Re: [MP3 ENCODER] MP3 encoding speed : LAME XING

2000-10-09 Thread engdev

 
  Using 160kbps for both LAME and Xing, encoding "Dire straits - telegraph road"
 
  LAME takes about 1.5 times longer than Xing.

 Is that the MMX version of LAME?

 Mark Powell - UNIX System Administrator - The University of Salford

Mark,

I don't know if it is a MMX version - how can I tell?

-help reports :
version 3.86
Win32 binaries from www.chat.ru/~dkutsanov/

Owen

--
MP3 ENCODER mailing list ( http://geek.rcc.se/mp3encoder/ )



Re: [MP3 ENCODER] MP3 encoding speed : LAME XING

2000-10-06 Thread Mark Powell

On Fri, 6 Oct 2000, engdev wrote:

  You're right Mark, compared to Lame 387 MMX --abr 128 Xing is only two
  times faster Bo)
 
  Regards,
  Wim Speekenbrink
 
 Using 160kbps for both LAME and Xing, encoding "Dire straits - telegraph road"
 
 LAME takes about 1.5 times longer than Xing.

Is that the MMX version of LAME?

Mark Powell - UNIX System Administrator - The University of Salford
Academic Information Services, Clifford Whitworth Building,
Salford University, Manchester, M5 4WT, UK.
Tel: +44 161 295 5936  Fax: +44 161 295 5888  www.pgp.com for PGP key

--
MP3 ENCODER mailing list ( http://geek.rcc.se/mp3encoder/ )



Re: [MP3 ENCODER] MP3 encoding speed : LAME XING

2000-10-05 Thread engdev

 You're right Mark, compared to Lame 387 MMX --abr 128 Xing is only two
 times faster Bo)

 Regards,
 Wim Speekenbrink

Using 160kbps for both LAME and Xing, encoding "Dire straits - telegraph road"

LAME takes about 1.5 times longer than Xing.

I thought the difference was greater, but I had been dealing with mono files
back then :)

Owen.

--
MP3 ENCODER mailing list ( http://geek.rcc.se/mp3encoder/ )



Re: [MP3 ENCODER] MP3 encoding speed : LAME XING

2000-10-01 Thread Wim

You're right Mark, compared to Lame 387 MMX --abr 128 Xing is only two
times faster Bo)

Regards,
Wim Speekenbrink

- Original Message -
From: "Mark Taylor" [EMAIL PROTECTED]
To: [EMAIL PROTECTED]
Sent: Saturday, September 30, 2000 6:19 AM
Subject: Re: [MP3 ENCODER] MP3 encoding speed : LAME  XING


|
| 
|  Well, I just compared speed: Xing is about 4 times faster than Lame in
VBR
|  mode.
| 
|  Regards,
|  Wim Speekenbrink
| 
| 
| Compare to "lame --abr 128".
| I bet Xing's VBR mode is much closer to LAME's ABR than LAME's VBR.
|
| Mark
|
| --
| MP3 ENCODER mailing list ( http://geek.rcc.se/mp3encoder/ )
|


--
MP3 ENCODER mailing list ( http://geek.rcc.se/mp3encoder/ )



Re: [MP3 ENCODER] MP3 encoding speed : LAME XING

2000-09-29 Thread Zia Mazhar

 The Xing MP3 encoder is exceptionally fast. Do they cut corners to get such high
 performance? Is their code written in assember?

Who knows how did they program it... But they did sacrifice some quality.



 What (if any) chance of getting similar speed out of LAME?

LAME is pretty fast too. I haven't benchmarked and compared both, though. Would you
please do it and post the results? :-)



 The Xing encoder is quite limited with regard to input sampling frequencies and
 output bitrates. Has anyone done testing on quality?

LAME definitely delivers higher quality than Xing. Xing isn't the worst encoder, but
isn't the best either.



 Owen


- Zia


--
MP3 ENCODER mailing list ( http://geek.rcc.se/mp3encoder/ )



Re: [MP3 ENCODER] MP3 encoding speed : LAME XING

2000-09-29 Thread Gabriel Bouvigne

 The Xing MP3 encoder is exceptionally fast. Do they cut corners to get
such high
 performance? Is their code written in assember?

 What (if any) chance of getting similar speed out of LAME?

 The Xing encoder is quite limited with regard to input sampling
frequencies and
 output bitrates. Has anyone done testing on quality?

The Xing encoder is not a reference in term of quality. It's right that it's
fast, but now the new FhG engine is as fast as xing, but with a better
quality.
Yes, Xing cuts corners. An example is that it only uses long blocks, and
thus doesn't compute short ones.
Yes, it contains a lot of assembler and mmx code.

And yes, there is a chance to get similar speed but with a lot higher
quality with lame, one day. The priority with Lame is quality, so quality
improvments are made first. Before optimizing to mmx a function, we must be
sure that this function won't be upgraded soon, this is why there is only a
few mmx on Lame.


Regards,


--

Gabriel Bouvigne - France
[EMAIL PROTECTED]
mobile phone: [EMAIL PROTECTED]
icq: 12138873

MP3' Tech: www.mp3-tech.org


--
MP3 ENCODER mailing list ( http://geek.rcc.se/mp3encoder/ )