All information in these pages is copyright (c) 1989-2003 by Roger Nichols.
All rights reserved. Permission for personal reference only, and may not
be reproduced by any method without written permission.
Better Late Than Never, But Better Never Late
by Roger Nichols
(What the hell does he mean by that? Does it have some bearing on digital
audio timing and whether the AES stream is clocked from the leading edge
or the trailing edge? Maybe we should just ask?) "Hey, Roger, what
does this mumbo jumbo mean?"
Well, I'm glad you asked. Mostly, it pertains to this column being late
again. It seems as though every four weeks or so another month goes by.
I have a good excuse this time, as I was out in California visiting friendly
manufacturers in Silicone Valley. I also found out that it is not named
after the area just above the female sternum.
Digidesign is hot on the release schedule for Pro Tools 4.0. Totally re-written
from the ground up, and chock full of new features. One tour around the
block with the new software and you won't know how you got along before.
PCI versions of the Pro Tools hardware has been shipping for a few months
now, with more PCI hardware on the way. Also new control surfaces are in
the works that will make Pro Tools feel like a hardware console instead
of a software program. I am working on the full review, so don't change
that channel.
Sonic Solutions was also on the visiting list, although they are a little
further North, in Novato, about 25 minutes past the Golden Gate bridge.
Sonic has developed the first authoring tools for DVD. I saw a demonstration
of the system and the first commercial software to be sent to the DVD pressing
plant. DVD is going to be an impressive format. Your favorite movies in
wide screen format with so many alternate language tracks that it will impress
even the Hippo families. (If you get that one, e-mail me. I'd be impressed.)
Multiple audio tracks of up to 96kHz 24 bit to supply all of the surround
sound you can stand is what mostly caught my attention.
Record companies saw lots of catalogue reissues when CDs superseded the
LP. Now they can re-release everything in multi-channel surround sound.
I am actually thinking about doing a few of them myself. With discreet surround
channels like DTS or Dolby Digital, you can pan something just where you
want it and it will stay there and sound good.
Sonic has also dropped their prices for an entry level system, so I see
the work station wars really heating up in the very near future. There will
be a full Sonic Solutions review soon also.
It Gives Me The Jitters
This jitter thing just won't go away. I still hear of people who insist
that their digital audio sounds better when stored on one certain brand
of hard disk. I hear tales of digital audio sounding better when transferred
with SDIF-2 cables (three separate cables for left, right, and word clock
with BNC connectors on each end) instead of AES. A major Japanese company
who shall remain nameless (JVC) says that digital audio is better if transferred
over Aluminum wires. They say that Aluminum works better than Gold or Silver.
Maybe they hope you will ingest the Aluminum, contract Alzheimer's disease
and forget that they made these claims.
Some people say that they can hear the difference between a CD-R that was
cut at 1X and a CD-R that was cut at 4X. The glass masters at the CD plant
are usually cut after transferring the audio from the CD-R at 2X. These
same people say that the CDs pressed don't sound as good as the ones that
are pressed from a 1X transfer to the glass cutter.
The theory is that the slight jitter in the spacing of the pits on the CD
cause jitter in the final digital audio output. After looking at the schematics
for a number of different CD players from Sony, Fisher, Denon, Kenwood,
JVC, Marantz, and Luxman, I can only conclude that THIS IS NOT POSSIBLE!!
If your player's output jitters because of pit spacing problems then IT
IS BROKEN, GET IT FIXED OR BUY A NEW ONE!
CD player timing is derived from an internal crystal that clocks the digital
audio out of memory, not directly off of the disc. This crystal provides
the reference for the servo that rotates the disc. The disc turns fast enough
to decode the pits and place the data in memory at the same rate that the
crystal is emptying the memory. If the memory is filling up, the disc slows
down, and if the memory is emptying, the disc speeds up.
Every piece of digital audio gear works exactly the same way. With a Tascam
or ADAT, if the memory is getting too full, the rotating heads slow down.
In all DAWs (Digital Audio Workstations), if the memory is getting empty,
the computer gets more data off the disk. With a Sony 48 track or a Mitsubishi
32 track, the tape speeds up or slows down as necessary to feed the memory
bin.
And finally:
Jitter doesn't matter except for A/D and D/A conversions. If the receiver
is properly decoding what the transmitter is sending, and a one or a zero
going in gives you a one or a zero coming out then it makes no difference
whether the jitter is less than a nanosecond or more than a fortnight. As
long as the data is buffered and re-clocked with low jitter for the conversion
process, it just doesn't matter.
When a bit is transmitted, the receiver attempts to discern the value during
the middle portion of the bit time. There can be a lot of jitter and still
give you the proper value of the bit. As an example, suppose our bit clock
was from 6am to 6pm. If you were told to go outside and look at the sky
somewhere around 12:00 and determine whether it was day or night, you could
have as much as five hours of jitter and still come up with the right answer.
If a one or zero in was not the same as the one or zero out, then computers
would not work. You would not be able to illegally copy software for your
friends. You would not be able to copy from one hard disk to another without
your bank balance changing randomly. ATM machines would dispense one amount
of cash and debit your account by a totally different amount (well maybe
that one was a bad example).
If you are getting bad transfers, then it is faulty hardware or bad cables
that are causing total destruction of the data stream. If an audio file
on one medium compares bit for bit with an audio file on another medium,
then it is the same file. They are the same, they will sound the same, there
is no argument that will convince me otherwise. End of discussion.
(It sounded like he was pissed. Did he ever explain the title to this column?
Not really, but I'm not going to ask. You do it!)
Return to EQ index