hi tapio, fons,
On 01/23/2014 01:29 AM, Fons Adriaensen wrote:
On Wed, Jan 22, 2014 at 04:58:29PM +, Lokki Tapio wrote:
One time window gives one direction estimate as it is an impulse
response which is analyzed. In ideal impulse everything (all
frequencies from one direction) are in one sample. In practise
impulses are spread in time, but the idea still works well.
to rephrase my priginal question: when you make a direction estimate for
a single reflection, do you assume it's a scaled dirac pulse, or do you
also take some "coloration" information by looking at the actual pulse
over a short time window, to measure effects such as low-pass filtering
etc.?
Doesn't this then rely on reflections being separated more than
the differences of time of arrival at the mics?
i don't think that's true in the general case, but there may be some
incidence patterns for near-simultaneous reflections which are
ambiguous. but thinking about that makes my head hurt :)
And even if not (if you really use correlation), as the density
of reflections increases, there will be two or more within each
time interval. In the reverb tail there will be many. How do you
separate them ?
if they reconstruct with DirAC, they don't need to. rather, the tails
would be decorrelated per speaker and played back as a non-directional
diffuse field.
It's possible to do such things using MDS and related techniques,
but none of those have been mentioned.
out of curiousity, what does MDS stand for?
--
Jörn Nettingsmeier
Lortzingstr. 11, 45128 Essen, Tel. +49 177 7937487
Meister für Veranstaltungstechnik (Bühne/Studio)
Tonmeister VDT
http://stackingdwarves.net
___
Sursound mailing list
Sursound@music.vt.edu
https://mail.music.vt.edu/mailman/listinfo/sursound