5 Aug
2009
5 Aug
'09
1:38 p.m.
On Wed, Aug 5, 2009 at 7:53 AM, Fons Adriaensen<fons(a)kokkinizita.net> wrote:
On Wed, Aug 05, 2009 at 10:18:57AM +0200, Arnold
Krille wrote:
That sounds like an interesting, fun (and difficult) project. But,
well outside the problem domain of jamin.
Jamin has been stable and useful mainly because it was always focused
on one specific task.
No reason not to reuse any useful parts of jamin for this new
application, of course. That has already been done once for a DJ
application.
--
joq
Some kind of PA-managment that does:
a.1) Frequency-X-over and EQ (and phase and that)
a.2) If EQ is to simple nowadays (as some Fons guy tells me:), sophisticated
digital filters...
b) Multiband-compressor (one per output or with independant frequency ranges)
and limiters for the outputs.
The Lake (later Dolby) processors used to be one of the
standard solutions for this (and are still used today),
see <http://www.dolby.com/professional/Live_Sound/>.
They are no longer sold AFAIK, one of the reasons being
that manufacturers of PA systems are integrating this
sort of technology, see e.g. <http://www.labgruppen.com/>
and <http://www.qscaudio.com/products/dsp/dsp30/dsp30.htm>.
Integrating xover filtering, EQ, and limiting and sometime
even the power amp brings some important advantages - it
removes many connections, gain controls and other things
that could be done or set wrong and render such processing
completely useless, and it allows to adapt the processing
to specific speaker units and combinations, usually selected
from a menu on the user interface, or by networked remote
control apps. A big PA system today is completely set up
and monitored via network connections, and this can include
things such as amplifier and even speaker temperatures.
You can do all of this on a PC, but it's a shaky solution
unless you do some further integration - for one there will
be far too many gain controls after the processor, making
nonsense of any carefully set limiter levels and exactly
matched xovers. One other problem is to find the proper
settings for your hardware, and that is one reason why this
job is today done mostly by the speaker manufacturers who
have precise information about the electrical, mechanical
and thermal limits of their products.
c) Automatic (and live!) comparison of the input
signal that is routed to the
outputs and the signal from one or more measurement microphones to adapt to
the room with only a mouse-click. And to re-adapt to the room when the crowd
is in.
d) Feedback-destroyer (which will probably be hard to do while c) is running
its adaption.
Not so simple but certainly possible and much more suited
to software based solutions.