[paranoia] Very slow paranoia on 24x CDROM

[Monty]

------- Forwarded Message

From: Wolfgang Weisselberg <weissel at netcologne.de>
Date: Sat, 2 Sep 2000 23:14:27 +0200

Trying to kill the keyboard, Rohan Beckles
(rohan.beckles at ndirect.co.uk) produced 2,5K in 70 lines:

> I have two machines running paranoia III version 9.7.  One is PII 266
> with a 32x Mitsumi FX-320S (I think), the other is a laptop with a PIII
> 650, with an NEC CD-224E 24x.

> No matter what I do, I can't get either of these drive to run anywhere
> near full speed.

The 24x and 32x is a rather theoretical speed. :-( What
happens is that a normal audio player will spin the CD with
varying speeds to guarantee a steady ~150k/sec of music data,
because the tracks at the outside (towards the end) of the CD
are much longer and thus one revolution gives you something
like 2.5 times the data than one on the inside (begin) of
the CD.  That is called CLV, constant linear velocity (the
speed measured over the head is constant)

Earlier CD-ROM drives did the same.  But if you do *not*
vary the speed, you do not have to speed up and brake down
the CD as much (which costs lots of time) and you gain a
speed boost towards the end of the CD.[1] This is called CAV,
constant angular velocity (meaning it spins at a constant RPM).
Of course Marketing brands the CD-ROM drives after the max
speed reached at the outermost tracks of the CD-ROM[2] and
rounds a bit as well ...

Now, what paranoia does is reading *audio* data, but not
the data as converted by the CD-ROM drive to analog data,
but rather the digital data as it is on the CD itself.  Since
music is rather differently encoded compared to computer data
(completely different protocol and all, including much less/no
error correction, at best unreliable fine positioning of the
head due to the prootocol, and so on), CD-ROM drives do handle
them differently ... and not all can handle digital reading
of audio data.

Now, with that in mind, you should not expect more than say 16x
or 12x throughput due to the CAV.  Then you will hit overhead
as the drive tries to read the digital data.  If you get a
6x-8x reading digital audio, you have a _fast_ drive.

> The net effect of this is that (using krabber) the laptop can process a
> WAV file faster than the CDROM can read the data from the CD drive.

That ought to be common, you do have a nice bit of CPU power
there.  Unless you do VBR encoding, that is. :-)

- -Wolfgang

[1] After all, you cannot spin CDs much faster than "40x",
    because that would actually rip the CD to pieces, at those
    speeds the CD is already being stretched quite a bit from
    the centripetal forces.  If you want to break that barrier,
    you better invest in more heads or something like that.
[2] We do not talk about audio data here!

------- End of Forwarded Message

--- >8 ----
List archives:  http://www.xiph.org/archives/
Paranoia homepage: http://www.xiph.org/paranoia/
To unsubscribe from this list, send a message to 'paranoia-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.