[LAD] twice as loud

[Charles Henry]

On Tue, Jul 27, 2010 at 11:57 AM, Ralf Mardorf
<ralf.mardorf at alice-dsl.net> wrote:
> On Tue, 2010-07-27 at 11:10 -0500, Charles Henry wrote:

>> Because psychoacoustics just hasn't been defined in a way to make hard
>> numbers stick.  The tendency in psychoacoustic experimental design is
>> to use discrete conditions (which gives better experimental power) in
>> order to show that an effect exists.  But this way, any given
>> experiment can't produce results that cover the whole space.
>> Generalization and extrapolation are limited.

> Masking theories used by audio codecs to compress audio signals don't
> work for people who are trained in listening, e.g. MP3 at any rate is a
> PITA. Trained people (here where I do live) do always here a loss in
> blind tests.

It's a problem for current models to accomplish compression without
losing some part of the audio quality.  Do you think there could be a
better (or best) basis for reducing the complexity and bitrate of
sound?  or is it just plain impossible?

> Before the brain does "math", are there any other senses involved to the
> interpretation of the input given by the ears? (Btw. I'm sure that math
> is just part of nature and can't describe nature, because it's just a
> part.)
>
> Regarding to the topic that there are two sound sources and you are
> thinking about a relation, try to imagine people who are autistic or who
> are 'normal', but having a panic attack, or try to remember a situation,
> when you barely were able to escape an accident. The filtering is
> completely different. Everybody of us is able to focus allegedly masked
> "things".

I agree--but I'll put a little more on the table here.  A strong model
is able to describe not only the big trends but the contextual,
situational, and personal sources of variation.  A model of that kind
of scope is very far off, in my mind.  I would be happy to have a
weaker model that generalizes well first.

> Perception, the interpretation of the input by all senses will change,
> regarding to the context, for audio even regarding to the tilting of
> your head to the body.
>
> Recordings and digital audio virtualization always has the lack of the
> experienced context. If it should be possible to completely gather
> everything of nature by math for specific situations, there still will
> be the context, a situation were everybody brains is able to count the
> peas that drop to the ground, after the glass of peas fall down to the
> ground.
>
> Remember your own experiences when you had or nearly had an accident.
> Time seemed to be slower and silent, but important sounds (regarding to
> your survival) become loud, while loud, but unimportant sounds become
> silent.
>
>