Am Fri, 02 Mar 2007 18:18:13 +0100 schrieb "Robert Leiking" <[EMAIL PROTECTED]>:
> > Hi Thomas, > > is there also a fixed point implementation available from mpglib or mpg123?? > Or is this SHARP DSP already in fixed point? The SHARC is 32-bit float, we even changed mpglib to skip the conversion to int for that. There is some code / macros in mpg123 for fixed point but they don't seem to work atm or are not complete. As I see it you use mpg123 in floating point mode - on an arch without floating point, you use kernel math emulation. I didn't benchmark how well this performs (yet... I have a 486 and a 386 system to play with - on the 486 mpg123 is clearly faster than mpg321, which uses mad). So if you want fixed point, you indeed better look into mad, because that's the point they are very proud of: fully fixed point math. For normal PC cpus, using the floating point math makes more sense - at least with mpg123 it is more efficient. Alrighty then, Thomas. _______________________________________________ mp3encoder mailing list [email protected] https://minnie.tuhs.org/mailman/listinfo/mp3encoder
