Re: armhf performance of vorbisgain

[Loïc Minier <lool@dooz.org>]

On Sun, Aug 22, 2010, Clint Adams wrote:
> This represents a rebuild with those flags of vorbisgain only, and not
> any of its library dependencies.

 Thanks for sharing the gnumeric spreadsheet.  Looking at the mean
 results, and summarising the percentage differences:

        sf      riku    hf

 real   100%    62.46%  46.89%
 user   100%    62.39%  46.69%
 sys    100%    86.40%  75.41%

 So assuming that's a fair comparison, "hf" consistenly wins over
 "riku":

 real: -24.93% time
 user: -10.88% time
 sys:  -12.72% time

 This benchmark is one encouraging data point, the more the merrier
 though  ;-)

   Thanks again for sharing these Clint,
-- 
Loïc Minier