Conference cookie::notes$archive:cd_v1

The 'sampling' frequency of 88.2 kHz is true in a sense, but this is
nothing unique. This method has been used by Philips/Marantz all the time
with a frequebcy of 176.4 kHz.

The disk itself contains samples at a frequency of 44.1 kHz. What these
'oversampling' players do, is to do the D/A conversion at a higher frequency
(I assume they interpolate to get the samples between, I have never seen an
exact description of the method).

The main advantagae is that you don't need those steep analog low-pass filters
that cause phase distortion at high frequencies.

Philips is claimimg to do some sort of digital filtering of the signal
before D/A conversion. Does anyone know any details?

Also, the player certainly doesn't have three lasers; rather the beam is split
into three and it may have three photodiodes or whatever they use.
This is a common feature too.

Never hear about double-sided CDs. Where will they put the label on?

As I understand it, the 3 laser beams are used to improve tracking.
If you are familiar with magnetic disk alignment techniques you
will see how it works.

One beam looks in the middle and the other two just to each side.
It is like driving your car down the middle of the lane by watching
the curb and the yellow line.

I have found that Tweeter has two kinds of salesmen.  The more
common kind don't really understand the technology.  The rarer
kind really do know what they are talking about, and will admit
the times they don't.  You have to find out which ones these
are and talk to them.

I know who to ask at my local Tweeter.  For other branches you
have to figure it out yourself.  Some smaller branches may
not have the expert there all the time.

I know about the Philips oversampling, they read 44.1K 16 bit samples
from the disk each second and digitally convert them into 176.4K 14 bit
samples.  Because they have more samples going into the D/A converter
the set of possible output sound has more constraints.  In particular,
the output in the frequencies in the range of 20-80 KHz is constrained
to a single value (and if the digital filter is perfect) there are no
signals in this range coming out.  This make a much easier job for the
analog filter to get rid of the high frequence stuff.

This may be what is happening, however the salesman was adimant that the
disk was reading those extra bits from the disk, and even said there were
more bits on the disk to read.  If oversampling was what was meant (the
salesman had never heard this term when I mentioned it) then the claim
that this player used more samples than any other is wrong -- the Philips
players use twice again as many.
						B.J.

That's about what I thought Philips was doing. As far as I know,
Yamaha does about the same thing but at half the sampling frequency...
in any case, the disk itself contains no more than 44.1 k samples
per second.

It is very interesting to follow the discussion on the advantages and 
disadvantages of the various methods (at least in the HiFi press over here).
There are people who claim they can hear the difference between an
oversampling player an a normal one, others say they cannot hear any 
difference.

At least you can see the difference if you look at the square wave
reproduction. Philips and other oversampling players have fully symmetric
'ringing' of the square wave, other are asymmetric.

Of the three major HiFi magazines here, one is convinced that the
difference is audible and even very significant, the other two say there
is no audible difference.

I tend to believe there is no audible difference, especially after reading
a test of 40 CD players in the previously mentioned magazine. They had
a subjective test on the sound quality too, praising all oversampling players
as far better than the others. Of the Akai player they said it had a typical
non-oversampling sound; apparently not being aware that Akai uses oversampling
too!!!!! (Of the four different player manufactured by Kyocera two have
oversampling, Akai and Rotel, the two others, MicroSeiki and Kyocera don't).

More serious reviewers see only one possible advantage with the oversampling
method: better long-time stability because of the digital filtering and
simple analog filter vs. a complicated analog filter.

I called Yamaha (they have an 800 number: 800/854-3264).

The player does do 2x oversampling.  They convert 44.2k 16 bit samples
to 88.4k 16 bit samples.  Philips converts them to 176.8k 14 bit samples.
Does anyone know which method is "better"?  Can "better" be defined in
this case?

Since this player reads a track ahead to do full ECC checking, it should
be able to play disks with more damage than other players.   Does anybody
know of any other player which does full ECC correction?

The person on the phone did not know anything about plans for double
sided disks, did not know that their player was "designed to play double
sided disks", and did not know where the salesman got that information
from.

Because of the oversampling and full ECC, there is a good chance I will
buy this player.
						B.J.

By the way, if you are interested, I have a couple of test of the
Yamaha player (I don't know if you believe in tests?!?!?). I can MAIL
you a summary.

There seem to be lots of curios players around. The early Akai players
had the 'standard' steep filters even though it has oversampling. The later
ones seem to have more reasonable filters. This is just to remind that
oversampling as such doesn't mean you have less phase problems.

You should also beware of the early Hitachi (DA-1000) players which had
bad D/A converters (recognisable on the BLUE power switch, as opposed
to the newer ones with a grey switch). The new ones have a Burr-Brown
D/A-converter.

By the way, I have a Toshiba player and am very satisfied (XR-Z70).
It costs about $500 here, has remote control, fast access times, phone jack,
adjustable output, and is reasonably small.

I have some technical papers on how Philips generates their samples.
They take L/R samples at 44 khz, giving 88 khz 8-bit samples.
They use a coding method that they call "EFM" ("Eight-to-fourteen
modulation"), which is a trick for tracking accuracy. Will send
a limited number of copies to interested persons.

There are usually several photodiodes in the laser reader assembly.  They
are for tracking.  In the case of the Philips, there are four.
One "reflected" data stream is split up by small prisms into
multiple streams.  Then, sums and differences are done to see
if the reader is too far "left" or "right", and real time correction
is done.  Ditto for focus on the pits -- the objective lens is
mounted on a little-bitty "voice coil", and can move up and down.
It's position is also adjusted in real time.  And, of course, the
photodiodes give you the output signal, too.  But more diodes
don't give you more anything in terms of signal.

we've done a fair amount of work here on CD's, as have the Tapes
Engineering folk in SHR.  We have some drives taken apart, etc.

don feinberg

I was just in Japan, and I found the Yamaha CD-X1 selling for
only $310.  The Yen is kind of weak just now (226 = $1).
Tweeter wants $595 for it here.

I don't know how the warranty works if you buy one in Japan.

CD disks in the stores were around $15.  But phonograph albums
are close to the same price.  Very expensive in Japan.

I reciently bought the Hatichi DA-3500 after looking at several other
units on the market. I am quite pleased with the unit. I talked with the
salesman at Tweeter in framingham about the difference between the Hitachi
and the Yamaha. After calling the Home office he told me that the Hitachi
and the Yamaha were basically the same units but that Yamaha produced the
semiconductors. 

As a Component Eng I have often run across Hitachi components but I have
never seen a Yamaha part. Does anyone know if this is true.

I was also told that Denon makes the unit but it is sold under the Hitachi
label due to some sort of Import problem.  Can anyone shed some light on 
this.

Paul

Any recommendations where to buy a CD2 in the Eastern MA/Southern NH area
and what the going price seems to be?  Also are these available via
mail order and if so for how much and where?

INANNA::FORTMILLER

    
    how about the cd-28, is that a good unit?