- From: Richard.Webb.my.foot@xxxxxxxxxxxxxxxxxxxxxxxxxxx (Richard Webb)
- Date: 10 Jun 2012 18:46:54 GMT
On Sun 2012-Jun-10 04:37, Don Y writes:
This is all a pretty tall order.
Actually, I already own implementations of the core technologies
involved. And, gluing things together is really not that
difficult: if you are here and this happens then do that.
<hmmm> I'm trying to get this straight in my head here.
Bear with me.
IF done at the individual speakers you're going to want to do
it with a combination of hardware and software I'd think.
If you constrain the problem "at the speaker", its a manageable
issue. I.e., attenuating one signal and mixing in a second
has very low cost. But, you have to set limits on just how
far you are willing to take this. I.e., can you mix in two
or three or five alarms concurrently?
okay, understood, but that brings up the question of what
drives the speakers. Are you pushing data digitally to each room, then using local d/a conversion to play the sounds?
Or, are you generating the analog audio at the media server, amplifying it and pushing it down speaker lines to the
various places? I know you probably discussed this in the
past in threads here, but i wasn't paying a lot of
attention, because at the tiem it wasn't relevant to what I
do so I probably blew past the threads. <sorry 'bout that
At some point, you have to decide that you are increasing the "cost"
for increasingly rarer events. And, look for other tricks to work
around those problems. For example, if someone presses the doorbell
just as the phone is ringing, delay the doorbell annunciator until
the phone annunciator has played out. Of course, if the phone and
doorbell will not both be routed to a speaker that is already
reproducing some source program, you can play both of them at the
same time -- via different speakers!
Yep, and as you and I think Bob noted elsewhere,
psychoacoustics comes into play here.
A lot of these judgments when mixing a show or something for broadcast are made by me using my ears at the time. If the
elements are prerecorded then I can manipulate and
"rehearse" until I get it right, but live right now I'm
always adjusting at least one element slightly to get
everything to sit right. The easy part would be such things as Carolyn is handling an important phone call and doesn't
want to hear the doorbell, or, Don's in the middle of a
project and would choose to ignore the phone.
Simple example, my workspace here, if I'm handling audio for hire I'm in the control room in the back of the truck so
that doesn't come into play, but consider my worksapce here.
i've got at least four speakers going to deliver sound just
in this room, and all four may be going at once. There's
the speaker for the uhf channel I monitor, the vhf ham radio channel as well, the high frequency radio which camps on a
frequency where I assist vessels at sea. Then there's the
speaker which is the output for my screenreader. The
priority always changes. The screenreader may be chattering away as I read this newsgroup when I hear a vessel call on
the hf. All of a sudden I want the screenreader to stop
chattering. NOw if one of the old duffers is on the local
vhf frequency complaining about his hemhoroids I get to
reach up and turn that channel's volume down as well <grin>.
I wouldn't want to consign all the different electronic
audio outputs around here to any sort of automation <grin>.
Could be done, but ...
<hmmm> This is a pretty complex undertaking I'd think.
Not really. The most complex part of it is coming up
with a way by which "users" can convey their preferences
to the system. I.e., how they can specify the "rules"
that the system will enforce -- without having to understand the
MIght work for some. I'd have so many rules though ... oh,
and add to the above occasionally another receiver listening to a news program on the radio.
LOts of balls in the air with this one. Where you are in
the house, what sort of program material, what sort of alert
or other audio needs to preempt the program material, etc.
etc.<ouch!> SOunds like you're talking about a media
server with oodles of processing power available to it.
No. All the media server has to do is push program material out to
speakers (and video displays). It doesn't have to decide *what* to
send where. Something else can command it based on other inputs
(i.e., the locations of the people in the house, their preferences,
the rules in place, etc.).
YEp, that's where your bluetooth comes in, telling it who's
where, but locally to each set of speakers there has to be
some processing power to do this. <hmmm>
Delivering audio and video in a timely fashion is a chore
but already solved. I'm working on the location resolution
system, currently. And, the expert system that implements the rules
is old technology. It's just putting it all together
and coming up with a clean interface that's the issue.
Okay, but I can see this one in "beta test" if not alpha for quite awhile.
| Remove .my.foot for email
| via Waldo's Place USA Fidonet<->Internet Gateway Site
| Standard disclaimer: The views of this user are strictly his own.
- Re: Mixing
- From: Don Y
- Re: Mixing
- Re: Mixing
- From: Don Y
- Re: Mixing