Re: Another linear to db question
- From: "Chris P. [MVP]" <msdn@xxxxxxxxxxxx>
- Date: Fri, 25 Nov 2005 10:31:08 -0500
On Fri, 25 Nov 2005 15:11:58 GMT, Bob Masta wrote:
> I'm not certain I understand, but I think what
> you want is just an expanded scale. You use
> the same math to convert to dB, throw away
> everything below -20 (or whatever), and expand
> what's left to fill the available space. That's
> just a matter of shifting and scaling, after a
> -20 dB threshold test. Anything below -20
> is set equal to -20, so that reduces the range
> to 0 to -20. Then you just scale that up to
> fill the space you formerly filled with 0-96 dB,
> by multiplying by 96/20 or whatever.
>
> Is that what you meant?
This is exactly what I do in my apps BTW. I use either a 48 or a 60dB
scale depending.
There is one more option however, which used to be very popular in the
analog world. That is to present a log*log scale, which has the effect of
"squishing" the quiet end. You would have to play with the math a little
to see if this gives the effect you are after, but essentially take your
inital dB result, divide by a magic number (100?), take the log and
multiply the result to fit your scale.
.
- References:
- Another linear to db question
- From: Dave
- Re: Another linear to db question
- From: Bob Masta
- Another linear to db question
- Prev by Date: Re: Another linear to db question
- Next by Date: Re: Capture audio generated with Microsoft Speech SDK
- Previous by thread: Re: Another linear to db question
- Next by thread: Re: Another linear to db question
- Index(es):
Relevant Pages
|
