normalisation, ordering, polarity and so on - which is not the case. In practice, it's FuMa, or it's SN3D.— Preceding unsigned comment added by 77.99.107.200 (talk) 10:06, 16 October 2017 (UTC)

Looks good to me. Things have moved on since this page was conceived. 74.205.216.186 (talk) 17:14, 16 October 2017 (UTC)