linuxaudio.org

[James Morris]
>On 15 July 2010 18:23, Harry Van Haaren  wrote:

quoted text
>> http://en.wikipedia.org/wiki/Denormal_number

>>

>> A nice "quick" fix for these is:

>> if ( number < 0.000000001)

>> =A0=A0=A0=A0=A0 number =3D 0.0;

>>

>> ;-)

>

>Don't forget about negative numbers ;-)

>

>n =3D ((n > 0.0 && n < 0.000000001) || (n < 0.0 && n > -0.000000001)) ? 0.=

0 : n;

I can't seem to remember the most sensible numeric limit for this=20

branching variant of denormal removal -- besides, it obviously depends=20

on the context in question and the use of floats versus doubles (as=20

does the use of fabsf versus fabs in what follows) -- but I am quite=20

positive that the following line, which replaces the above one, is not=20

only easier to read, but also stands a good chance of executing=20

quicker on most contemporary CPUs due to the elimination of at least=20

one branching instruction:
=09n =3D fabs(n) < 1e-9 ? 0 : n;=20
Cheers, Tim