new article

[archive]

From: Dan Nelson <dprimary@x??????xxxxx.xxxx
Date: Wed Nov 17, 1999 7:18 am
Subject: new article

I've been working on some pages for the site, but I'm not sure if this
one is very clear. What do you guys think?
Bob Hodas had a similar explanation in the Feb. 99 Mix. But, I feel he
made it too simple and it could cause some problems. If you can find the
issue it is a very good article. It took me many years to figure out
many of points he makes in the article. Where was Bob 15 years ago?
A local acoustician told me about this trick a few years ago, so it must
common knowledge to acousticians.

Dan Nelson

Placement of absorption materials

Reducing First Order Reflections

In the mix position You will want to setup what is known as a
Reflection Free Zone (RFZ). In the RFZ you will want to eliminate any
early reflections of sound. The human brain interprets any reflection
more than 19-20 ms as a separate sound from the original source. Any
reflection that reaches the ear in under 19 ms the brain interprets it
as part of the source this tends to smear the sound from the monitors,
so the goal is to minimize these first reflections.
Any surface within 11 feet of the mix position could cause a first
order reflection, but treating every surface within 11 feet of the mix
position is not possible or necessarily desirable. In the area of the
room between the monitors and the mix position any surface that has a
21.5 foot complete path between the monitor and the mix position can
also cause a reflection.

One way to find the areas that need to be treated, will require two
people and a mirror (a plastic one is preferred since you may have to
tape it to a pole to reach some areas)
To help find the placement location have one person move a mirror along
the side walls of the control room with the other person siting at the
mix position. Any point the monitor can be seen from the mix position in
the mirror mark those points on the wall. Repeat this process for the
ceiling and back wall. Any of these points that are less then 11 ft from
the mix position will cause early reflections and is where you will
want to locate absorption materials. In the front half of the room
measure from the monitors to the marked reflection point to the mix
position if it is less than 21.5 feet be sure to treat those areas as
well. You may want to treat an area larger than what you have marked,
since you won't always be in the exact mix position.

The reason for the two measurements is, if you measured 22 feet from the
speaker to the back wall and back to the mix position your reflection
time is still under 19 ms because the monitor is a few in front of you
the sound has traveled 3-4 ms before it reached your ears so the
reflection will reach you in 15-16 ms. In the area in front of the mix
position you could have a reflection from something more than 11 feet
away reach you in less than the 19 ms since the spot is is less than 11
feet from the monitors.

[archive]

From: Mark Plancke <soundtch@x??x.xxxx
Date: Wed Nov 17, 1999 3:59 pm
Subject: Re: new article

>From: Dan Nelson <dprimary@e...>
<mucho good stuff snipped>

Dan

What is the prefered absorbtion required at these
reflection positions? Some wideband absorbtion like
4" 703 or is 2" 703 enough for this application?

Mark Plancke

SOUNDTECH RECORDING STUDIOS
http://SOUNDTECHRECORDING.COM
Windsor, Ontario Canada

[archive]

From: Dan Nelson <dprimary@x??????xxxxx.xxxx
Date: Wed Nov 17, 1999 6:32 pm
Subject: Re: new article

It would depend on your rt60 calculations and the space you have. But most of
the time I use 2' 703 or one of the foam products on the side walls and
ceiling and just 703 2" to 4"on the front wall (depending on the room) with
diffusers on the back wall.

Dan Nelson

[archive]

From: "Greg" <gfreitag@x??x.xxxx
Date: Thu Nov 18, 1999 1:49 am
Subject: Re: new article

>In the mix position You will want to setup what is known as a
>Reflection Free Zone (RFZ). In the RFZ you will want to eliminate any
>early reflections of sound.

True

> The human brain interprets any reflection
>more than 19-20 ms as a separate sound from the original source.

I must take issue with the above statement. Any reflection that
arrives within 50ms will be perceived as the same sound, this
is called the Haas effect. Try talking in a situation where the
reflections arrive outside the Haas zone, the brain interprets
the reflection as a discrete event. The time needed between
the direct and reflected energy, sometimes referred to as the
ITD gap (initial time delay gap) is a bit more complex than
what is stated above. Useful numbers can be as low as 10ms,
it is dependent on the ratio of the direct to reflected energy.
That being said, 20ms to 30ms is a good target.

Greg

[archive]

From: Dan Nelson <dprimary@x??????xxxxx.xxxx
Date: Thu Nov 18, 1999 6:14 am
Subject: Re: new article

My mistake, I was combining ITD gap and Haas effect into the Haas effect.
Do you have any insights into ITD gap and the Haas effect? After reviewing
the sources I have, I still find it difficult to explain in a meaningful
way.

I'm also trying to find all my sources for the 11-12 feet 19-20 ms for the
first order reflections, Scott seems to have run across some.

Dan Nelson

[archive]

From: SRF7@x??x.xxx
Date: Thu Nov 18, 1999 11:04 pm
Subject: Re: new article

In a message dated 11/18/99 1:14:49 AM Eastern Standard Time,
dprimary@e... writes:

> I'm also trying to find all my sources for the 11-12 feet 19-20 ms for the
> first order reflections, Scott seems to have run across some.

Malcolm Chisolm advised me regarding the design of my control room that I
needed a minimum of 11.5' behind my head (I think that was the figure he
used).

If that certification doesn't leave you breathless enough to just take it on
faith and move on, my understanding is that this limitation is pertinent to
practical real world applications (not the lab conditions under which
psycho-acoustics such as the Haas effect are measured)... best I can recall,
problems of "smearing" initial sounds with reflected sounds can be made to
occur even within the 20 ms gap as Greg points out... and as he advises they
can exist in even longer time domains, or for that matter, signals can
coexist in shorter time frames without causing this effect... relative volume
comes into play as well as the methods by which humans located the source of
a sound (near coincident (in time) sounds perceived to come from different
locations are still recognized as different sounds from different sources).

AES has an anthology on stereophonic techniques which gets into this stuff
which can be ordered from their website:

<A HREF="http://www.aes.org/publications/anth.cfm">http://www.aes.org/publicat
ions/anth.cfm</A>

I believe what Mr. Chisolm was trying to clue me in on in his advice to me
that I needed a minimum distance to the rear wall was that by taking
reasonable steps to treat the rear of the room with diffusion I could be
successful in curing unwanted smearing of the reverb trails into and over the
initial signals, provided I had about 11.5' feet behind my head... but that
without this minimum of distance I would have problems even if the room
treatment was extensive. He did not express his advice in this precise
context, but rather in the context of creating a "natural reverberant field"
and the use of diffusion to help create one, but obviously the topics overlap.

Put another way... if you only have 8' behind your head, by the time you have
made the rear of the room sufficiently absorbent to avoid smearing (by
reducing the amplitude of the reflected waves), you have also truncated the
overall reverb trails to a point that unnatural attenuation will arise...
that is to say the room starts sounding too dead.... and therefore unnatural.

On a side note, I think that 12' as a minimum target for behind the head
distance is a better figure to quote as this provide barely adequate room for
some type of producer's desk, a bit of floor space for foot traffic, a coffee
table and a couch to fit between the mix position, and the rear wall. Of
course this isn't a matter of acoustics, but it is relevant to overall design.

The draft beer system must of course be within reach of the mix position, and
thus is more a question of minimum room width.... but that's a thread of a
different color.

Scott R. Foster

[archive]

From: "Greg" <gfreitag@x??x.xxxx
Date: Fri Nov 19, 1999 2:32 am
Subject: Re: new article

>From: Dan Nelson <dprimary@e...>
>My mistake, I was combining ITD gap and Haas effect into the Haas
>effect. Do you have any insights into ITD gap and the Haas effect?

A fairly good definition can be found at:
http://www.rane.com/par-h.htm

If we look at the example of two speakers with a mono source our
brains produce a phantom source in the middle. If we start adding
delay to one of the speakers the center image will begin to shift to
the undelayed speaker. At some point (about 20ms) the delay will
be great enough that the sound will now appear to be coming only
from the undelayed speaker. As everybody knows it is also possible
to shift the phantom image by changing the level of one of the two
speakers through the use of a pan pot or balance control.
So it is possible to shift the apparent source of the image by:

1) Using short time delays
2) Level differences between the two sources

An early reflection can be thought of as a separate speaker delayed in
time. If we imagine an ideal situation where the reflecting surface
has an absorption coefficient of 0, (100% reflective), and also
imagine a speaker that has uniform polar response, in other words
the off axis response is the same as the on axis response. Since the
reflected sound would be identical to the sound produced by the
speaker, we would now hear a phantom source some place
between the two sources. In the real world the spectral response
of the reflections will not be the same as that of the speaker on axis
for two reasons:

1) The off axis response vs. frequency of the speaker will not be the
same as the on axis response.
2) The absorption coefficient of the reflecting surface vs. frequency
will vary.

The ratio of the direct to reflected sound is now frequency dependent
and our perception of the location of the sound becomes dependent
on the frequencies involved. The complex sounds we listen to are
composed of many frequencies which appear smeared between the
the speaker and the reflected image.

Greg

[archive]

From: SRF7@x??x.xxx
Date: Fri Nov 19, 1999 3:49 am
Subject: Re: new article

In a message dated 11/18/99 9:56:19 PM Eastern Standard Time,
gfreitag@s... writes:

<< 1) The off axis response vs. frequency of the speaker will not be the
same as the on axis response.
2) The absorption coefficient of the reflecting surface vs. frequency
will vary. >>

Also locational cues are absolute phase dependent (to some extent)... its is
possible to create apparent sound locations outside the stereophonic field
(not at or between the speakers) by manipulating absolute phase ... freaky
world.. init it?

Scott R. Foster

[archive]

From: Dan Nelson <dprimary@x??????xxxxx.xxxx
Date: Fri Nov 19, 1999 5:41 am
Subject: Re: new article

So the 20 ms ITD gap allows you to hear the time domain localization cues
from the recording room. Correct?
This must be were the 19-20 ms 11-12 feet minimum first reflection comes
from.

Does this also work when the recording is from a concert hall and the first
reflections would by greater then 20 ms? Or is 20 ms close the max delay
needed to locate a sound far left or right because of the human head, and
the reflections do not help localization's cue if you are in the same room
as the source.

Do you know if any books on psychoacoustics get more in depth on the Haas
effect? most of the books on acoustics have one line explanations. The rane
page is the best I've seen.

Dan Nelson

[archive]

From: Tony@x??x.xxxxx.xx.xxxxxxxxxxxxxx)
Date: Fri Nov 19, 1999 7:29 pm
Subject: Re: new article

The International Telecommunications Union (ITU) and the European
Broadcasting Union (EBU) have recommendations for rooms for critical
listening. These include a requirement on early reflections:
levels of reflections earlier than 15ms relative to the direct sound
should be at least 10dB below the direct sound for all frequencies in
the range 1kHz to 8kHz.

This seems adequate to achieve good stereo imaging. However it is not
always easy to achieve even this limited requirement. Absorption is
not always enough, it is often necessary to use some angled surfaces
to reflect sound away from the listening position.

Once you start killing reflections you have got to kill all the major
ones. Leaving just one biggish one can be worse than lots. Don't
forget the desk top.

As has been said, you can get a good indication of what reflections
need to be dealt with with a mirror. But for professional work, to
find out whether you have really killed the reflections you need an
FFT analyser or system such as MLSSA.

--
Tony Woolf
(Tony Woolf Acoustics)

[archive]

From: Tony@x??x.xxxxx.xx.xxxxxxxxxxxxxx)
Date: Fri Nov 19, 1999 7:46 pm
Subject: Re: new article

In message <002201bf3236$5415c980$2f0b49d8@G...> "Greg" writes:
From: "Greg" <gfreitag@s...>
> So it is possible to shift the apparent source of the image by:
>
> 1) Using short time delays
> 2) Level differences between the two sources

But if you shift it off centre using one method, and bring it back
again using the other method, you get an unstable image. That's why
it's important to have the speakers exactly equidistant from the
listener. You can't really compensate for a time difference with a
balance control.

--
Tony Woolf

[archive]

From: Jon Best <jrbest@x???xx.xxxx
Date: Fri Nov 19, 1999 3:05 pm
Subject: Re: New Article

Now, I am definitely a relative newbie to acoustics, but is there a reason
_not_ to use thicker foam, if you've got room? I've been in a number of
control rooms that use a good bit of 2" and are pretty light on thicker foam
and bass trapping,
and they all seem to share the same ringy midrange. In my limited experience,
I've preferred rooms that used 4"+ foam pretty much wherever they needed foam
at all (again, if there's enough room that it's not awkward).

Jon Best
Sales Weasel From Mars

[archive]

From: "Dave Martin" <dave.martin@x?????xxxx.xxxx
Date: Fri Nov 19, 1999 3:28 pm
Subject: Re: New Article



Jon,

I think that foam is a different issue than 703; I assume that you're
talking about the Auralex stuff (or something like it) - an acoustical foam,
while 703 is a dense fiberglass. I think that the figures would be quite
different for the two..

Dave Martin
DMA, Inc.
Nashville, Tennessee
dave.martin@n...

[archive]

From: Dan Nelson <dprimary@x??????xxxxx.xxxx
Date: Fri Nov 19, 1999 4:50 pm
Subject: Re: New Article

2" 703 performs close to how 6" of foam performs. I have a hard time working
in a room with just 2" foam in the room. They are usually built by someone that
didn't do any research into acoustics, they just started gluing foam to the
wall. The foam manufactures could do a better job educating their customers.
Aurally tries with the acoustics 101 but most of it is about soundproofing.

Dan Nelson

[archive]

From: "CactusFire (A.S.)" <cactusfire@x??????xxxxx.xxxx
Date: Fri Nov 19, 1999 5:33 pm
Subject: Re: new article

I think that the 20 ms thing does not orignate in theory, but as the result of
some study where the halls with the best acoustics were studied and the
correlating factor was the 20 ms time delay before first reflections. If I
remember correctly, this is talked about in Malcolm's control room article, as
well as in two of the three F. Alton Books.

Andy

[archive]

From: Wayland Dong <wdong@x??x.xxxx
Date: Fri Nov 19, 1999 7:49 pm
Subject: Re: New Article


I don't know much about foam, but the main difference between 2" and 4"
703 is in the 125 Hz octave band (from the Owens-Corning numbers).
Neither is terribly effective below 100 Hz, and they are not too
different at higher frequencies. So 4" can be useful for bringing the
low bands in line with the desired RT curve, but is not necessarily
necessary (i.e., you should not automatically assume that 4" is worth the
extra space or cost, at least not everywhere in the room.)

Wayland

[archive]

From: "Greg" <gfreitag@x??x.xxxx
Date: Fri Nov 19, 1999 10:47 pm
Subject: Re: new article

>From: Dan Nelson <dprimary@e...>

>So the 20 ms ITD gap allows you to hear the time domain localization
>cues from the recording room. Correct?

If I am reading you correct then yes.

>This must be were the 19-20 ms 11-12 feet minimum first reflection
>comes from.

Yes but this is a simplification since the level of the reflections
need to be taken in account. Any early reflections that are 30
dB lower than the direct field can be ignored.

>Does this also work when the recording is from a concert hall and the
>first reflections would by greater then 20 ms? Or is 20 ms close the
>max delay needed to locate a sound far left or right because of the
>human head, and the reflections do not help localization's cue if you
>are in the same room as the source.

I am not sure how well the psychoacoustic's understands the mechanisms
involved. The 20 to 30ms was chosen by looking at the time-energy
plots for concert halls that were subjectively deemed to have superior
acoustics.

>Do you know if any books on psychoacoustics get more in depth on the
>Haas effect? most of the books on acoustics have one line
>explanations. The rane page is the best I've seen.

No, my understanding comes from putting together bits and pieces from
various sources. You might want to look at publications in JAES. Don
Davis was the person responsible for what is called the LEDE concept
(Live End Dead End). The idea is to reproduce the time-energy
signature of a concert hall, a simulator so to speak. We have only
been touching on one aspect of the LEDE room, the ITD gap. There
are other requirements on the first significant reflections which have
not been touched on and the requirement for diffusion. I will try to
locate the exact papers that were published by Don Davis sometime
this weekend and post them here. One more thing I would like to
address. Several people have refered to RT60, reverberation time
in a acoustically small room does not exist. By definition the RT60
is a statistical measurement that is made at some point beyond the
critical distance. The "critical distance" is the point in a room that
the level of the reflected sound is equal to the direct sound. In a
acoustically small room the critical distance does not exist.

Greg

[archive]

From: Dan Nelson <dprimary@x??????xxxxx.xxxx
Date: Sat Nov 20, 1999 7:51 am
Subject: Re: new article

That is my understanding as well.
But the absorption coefficient calculation for a rt60 is useful as a
starting point to figure out the needed absorption

Dan Nelson

Greg wrote:

> One more thing I would like to
> address. Several people have refered to RT60, reverberation time
> in a acoustically small room does not exist. By definition the RT60
> is a statistical measurement that is made at some point beyond the
> critical distance. The "critical distance" is the point in a room that
> the level of the reflected sound is equal to the direct sound. In a
> acoustically small room the critical distance does not exist.

[archive]

From: "Ty Ford" <tford@x????xxx.xxxx
Date: Thu Jan 1, 1970 4:59 am
Subject: Re: new article

>address. Several people have refered to RT60, reverberation time
>in a acoustically small room does not exist. By definition the RT60
>is a statistical measurement that is made at some point beyond the
>critical distance. The "critical distance" is the point in a room that
>the level of the reflected sound is equal to the direct sound. In a
>acoustically small room the critical distance does not exist.
>
>Greg

Hi Greg,

How does that work? I would think that if the room were small and
reflective enough that the reflected sound could easily be almost as large
as the direct sound (minus standard reflective losses). It would sound like
your typical bathroom. Not an ideal recording or monitoring environment, but
certainly possible. What am I missing?

Regards,

Ty Ford

Ty Ford's equipment reviews and V/O files can be found at
http://www.jagunet.com/~tford

[archive]

From: SRF7@x??x.xxx
Date: Tue Nov 23, 1999 6:09 am
Subject: Re: new article

In a message dated 11/20/99 9:18:46 AM Eastern Standard Time,
tford@j... writes:

> >address. Several people have refered to RT60, reverberation time
> >in a acoustically small room does not exist. By definition the RT60
> >is a statistical measurement that is made at some point beyond the
> >critical distance. The "critical distance" is the point in a room that
> >the level of the reflected sound is equal to the direct sound. In a
> >acoustically small room the critical distance does not exist.
> >
> >Greg

> Hi Greg,

> How does that work? I would think that if the room were small and
> reflective enough that the reflected sound could easily be almost as large
> as the direct sound (minus standard reflective losses). It would sound like
> your typical bathroom. Not an ideal recording or monitoring environment, but
> certainly possible. What am I missing?

> Regards,

> Ty Ford

Best I recall...

Sound encounters uncompressed (and unexpanded) air and uses energy to conform
the static (or at least incoherent) medium (air at rest, or at least not
conforming to the waveform) into a transferring medium (vibrating air
conforming to the waveform) as the wave travels (by virtue of creating
coherent conformance in the air mass to the waveform). Sound waves propagate
in 3 dimensions and therefore the energy necessary to propagate a given
amplitude wave rises to the 3rd power per unit of linear distance from the
sound source (or more rationally sound diminishes in power at the rate of the
inverse of the cube root of distance from the source ... that is to say
really fast). The foregoing notwithstanding, its not that simple, the
waveform encounters an additional impediment in that air gets warmer even
while transferring the wave (air is not a frictionless medium) and thus
absorbs some of the energy as heat. Suffice to say, it takes a lot more
power to make a given level of noise 10' feet away than it does 1' away, and
the change in requisite power exceeds the cube of the increase in linear
distance.

For this reason, even given Sabin numbers of zero (perfect reflection) for
the walls, sound power is dropping off at an enormous rate for reflected
sounds because such sounds are of course twice the distance to the wall
further away (past your head to the wall and then back) from you than the
sound source itself. Throw in a Sabin number for the wall and a bit of
humidity in the air and its gets worse ... at least as far as the propagation
of the sound is concerned.

Thus, you have to get a good distance away from the source for the sound
(called critical distance, the calculation of which is a matter of room
volume, but is altered by the frictional quality of air and the overall Sabin
content of the room) before there is any possibility of the perceived
amplitude of the reflected sound equaling the perceived amplitude of the
source.

In a small you room this never happens (the critical distance does not exist)
for the simple reason that you can't get far enough away from the sound
source for the reflected signal to reach parity in amplitude. Thus you can't
measure the time to reach a 60 dB drop from initial amplitude at the critical
distance because there is nowhere to stand to take the measurement (no
critical distance).

Its late... I hope this makes sense ... email me if it doesn't and I'll try
again.

Scott R. Foster