Early reflections: 5 ms time window for sound source location?

[[email protected]]

The loudspeaker placement method as proposed by Audio Physic (http://
www.audiophysic.de/aufstellung/index_e.html) is based on the idea that
source location is happening within the first 5 ms, after which first
reflections are suppressed and don't disturb any more. The basis for
this method is very likely to be found in Audio Physic's AES paper
4246 (http://www.aes.org/e-lib/browse.cfm?elib=3D7529) where it is said
that for delays of 1-5 ms a broadening of the image and sometimes
image shift is occurring. Reflections in the sub 5 ms range should
therefore be minimized.

Reference is made to the chapter "Law of the first wavefront" in
Blauert's "Spatial hearing", 1974 edition. I checked that chapter of
the 1974 and 1983 editions and could not find any mentioning of such 5
ms window. Summing localization is mentioned, but the time window is
about 1 ms, then the precedence effect takes over.

My question is, where could this 5 ms time window come from? Any
particular literature?

Klaus

[Arny Krueger]

> wrote in message

> The loudspeaker placement method as proposed by Audio
> Physic (http://
> www.audiophysic.de/aufstellung/index_e.html) is based on
> the idea that source location is happening within the
> first 5 ms, after which first reflections are suppressed
> and don't disturb any more. The basis for this method is
> very likely to be found in Audio Physic's AES paper 4246
> (http://www.aes.org/e-lib/browse.cfm?elib=7529) where it
> is said that for delays of 1-5 ms a broadening of the
> image and sometimes image shift is occurring. Reflections
> in the sub 5 ms range should therefore be minimized.
>
> Reference is made to the chapter "Law of the first
> wavefront" in Blauert's "Spatial hearing", 1974 edition.
> I checked that chapter of the 1974 and 1983 editions and
> could not find any mentioning of such 5 ms window.
> Summing localization is mentioned, but the time window is
> about 1 ms, then the precedence effect takes over.
>
> My question is, where could this 5 ms time window come
> from? Any particular literature?

Wherever it comes from, it can't be the whole story. I sense that a
timing-based model of spatial sensing is being promoted, which unfortunately
completely ignores HRTF's.

This statement from their web site:

"The spatial information relating to a sound source is mainly detected by
the brain in the first five milliseconds, everything which follows will at
first be ignored. Sound travels at 1.7 metres (5.6 ft) in five milliseconds.
If the sound of a reflection has to travel 1.7 m (5.6 ft) more than the
direct sound of the source, then it will have no influence on our spatial
perception."

is incredibly wrong when it says: If the sound of a reflection has to travel
1.7 m (5.6 ft) more than the direct sound of the source, then it will have
no influence on our spatial perception."

Of course, sounds delayed by more than 5 milliseconds can affect our spatial
perception. If strong enough and at the right frequencies it can completely
blur spatiality.

[Gary Eickmeier]

"Arny Krueger" > wrote in message
...

> Of course, sounds delayed by more than 5 milliseconds can affect our
> spatial
> perception. If strong enough and at the right frequencies it can
> completely
> blur spatiality.
>

No, not "blur" spatiality. The reflected sound is what CAUSES spatiality. I
recently set up a demonstration at our local audio club that showed how my
Image Model works. My test speakers are omnidirectional and have more output
to the rear than toward the audience. They were placed 5 feet out from the
front wall and the side walls (for a 10 mS delay). I demonstrated very
conclusively that not only were the sound images not destroyed, the
reflected sound (of the SAME frequencies as the direct sound) caused an
Auditory Scene to form between and behind the speakers, the sound coming
from points outside of the speakers themselves and having greater depth and
spaciousness. I showed this visually to the group with an orange cone that I
placed where the sound seemed to be coming from, as directed by members of
the audience. A panned human voice, for example, was heard to come from
about 4 feet behind the plane of the speakers, and it remained so even at
the stereo extremes left and right, the sound NEVER seeming to come from the
speakers themselves.

This is a very mysterious, highly sought after effect, but it is simply
caused by the reflected sound from behind and beside the speakers. It is a
grave mistake to kill those reflections with the schemes proposed in the
article under discussion.

Gary Eickmeier

[Trevor Wilson[_3_]]

Gary Eickmeier wrote:
> "Arny Krueger" > wrote in message
> ...
>
>> Of course, sounds delayed by more than 5 milliseconds can affect our
>> spatial
>> perception. If strong enough and at the right frequencies it can
>> completely
>> blur spatiality.
>>
>
> No, not "blur" spatiality. The reflected sound is what CAUSES
> spatiality.

**Incorrect. ANY reflected sound will cause 'blurring'. As anyone who has
experience with an anechoic (or near anechoic) environment will attest. Mr
Krueger is correct.

--
Trevor Wilson
www.rageaudio.com.au