STI vs. ALcons

[Efren Pineda]

in sound system design, the usual approach in judging speech
intelligibility is to determine ALcons to be less than or equal to 10%.
when i attend a seminar handled by BOSE engineers they told us that they
do not follow ALcons principle because their speakers have low Q
(directivity factor) and gave them a value greater than 10%. so they use
the STI in judging the speech intelligibity in their design. they have
incorporate this in their sound system software, 'The Modeler', a MAC
based program that allows the designer to see whether the system will
provide better intelligibility once the system is installed.

is there an eqution that will give a specific value of STI? what are the
factors that affect this quantity?

we use the JBL Cadp2, also a sound system design software, in our
company that uses ALcons in judging speech intelligibility. so far it
helps us a lot in most of our installations.

anyone who wants to react on this?
it is very much appreciated since this will help us a lot in improving
our sound system designs.

thanks,

efren

[douglas....@crc.doc.ca]

In article <95093017...@scninet.org> efren....@scninet.org (Efren Pineda) writes:
>in sound system design, the usual approach in judging speech
>intelligibility is to determine ALcons to be less than or equal to 10%.
>when i attend a seminar handled by BOSE engineers they told us that they
>do not follow ALcons principle because their speakers have low Q
>(directivity factor) and gave them a value greater than 10%. so they use
>the STI in judging the speech intelligibity in their design. they have
>incorporate this in their sound system software, 'The Modeler', a MAC
>based program that allows the designer to see whether the system will
>provide better intelligibility once the system is installed.
>
>is there an eqution that will give a specific value of STI? what are the
>factors that affect this quantity?
>

Well I like THIS ruler better because it gives me the answer that I want...

Isn't ALcons at it's root a measure of how many syllables are missed by
a group of listeners when they take dictation from someone reading a
list of nonsense syllables, a high proportion missed indicating that
normal speech will be difficult to understand?

Unless BOSE are saying that the intelligibility of speech cannot be used
to measure the intelligibility of speech, there must be some confusion here.
My reference works on this subject are at home, I'll do some reading
overnight and get back to you.

[HARALD STEINDL]

Hello!

>in sound system design, the usual approach in judging speech
>intelligibility is to determine ALcons to be less than or equal to 10%.
>when i attend a seminar handled by BOSE engineers they told us that
they
>do not follow ALcons principle because their speakers have low Q
>(directivity factor) and gave them a value greater than 10%. so they
use
>the STI in judging the speech intelligibity in their design. they have
>incorporate this in their sound system software, 'The Modeler', a MAC
>based program that allows the designer to see whether the system will
>provide better intelligibility once the system is installed.

Does this help the customer in any way, whether you try to predict the
performance in this or the other way? No!
The customer wants to have a good (=understandable) system.

I am more into practical use than %Alcons and STI.
My experience is, (and I am not alone) that a low Q-speaker needs very
good or perfect room acoustics to provide good quality.

But go into a gymnasium or sportshall and install some 802's and hear
for yourself! ;-)))

I will stay with my Community's and other real-Q-systems.
People tend to mix No-Q with low-Q.

Bye

Harald

MOCOM GmbH, Vienna, Austria