[vorbis] OGG IS BEST, BUT UNUSABLE! (very important)]

[Jack Moffitt]

had to share this one :)

jack.

----- Forwarded message from [redacted] -----

Delivered-To: jack at i.cantcode.com
From: [redacted]
To: <feedback at vorbis.com>
Subject: OGG IS BEST, BUT UNUSABLE! (very important)
Date: Sat, 27 Jan 2001 10:08:38 +0100
X-Priority: 3
X-MSMail-Priority: Normal
X-Mailer: Microsoft Outlook Express 5.50.4133.2400
X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4133.2400

Hello,

I've been studying mp3 sound quality for a long time and it was a great surprise to find an other decent encoder.
OGGVorbis has superior sound quality to mp3 ABR/CBR 128, that's a fact.
So why did I call it 'UNUSABLE'?!?

The whole failure is hidden in OGG's philosophy. Let me put this in such way:
Suppose I had a very diffucult piece of music (but of course I don't know it, cuz it rather can't be determined by ear).
I encode this piece with Lame VBR -V 1 and I'M COMPLETELY SURE IT'LL BE ABOUT 98% transparent to the original. Sometimes the piece
will be encoded at 200 kbps, sometimes 105 kbps (these are real cases). It'll almost never (98%) contain artefacts, even the
latter-105kbps file.
Now, the same difficult piece which was classified by Lame's sophisticated algorithms as a 200kbps put to OGG with ABR 128kbps.
What the output will be like?:
- it'll be encoded at about 145 kbps - not 128 kbps (which is still not enough, even for OGG)
- it'll sound as a piece of s... - completely distorted and muffled
Of course You may say: "You could have taken 192 kbps in OGG!"
But now what about the mp3 105kbps-piece? Ogg would encode it transparently probably using even 95 kbps! Why shuld I take the 192
bitrate then? So every piece has to be encoded three times and compared to original...

Am I sayin that OGG is worse that Lame VBR? NO! I'm just trying to convince You that the whole OGG's ABR (not real VBR, cuz it's
only Average Bit Rate) philosophy is just stupid!! You may say that it is easier to use for all those lame mass of people. Again NO!
Such people aren't interested in anything other than mp3 CBR 128. Probably 90% of Vorbis users are real pros/audiophils, who're
waiting for an alternative. There are no more professional attitude than the REAL VBR, where one determine the lowest and highiest
bitrates as well as the algorithm (-V 0 to -V 9 in Lame). All right, but where to take the VBR procedures from? Well, please open
latest lame's 3.87 source file, and copy-paste appropirate selection;). That's just it. Lame is also open source, and VBR routines
aren't Fraunhofer property, but programmers' who contributed to lame. Of course I don't want You to get rid of existing 5 ABR
routines.

I hope You found my opinions helpful in further OGG development. I wish OGG was the best encoder ever, and I wish You to succeed in
doing it.

Regards!

[sig redacted]

----- End forwarded message -----

--- >8 ----
List archives:  http://www.xiph.org/archives/
Ogg project homepage: http://www.xiph.org/ogg/
To unsubscribe from this list, send a message to 'vorbis-request at xiph.org'
containing only the word 'unsubscribe' in the body.  No subject is needed.
Unsubscribe messages sent to the list will be ignored/filtered.