Re: [Sursound] Ambisonics not on the cheap

[David Worrall]

Hi Ben.
Thanks for your overview and relating of experiences, Ben. Much appreciated. 
You said
>> We ended
>> up piecing together an 8 channel system for a 1700 seat hall that was not
>> accurate in the slightest. However, my professor and the sound designer were
>> both blown away by the precision of what was essentially glorified panning,
>> despite the large hall.

A couple of rwsponses, FYI.
One of the things that I think gets lost in all the technical chatter about 
ambisonics is how "robust" it actually is:
Sure, you can get better performance by spending $squillions on gold cabling 
etc etc but it is surprising how forgiving of abuse it is. Dave Malham will 
remember and experiment I did in York in 1996 before the advent of 
(affordable?) 4 track digital recorders - I took did some field recordings with 
their (ST250, I think it was) onto two unsynced DAT recorders, made a 4ch 
reconstruction in the studio and on playback, whammo! - up came the soundfield 
like it had been recorded with precision!

On the Dome front - and the number of speakers depends on lots of things - do 
you have any photos? One I found was not to place a speaker at the very top. 
Here's some domes I've built and used ambisonics with:
http://www.avatar.com.au/worrall/index.php/polymedia-event-theatres
+ a 3rd one is on the way:
http://www.avatar.com.au/worrall/index.php/current-projects/35-polymedia-event-theatre-no-3

I'm working towards a residency in Atlanta later in the year - and they've 
expressed an interest is spaces and configurations for sonification, so it may 
be that we can touch base and compare notes/benefit from knocking heads 
together.

cheers, David (W)

On 30/07/2011, at 1:43 PM, Ben Bloomberg wrote:

> Ok, at the behest of Jörn, I am going to send some links to various things.
> Feel free to tear them apart! Also for what it's worth, the setup pictured
> in the first article is about the worst sounding rig we've put together. The
> speakers are really not that bad, though. I wish we had given them some
> other photos.
> 
> http://blog.bowers-wilkins.com/sound-lab/the-future-of-surround-sound-from-mit/
> http://bot23.com/tag/ambisonics/
> 
> We also had a paper in NIME this year that has a description of our current
> project:
> http://www.nime2011.org/proceedings/papers/J01-Jessop.pdf
> 
> Ben
> 
> 
> 
> On Fri, Jul 29, 2011 at 12:05 PM, Ben Bloomberg  wrote:
> 
>> Hi Everyone,
>> 
>> Well, now that Sarang has spoken up, he's given me the courage as well.
>> I've been lurking on this list for several months now. Just to be clear, I
>> think you're all high-priestesses!! It's been incredible to read the
>> discussion (most of which has been way over my head, but I try to slog
>> through) and hear what people have to say.
>> 
>> My experience with ambisonics is odd, having sort of jumped into the deep
>> end after suggesting ambisonics as a solution for a school project 3 years
>> ago (with little knowledge of math, let-alone spherical harmonics). We ended
>> up piecing together an 8 channel system for a 1700 seat hall that was not
>> accurate in the slightest. However, my professor and the sound designer were
>> both blown away by the precision of what was essentially glorified panning,
>> despite the large hall. We used the ICST Max 4.5 externals.
>> 
>> Since then, we've implemented our own parallel processing audiounit/OSC
>> based system and we use ambisonics as an abstracted panning system for
>> synthetic sound environments on most of the performances we put together. My
>> professor loves the idea that we can encode something and play it back on
>> any speaker configuration with relatively consistent results (given the
>> halls we play are vastly different anyway, usually the ambisonic
>> configuration is the least of our concerns).
>> 
>> So my experience is almost purely empirical/trial-and-error. We've done
>> systems for 3 people and systems for 2000 people, some very high density and
>> some very low density. Mostly we just put up speakers and listen to see what
>> we think, or take the live-sound fudging approach of blurring multiple
>> systems together for different seating areas. Having taken some more math
>> classes, things are starting to become a bit clearer now, but I'm still
>> learning (that never stops!). Some of our equipment sponsors have asked me
>> to write things about the projects for them, which have ended up on the
>> internet and probably contain some mis-information.
>> 
>> That said, I have a two questions and an offer.
>> 
>> First, I'm really interested in ambisonic decoder and encoder weightings.
>> From what I read, they provide a means to used fixed-point processing.
>> However, playing with them results in vastly different sounding results. I
>> made a really simple script to visualize the phase of a signal at the output
>> of decoders with different weightings, but I'm not sure I understand what
>> I'm seeing or how that has an effect on what we hear. I will try to ma

Re: [Sursound] Ambisonics not on the cheap

[Ben Bloomberg]

Ok, at the behest of Jörn, I am going to send some links to various things.
Feel free to tear them apart! Also for what it's worth, the setup pictured
in the first article is about the worst sounding rig we've put together. The
speakers are really not that bad, though. I wish we had given them some
other photos.

http://blog.bowers-wilkins.com/sound-lab/the-future-of-surround-sound-from-mit/
http://bot23.com/tag/ambisonics/

We also had a paper in NIME this year that has a description of our current
project:
http://www.nime2011.org/proceedings/papers/J01-Jessop.pdf

Ben



On Fri, Jul 29, 2011 at 12:05 PM, Ben Bloomberg  wrote:

> Hi Everyone,
>
> Well, now that Sarang has spoken up, he's given me the courage as well.
>  I've been lurking on this list for several months now. Just to be clear, I
> think you're all high-priestesses!! It's been incredible to read the
> discussion (most of which has been way over my head, but I try to slog
> through) and hear what people have to say.
>
> My experience with ambisonics is odd, having sort of jumped into the deep
> end after suggesting ambisonics as a solution for a school project 3 years
> ago (with little knowledge of math, let-alone spherical harmonics). We ended
> up piecing together an 8 channel system for a 1700 seat hall that was not
> accurate in the slightest. However, my professor and the sound designer were
> both blown away by the precision of what was essentially glorified panning,
> despite the large hall. We used the ICST Max 4.5 externals.
>
> Since then, we've implemented our own parallel processing audiounit/OSC
> based system and we use ambisonics as an abstracted panning system for
> synthetic sound environments on most of the performances we put together. My
> professor loves the idea that we can encode something and play it back on
> any speaker configuration with relatively consistent results (given the
> halls we play are vastly different anyway, usually the ambisonic
> configuration is the least of our concerns).
>
> So my experience is almost purely empirical/trial-and-error. We've done
> systems for 3 people and systems for 2000 people, some very high density and
> some very low density. Mostly we just put up speakers and listen to see what
> we think, or take the live-sound fudging approach of blurring multiple
> systems together for different seating areas. Having taken some more math
> classes, things are starting to become a bit clearer now, but I'm still
> learning (that never stops!). Some of our equipment sponsors have asked me
> to write things about the projects for them, which have ended up on the
> internet and probably contain some mis-information.
>
> That said, I have a two questions and an offer.
>
> First, I'm really interested in ambisonic decoder and encoder weightings.
> From what I read, they provide a means to used fixed-point processing.
> However, playing with them results in vastly different sounding results. I
> made a really simple script to visualize the phase of a signal at the output
> of decoders with different weightings, but I'm not sure I understand what
> I'm seeing or how that has an effect on what we hear. I will try to make a
> comped image of all the plots and send it out.
>
> Second, we will be putting up a fairly large ambisonic system hanging from
> a 20 by 20 foot truss in the next month. I have heard that the more speakers
> you can add to a system, the better it will sound. Is that true? With higher
> speaker counts >20, we've had the best results by removing decoder
> weightings. We seem to hear a better sense of localization for our synthetic
> recordings. I had an idea that this changes the decay rate of "cosines" in
> what is essentially a vector projection, so each speaker occupies a smaller
> slice of the surround field "pie". Is that anything close to correct?  What
> would be the optimum number and configuration of speakers for a 20' dome?
> (assuming equipment is not an issue)
>
> Ok, now the offer! Being at a place that has a lot of resources, but little
> time and man-power, I wanted to offer up our inventory of equipment to the
> community. We have about 90 speakers that we keep around for projects or
> touring. If anyone is working on a project that could benefit from extra
> speakers, drop me a line. If we've got the gear free, you're welcome to it.
> It's a mish-mash of Bowers and Wilkins, Duran Audio Axys, and Mackie 824s.
> We drive our systems with MOTU, RME, and SSL MADI interfaces. Also, if
> people are interested in testing things on this 20x20 dome and want to come
> down to Boston (it'll be up in September and October), we'd be honored to
> have you as guests!
>
> Best wishes,
> Ben
>
-- next part --
An HTML attachment was scrubbed...
URL: 

___
Sursound mailing list
Sursound@music.vt.edu
https://mail.music.

[Sursound] Ambisonics not on the cheap

[Ben Bloomberg]

Hi Everyone,

Well, now that Sarang has spoken up, he's given me the courage as well.
 I've been lurking on this list for several months now. Just to be clear, I
think you're all high-priestesses!! It's been incredible to read the
discussion (most of which has been way over my head, but I try to slog
through) and hear what people have to say.

My experience with ambisonics is odd, having sort of jumped into the deep
end after suggesting ambisonics as a solution for a school project 3 years
ago (with little knowledge of math, let-alone spherical harmonics). We ended
up piecing together an 8 channel system for a 1700 seat hall that was not
accurate in the slightest. However, my professor and the sound designer were
both blown away by the precision of what was essentially glorified panning,
despite the large hall. We used the ICST Max 4.5 externals.

Since then, we've implemented our own parallel processing audiounit/OSC
based system and we use ambisonics as an abstracted panning system for
synthetic sound environments on most of the performances we put together. My
professor loves the idea that we can encode something and play it back on
any speaker configuration with relatively consistent results (given the
halls we play are vastly different anyway, usually the ambisonic
configuration is the least of our concerns).

So my experience is almost purely empirical/trial-and-error. We've done
systems for 3 people and systems for 2000 people, some very high density and
some very low density. Mostly we just put up speakers and listen to see what
we think, or take the live-sound fudging approach of blurring multiple
systems together for different seating areas. Having taken some more math
classes, things are starting to become a bit clearer now, but I'm still
learning (that never stops!). Some of our equipment sponsors have asked me
to write things about the projects for them, which have ended up on the
internet and probably contain some mis-information.

That said, I have a two questions and an offer.

First, I'm really interested in ambisonic decoder and encoder weightings.
>From what I read, they provide a means to used fixed-point processing.
However, playing with them results in vastly different sounding results. I
made a really simple script to visualize the phase of a signal at the output
of decoders with different weightings, but I'm not sure I understand what
I'm seeing or how that has an effect on what we hear. I will try to make a
comped image of all the plots and send it out.

Second, we will be putting up a fairly large ambisonic system hanging from a
20 by 20 foot truss in the next month. I have heard that the more speakers
you can add to a system, the better it will sound. Is that true? With higher
speaker counts >20, we've had the best results by removing decoder
weightings. We seem to hear a better sense of localization for our synthetic
recordings. I had an idea that this changes the decay rate of "cosines" in
what is essentially a vector projection, so each speaker occupies a smaller
slice of the surround field "pie". Is that anything close to correct?  What
would be the optimum number and configuration of speakers for a 20' dome?
(assuming equipment is not an issue)

Ok, now the offer! Being at a place that has a lot of resources, but little
time and man-power, I wanted to offer up our inventory of equipment to the
community. We have about 90 speakers that we keep around for projects or
touring. If anyone is working on a project that could benefit from extra
speakers, drop me a line. If we've got the gear free, you're welcome to it.
It's a mish-mash of Bowers and Wilkins, Duran Audio Axys, and Mackie 824s.
We drive our systems with MOTU, RME, and SSL MADI interfaces. Also, if
people are interested in testing things on this 20x20 dome and want to come
down to Boston (it'll be up in September and October), we'd be honored to
have you as guests!

Best wishes,
Ben
-- next part --
An HTML attachment was scrubbed...
URL: 

___
Sursound mailing list
Sursound@music.vt.edu
https://mail.music.vt.edu/mailman/listinfo/sursound