2007/4/26, Paul Martin <<a href="mailto:pm@nowster.zetnet.co.uk">pm@nowster.zetnet.co.uk</a>>:<div><span class="gmail_quote"></span><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
If the W channel is stored in the AMB file at a reduced level, and<br>I'm decoding from an AMB file<br></blockquote><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
[...]<br>OR, am I to assume that the 3dB reduction in the W channel is a<br>prerequisite the Furse-Malham decodes already take into account</blockquote></div><br>Actually, Furse-Malham equations describe a second-order Ambisonic encoder. (The scalability of Ambisonics makes it possible to use a subset of them (first three) for 1st order, planar encoding.) There is a -3 dB coefficient in the first of them
<br><br>W = 1/sqrt(2) Ó a(n) = c Ó a(n),<br><br>in the other words we obtain the W signal by dividing the sum of all sound sources by sqrt(2). Anyway, the exact solution (velocity decode) for a rectangle rig should not assume any coefficient
<br><br><span class="q" id="q_1122a930acc215a9_0">a(n) = W/c + X/cos ö</span><span class="q" id="q_1122a930acc215a9_0">(n)</span><span class="q" id="q_1122a930acc215a9_0"> + Y/sin ö</span><span class="q" id="q_1122a930acc215a9_0">
(n)</span><span class="q" id="q_1122a930acc215a9_0">, <br><br>[ö</span><span class="q" id="q_1122a930acc215a9_0">(n)</span><span class="q" id="q_1122a930acc215a9_0"> is the angle of each speaker's position, measured anticlockwise from forward direction], so
<br>as long as you divide W by the coefficient used for encoding, you still have velocity decode.<br></span>