Re: [MP3 ENCODER] Speeding up window_subband

[Takehiro Tominaga]

Hi All,

I've merged Naoki's new IDCT code for filter subband into my tree,

but..... but it is not faster than the old brute force calculation
algorithm, and obviously slow (about 20% for filtering itself, 5% for
overall encoding process) on my celeron 464 linux box.

I'll optimize the coding and put it on the web or merge it to the
current LAME tree. Pls test it various environment..

uum, it's hard to speed up.... It seems simple algorithm is fast on
the current computer, even it containing many multiplication.

BTW, Naoki,

>>>>> "N" == Naoki Shibata <[EMAIL PROTECTED]> writes:

    Takehiro> Your new code looks like very similar to my snapshot(but
    Takehiro> is not on the net yet :p). It was mentioned in the first
    Takehiro> mail of "the road to the next(v4.00?)" thread.

    N>   I see. It seems that mdct_short and mdct_long are also
    N> speeded up by introducing faster algorithm.  Have you already
    N> made it?

No :) 
I wanna heard it, but I'm afraid it is slow too...
---
Takehiro TOMINAGA // may the source be with you!
--
MP3 ENCODER mailing list ( http://geek.rcc.se/mp3encoder/ )