Conference cookie::notes$archive:cd_v1

Thanks for the review. I've been thinking of getting one of these to do
double-duty as a portable *and* a car CD player (of course, I figure I'd need
to get a amp for the car -- or will the D-5 drive car speakers by itself?).
I'd like to have a CD player for the car, mostly, but I've been (understanda-
bly) nervous about having it ripped off. Getting one that I can port in and
out seems like the ideal solution.

--- jerry

	Thanks again for the review. Is the Ac adapter and the
	external battery case included in the price??

	>>>Peter

I believe the battery case is; without one of those two you can't use the
thing because it is WAY too small to hold 6 'C' cells.

By the way, could someone out there explain why 'double sampling' can improve
the high end of CD sound?  I didn't think you could get any more information
off of the disc than what was put ON it, at 44.1 kHz!  Doesn't 'double sampling'
simply read each word off the disc twice?  I don't understand what that could
gain you.

Confused,

John Francini

"Double sampling", or in the more general case oversampling, has been discussed
earlier in this file and also in AUDIO.NOT, but here is a summary.

Oversampling has nothing to do with getting the information off the disk or
error correction or recovery, it is part of the processing of the sound.

The 44.2K signal rate can uniquely specify all of the audio frequency
components up to 22.1K.  Any higher frequencies on the input have no
effect on the bits and an infinite number of output waveforms can be
generated by the D/A conversion and be consistent with the input.
The high frequency sound is probably inaudible, but there could be
a noticeable beat effect, it could be irritating to animals or people
with good hearing, and could damage audio equipment if it has a high
level.  All of this "noise" should be filtered out.

If the filtering is done by analog techniques then the slope of the filter
will either take away from the high end of the "good" signal, or leave
part of the "bad" signals.

Oversampling is running the D/A converter at a higher rate by inserting
zero samples into the input data stream.  If you double the number of
signals then you have a data stream which specifies all signals up to
44.2K, with all of the signals in the range of 22.1K to 44.2K being zero.
Now it is much easier to use an analog filter to get rid of the remaining
high frequency noise without touching the signal.  Philips uses quad
oversampling -- that give you a clean signal up to 88.4K.
						B.J.

I believe that the battery case is optional as well as the headphones.  The
machine comes with a ac adapter and a patch cord to convert from sub-mini
jacks to phono plugs.  This way they can sell it ready to use with a home
system and also have portablity as an option.

Of course, the real reason is because an ac adapter and patch cord is cheaper
than battery case and headphones.

I believe that the Battery cale

                             ^---case costs $50 and the headphones go for
about $80.00.

I also read (in a consumer electronics trade rag) that the CD industry was
worried about the Diskman hurting the rest of the industry by forcing a
rapid price erosion.  At a price of $300 list, consumers will expect similar
prices across the board.  The article further said especially since the unit
will most likely be discounted, possible down to the $200 level.  No time 
frames were given.

Jim Pappas

There is a write up of the D-5 in the December '84 Stereo Review which lists
the following prices..

	Player				D-5		$299.95
	Carrying case / Battery pack	EBP-9LC		  49.95
	Car battery cord		DCC-120		  36.50
	Sony headphones ("recommended") MDR-M77		  85.00

		..Paul.

(re: .4 "yet another explanation of double/quad/kilo sampling"

Do the multiple sampling schemes actually use a ZERO value between
actual values or some interpolated value?  It seems to me that using
zero would inject harmonics of the basic sampling rate.

__Rich

	re.-1
		I thougt that you used a value between the current
		sample and the next coming one (for double sampling).


	>>>Peter

Re: .7,.8

I am not sure what values are put in to the data stream.  I have been told
many times that the values are zeros and a couple of times that they are
interpolations so I have started repeating the "zeros" version.

If anyone has the definitive story, or feels that they can work out the
math themselves to see what values work, please let me know.

						B.J.

RE: .7, .8 and .9

Somewhere recently (probably on the USENET net.audio distribution) I skimmed
a lengthy and scholarly description of the Philips quad-sampling technique.
According to this report, zeroes are indeed inserted in the binary data
stream.  The report went on to describe what happens next, but I didn't
really understand the math or electronic theory involved.  I do recall that
the end result was better fidelity out of the analog filters.

Ward

Some more info on the Sony D-5:

In case there are still any questions, the D-5 comes only with the AC adaptor
and a patch cord from the stereo mini-jack to male RCA connectors. It does
not come with headphones or the battery case. I tried mine with the pair of
headphones that came with my Walkman and found that the dynamic range and
high end frequency response of the CD made these type of headphones sound
pretty lousy. They are really made for radio and tape. I found using a really
good pair of standard size headphones sound really great. The headphone amp
in the D-5 does a really good job driving such headphones. It won't melt
your brain but it is plenty loud. Also, my headphones are inefficient so
more effiecient phones would be even better. The prices mentioned earlier
for the headphones, battery case and car adaptor to me seem to be quite
outrageous. I would buy non-Sony stuff myself. You can get a pretty fair
pair of headphones for $80. Also, you can get a 9V car adaptor at Radio
Shack for I believe about $10. It has plenty of power to drive the D-5.
I don't know what the Sony battery case is like, but you could make one
from parts at adio Shack for about $6.

What I would really like to report in this note is that I now have wired
up my D-5 in the car. I have never heard anything like it in a car before.
Its like having a whole new stereo. I have a nice sound system in my car
as is, subwoofers, front/rear speakers, separate power amps on all, and
a crossover/equalizer (all Proton stuff). Speakers are all ADS. I built
a small switch box to select the input to the system between my radio/tape
dec and the D-5. It also has a requlator in it to provide the 9volts for
the D-5. I have had the D-5 setup in my car for about 2 weeks now and
have not used the radio/tape dec since. The sound is really amazing.
I
can make the whole car shake wit
 the bass, and the high end isIincredibly smooth and clean. I can place the D-5 in the glove compartment
or on the front seat. In either place I have never had it skip once. In
fact it has performed flawlessly since I got it. I never leave it in the
car when I am not in it, put its so small and easy to connect/disconnect
that this is not problem. Takes all of 2 or 3 seconds. One time I did
leave the unit in the car over a really cold night to see how it would
perform being freezing cold. I just popped in a disc and away it went.
No problems at all. I think the D-5 is truly an amazing piece of hardware.
I am currently driving the input too my system from the headphone output
of the D-5 since this gives me control over the volume. However, I plan
on putting a volume control in the same box with the regulator/switch
as soon as I receive it (I had to order it) and driving it with the
line outputs of the D-5. However, I don't really think it will make
any difference in the sound.

In sound comparisons with my Yamaha CD-2 I have found that the high end
of the Yamaha is cleaner than that of the D-5. I am assuming this is
due to the oversampling and digital filtering of the CD-2 which the D-5
does not have. But in the Car, it is truly awsum! Given you would pay
at least $300 for a car radio/tape deck the D-5 is not a bad deal at
all. If your into CD's in the home, if you even try them in the car
you'll never listen to your radio or tape again. If anyone has any
futher question, let me know.

Re oversampling...

What oversampling does is effectively create a data stream with twice
(or 4 times) as many data points as the origional data stream.  The higher
speed data points emitted by the sample rate converter are a weighted
average of the origional samples padded with zeroes; effectively its
a way of producing intermediate samples that are interpolated properly
between the actual data.  This weighted average is called a finite
impulse response digital filter and can be designed to have very accurate
linear phase; if you see a reveiw of a CD player that has a very symmetrical
looking impulse response it means that its filter has liner phase.

The advantage of doing digital filtering is that now the analog filter's
stopband is not a mere 1kHz away, its over an octave away and only mild
low-tolerance filtering is needed.  And the digital filter will not drift.
Also, its actually harder to phase equalize a sharp cutoff analog filter than
to do the filtering itself!  Hence the players I've read reveiws of without
oversampling have not had linear phase response.

I've read a few papers on the Phillips version and think they did their
design work very well.

In terms of filtration there's no sonic advantage of using oversampling over
direct analog filtering per se, though it may be more economical to build
an oversampled unit since any trimming of the lowpass filters is effectively
eliminated.  The linear phase aspect is sonically unimportant (as long
as both channels are precisely matched in their phase) but nice to have.

One other possible advantage of oversampling is if the D-A were to drift
causing excess quantization noise (above the least significant bit) the
oversampling converter may tend to reduce the audibility of the noise.
Again, that may be unimportant in practice since there are design techniques
for making D-A converters inherently linear down at the LSB level and
CD manufacturers use these techniques.

What all this means:  oversampling is another way of doing it but doesn't
make any practical difference.

- Jim

If I recall from the writeup in Stereo Review, the D5 has about a 2 dB
rolloff of the highs at 20kHz. I suspect that would be more likely to
explain the difference you hear, since it's a lot more blatant than 
single vs oversampling, unless the rolloff was only a characteristic of
the unit Stereo Review received.

Re: .13

The rollof is a result of the lack of oversampling, I got a good
refresher of why from the Nak OMS-7/OMS-5 brochure that I re-typed.
If you don't oversample, you have to sharply attenuate the frequencies
after 20 kHz, unfortunately, analog filters are not perfect, and some
of the attenuation occurs below 20 kHz.

					    -Dave

I doubt how many people can hear -2dB at 20 kHz (assuming the rollof starts at,
say, 17 kHz)????

I personally do not hear much above 17 kHz, and even at 17 kHz it has to be
pretty loud. I would say that most of the signals (if any) above 15 kHz on
most CDs (or any other media) will be below theshold anyway.

Well, that applies to most humans anyway; don't know whether the rest of
the participants to this notesfile are cats and dogs...

	I just bought my wife a D5 for Hanukkah. It is an AMAZING little
box. They are hard to come by here in the Chicago area, and no one is
discounting the $299 list price. My wife has easilly matched that total
in buying disks to feed it in the past week! One of the disks we already
had as one of those "audiophile" recordings, from a digital master. I
tossed the disk on my turntable (~$600), and the CD in the D5, and
played the two in synch. Switching back and forth between them is quite
enlightening! The best way I have found to describe it is to be outside
a concert hall, while someone opens and closes the door! Its kind of
like the movie "The Wizard of OZ", when it switches from B/W to color. 

	I too plan on rolling my own portable power pack. Since the unit
seems to run on 7.2v (6 Ni-Cads) to 9.0v (6 Alkaline), I plan on using a
couple power packs for my electric R/C cars. I have a 7.2v 1.5 AH pack
made up of 6 nicad cells, and an 8V 2.5 AH pack made of 4 2v "D" gell
cells. I just plan to make a connector for the back of the D5 that
matches the R/C car, and I should be set to go. My understanding is that
the unit will only run for about 2 hrs on the Nicad pack, and if you go
alkline, that it will run for 4 hrs. Given a $1.00 each price for C
cells, thats $6 a fill! You can easilly spend more in throw away
batteries, than you do on the player or the disk! I'll stick to the
rechargables! 

	Now for the problem. We have occasionally noted a faint high
pitched tone coming thru both channels of the unit. It is intermittent,
usually hapenning at the beginning of a disk, but sometimes starting in
the middle of a playing disk. If you skip around the disk, and back to
where you started it usually goes away, but will most likely come back
if you "spin down" and restart the unit. It has hapened intermittently
on several different disks, and both with headphones, and when playing
thru the stereo system. Once it shows up, it is usually repeatable for
some time, and then goes away. My rock-lined ears estimate that the tone
is around 10KHz. It is much fainter than peak music volumes, but can be
heard quite clearly in soft passages and between tracks. 

	Has anyone got any ideas on what the problem could be? Has
anyone seen\nees\heard this on a CD player before? Are all of my disks
bad? Could this be an environment problem? I really hate to send it
back, as we live the unit so much, and I can see it taking 1-2 months to
get it back, only to have the problem reoccur because they could not
make it fail. By the way, I called SONYs toll free number and asked the 
same questions. They would only refer me to my local authorized SONY 
shop, or suggest that I take it back where I bought it (which I will, 
unless someone has a better idea on what the problem is). I also asked 
about the service manual, and was told that it was available for $5.00. 
I expected them to want a lot more for it. (they only wanted $15 for the 
manual for my SL2700 BetaHiFi!) I shal likely pick up both of them.

I have a D-5 and have been using it quite a bit for the month or so I have
had it. I have not noticed anything like youd described. If you have not
had it long, I would just take it back to where you got it and exchange
it for a new one. I sure your discs are fine since you said you can repeat
parts that made the noise and it will not be there. I have around 50 discs
now and have not had a bad one yet. Unless you include the one that had
a defective case (the top came off when you opened it up).

	I did just that. I brought it back to the store in the Chicago
area where I bought it, expecting that they would either take it into
their own shop, or send it to SONY for repair. They happened to have one
in stock, and traded mine for the new one. The new one has been fine so
far, and they sent the old one back to SONY for replacement. I certainly
didn't expect this kind of service, when they can sell these little guys
as fast as they come in. I guess thats why I have been dealing with that
store for 8 years. For anyone in the Chicago area looking for a
reputable dealer, they are United Audio, with 5 stores in the area. They
aren't the high pressure places, and their prices are VERY competitive
(they say that they will match any one elses advertized prices). The
only thing I went elsewhere for was my SL2700, as mail orded saved me
about $350, including the sales tax. 

I received my February Audio magazine recently and in the back, there is
a store in Braintree MA. selling the D5 for 218.00.  This is of course
plus shipping and sales tax to Mass customers.  I do not know if they are
selling other than mail order but there is the potential to save a little
there.

A while back I reported that the industry was a little upset with Sony when
they introduced the D5 with a suggested retail of 299.00.  They were upset
that it might cause too rapid a price erosion in the home player units as 
well.  The article stated "especially since it is likely to be discounted
to the 200.00 range.

Well,  Let the price wars begin!!

Jim Pappas