Localizing low frequency sound

Archived from groups: rec.audio.tech (More info?)

In article <hr8641tjol8de9r9q9g42cmrjok62f09vt@4ax.com>,
Per Stromgren <per.stromgren@telia.com> wrote:

> Scott Dorsey wrote:
> > The human ear can't perceive real imaging at 20 Hz, but at 100 Hz it sure
> > can, and most of the satellite-sub systems have substantial subwoofer output
> > well above that.
>
> Are you sure about that? I have seen very different data about the
> lower frequency at which we can localize sound sources. One difficulty
> when trying this is that we often use loudspeakers that have gross
> distortion at bass frequencies, which makes them generating rather
> loud second and third tones harmonics. I do not accuse of such a
> mistake, but I'm curious as to where you got your data.
>
> My own experiments with Quad ESL 63s does not support your 100 Hz,
> rahter the vicinty of 200 Hz, but they are not scientific, just hobby
> work in my living room. My B139-driven tranmisson line has higher
> distotion(*) and give the effect I refer above.
>
> Per.
>
> *) The Quads generate lower bass distortion than the B139 when driven
> with spec, but the B139 can play louder, of course.
>
>

IME/IMO A simple single or dual sub setup in a large-ish room *may* be
localized, especially when pushed near its limits mechanically. But an
array of said subs or an array of IB subs with effortless reproduction
can be very engulfing and hard to localize.

All the YMMV's and whatnot apply as these are my ears and feelings.
Pyschoacoustics play a big role after all.

hth,

--
Cyrus

*coughcasaucedoprodigynetcough*

Archived from groups: rec.audio.tech (More info?)

"Per Stromgren" <per.stromgren@telia.com> wrote in message
news:hr8641tjol8de9r9q9g42cmrjok62f09vt@4ax.com
> Scott Dorsey wrote:
>> The human ear can't perceive real imaging at 20 Hz, but at 100 Hz
it
>> sure can, and most of the satellite-sub systems have substantial
>> subwoofer output well above that.
>
> Are you sure about that? I have seen very different data about the
> lower frequency at which we can localize sound sources. One
difficulty
> when trying this is that we often use loudspeakers that have gross
> distortion at bass frequencies, which makes them generating rather
> loud second and third tones harmonics. I do not accuse of such a
> mistake, but I'm curious as to where you got your data.

IME Scott is pretty close to the truth, based on lots of experience
with robust and low-distortion subs.

The hidden variable here is the relative positioning of the sub and
the sats, as well as their size.

> My own experiments with Quad ESL 63s does not support your 100 Hz,
> rather the vicinty of 200 Hz,

This statement is consistent with what Scott actually said depending
on how you read the statements. If one reads your statement it can be
construed to mean that you heard imaging problems at 200 Hz, which
might be true of a system that also had imaging problems at 100 Hz.

>but they are not scientific, just hobby
> work in my living room. My B139-driven tranmisson line has higher
> distotion(*) and give the effect I refer above.

> *) The Quads generate lower bass distortion than the B139 when
driven
> with spec, but the B139 can play louder, of course.

Archived from groups: rec.audio.tech (More info?)

On Fri, 25 Mar 2005 05:44:19 -0500, "Arny Krueger" <arnyk@hotpop.com>
wrote:

>
>IME Scott is pretty close to the truth, based on lots of experience
>with robust and low-distortion subs.
>
>The hidden variable here is the relative positioning of the sub and
>the sats, as well as their size.

OK. You mean that in practice the localization of bass is dependent on
these variables?

My question could be seen from two wiewpoints: 1. what is a good upper
cut-off frequency for a sub concerning localization? and 2. what is
the theoretical low frequency where we can localize the sound?

If the answers are different, why?

When I wrote my question, I thought of number 2, actually. I haven't
seen any controlled experminets of this second kind, but I don't have
access to JAES, were I suppose is the first place to look.

Per.

Archived from groups: rec.audio.tech (More info?)

Per Stromgren wrote:

> My question could be seen from two wiewpoints: 1. what
> is a good upper cut-off frequency for a sub concerning
> localization?

I'm running a pair of ATC 9" studio bass boxes off of the centerline
with the pair of KEF Coda's that provide audio for this desktop.
Cross-over is 115 Hz, third order, and the bass box setup is tucked away
behind furniture and aimed at the wall near the corner for improved room
loading.

Delaying their signal so that it no longer arrives first would perhaps
help, but as it is they have to have their level set a couple of dB's
lower than it ought to be to prevent "onesidedness" in the bass from
being obvious. It would be better to separate the bass cabinets, but
that would make the placement of the work table less comfortable.

With your ESL's you definitely should do what I understand you to have
done: place one sub behind or below each. I am using a small Sentec
x-over btw. .... As I recall the ESL specs 115 Hz 18 dB pr. octave is a
very sensible x-over to use with them.

There was also a neat electronic x-over, based on a National
Semiconductor application note, in Radio & Television (swedish
magazine), somewhere around 1977 or 1978.

> and 2. what is
> the theoretical low frequency where we can localize the sound?

> If the answers are different, why?

One issue that comes to mind is that a subwoofer setup has to cross-over
smoothly to some other loudspeakers, cross-over issues could be one way
of stumbling over "unassumed audibility".

> When I wrote my question, I thought of number 2, actually. I haven't
> seen any controlled experminets of this second kind, but I don't have
> access to JAES, were I suppose is the first place to look.

Try taking a closer look at http://www.aes.org .... oh, and you could
also try contacting the swedish chapter.

> Per.


Kind regards

Peter Larsen

--
*******************************************
* My site is at: http://www.muyiovatki.dk *
*******************************************