Spectrum analysis in JAMin is incorrect.

After changing the update frequency for the FFT analysis in JAMin, I 
thought I'd see if I could fix another long-standing complaint of mine: 
the graph never even gets close to the top of the window, so while there 
might be really interesting things going on at low volume levels, I 
can't see them.

It seems like the signal is getting some automatic gain control 
(compression) before the FFT stage, separate from the compression that 
JAMin is doing itself, to keep the visual of the signal visible. Passing 
several test tones through to see what they looked like, I discovered 
this in the 30-band spectrum page:

http://www.naptastic.com/david/jamin-graph-wrong.png

The input being analyzed is white noise. It displays correctly on the 
HDEQ page, but here, well, that ain't what white noise looks like. (You 
can ignore the peak indicators being in weird places; that's from the 
other test tones I ran.)

So, I'm gonna jump back in the code right now, but I thought y'all might 
want to be aware, there are some gremlins in there that it seems no one 
has noticed before.

------------------------------------------------------------------------------
Enable your software for Intel(R) Active Management Technology to meet the
growing manageability and security demands of your customers. Businesses
are taking advantage of Intel(R) vPro (TM) technology - will your software 
be a part of the solution? Download the Intel(R) Manageability Checker 
today! http://p.sf.net/sfu/intel-dev2devmar