Conference cookie::notes$archive:cd_v1

I seriously doubt you'll hear the difference between the dbx and *any*
other cd player. Any response aberations are generally of the magni-
tude of 10ths of a db at any given frequency, or .00X% THD, or some
other absurdly small figure. I don't believe you'd hear the difference
between players except in an A/B test and then couldn't say which one
was "better". For all practical purposes, repeat, practical purposes
you can consider a properly designed and functioning cd player absolutely
flat and distortionless. You're more likely to get a different sound by
moving the couch in your listening room than you would between units.

Check out the features you want, programming, suseptability to jarring,
warrantee, serviceability, ect and trade these off against $$$. Don't
hang yourself with the specs.

Does anyone consider my first sentence an invitation to "war"?

edd

Yes.

The key to your statement is: "properly designed."  Only the MOST expensive
Sony and the Phillips based players fit that bill - the rest use only one
DAC and suffer from interchanel phase shift, or don't use digital and analog
filtering with oversampling.

Now comes a neglected factor: a properly executed design.  This criteria
dismisses most of the Phillips, and the rest of the improperly designed units.
Note:  How can you expect a manufacturer to use premium parts in a player
with an inferior design - why would they bother?

I've done much research and many mods on my Magnavox (Phillips) player (and
I'm not done yet).  They cut corners.  It doesn't really matter in the
servo system as long as it works, but it counts alot in the simple audio
section.  They should have provided two well regulated, higher voltage supplies
for the audio section instead of sharing the servo section's.  They should
have used larger, better electrolytics, more film caps, some micas instead
of ceramics, and more and higher tolarance metal film resistors.

The Magnavox isn't a bad player at all, this is typical.  All the compromises
are done for cost - and don't assume that by paying an extra $40 it would
have taken to do it right, you will get a better player from someone else.

I sure hope the A to D circuitry at recording studios is more "perfect" than
what is commonly found in players.  I can't picture them using teflon caps
in the sample and hold circuits, and certainly not in the new DAT recorders.

So, there.  Happy?

Mark

re: -1  And now for "Shades of Julian Hersch...". Points taken. Granted
there are theoretically better designs and better implementations, but
are they of an audible magnitude? Certainly a single D/A must cause some
degree of phase shift by nature, leading to a lack of imageing. But, and
(scratch "and") are you going to hear it? During an A/B test, probably.
My Sound Concepts IR2100 allows me to vary phase relationships between
channells and I can hear the difference "as it occurs", but small changes
become "lost" seconds later without a reference point to compare them
against.

I believe the argument for or against replacing caps, filters, ect. be-
comes moot when taken in reference to the original note. Engineers may
know there is a "better" sound being produced (and hence will "hear" it)
but should Joe Consumer spend his hard earned bucks chasing the Grail?

I don't think the difference between players of one design vs. another
is great enough to get hung up on. For the most part, they all sound
great. Sit back and enjoy the music. 

Edd

On the phase-shift note.  I have asked this question before, but without
a satisfactory answer:  Are not the bits laid down sequentially on the disk?
I always thought that the bits for one channel came before the bits for the
other channel on the disk.  What good would 2 DACs do you in that case?
Does the DAC which picks up the first sample hold it until the 2nd DAC gets
its sample?  That would be the only way to eliminate phase shift.

Tin Ear advocate of the Consumer Reports school of audio equipment review.

Sorry Eirikur, this is really digresing.

First, let me admit a mistake.  ADS now has a 2x oversampling player with
dual 16 bit D/A converters.  Second, I'll admit that there are two
compromising aspects to Phillips players.  First, they don't quite have
16 bit resolution.  They use some fancy interpolation and filtering but
then use only a 14 bit D/A converter.  Second, there are some spurae
at 1/8, 1/4, and 1/2 of the sampling frequency, yielding -50db peaks
at 22, 44, and 88khz.  Mission uses additional filtering to attenuate
the 44 and 88khz peaks.

Here is a block diagram for the Phillips player:

Reader -> Demodulator -> ERCO -> CIM -> FIL+-> DAC -> Active filter -> out
					   +-> DAC -> Active filter -> out

OK, those are the major components for the decoding section.
Reader gets an analogue signal reading data off the disk.
Demodulator decodes the data into binary data and provides sync.
ERCO detects and corrects soft errors.  It flags the CIM for hard errors.
CIM (Concealment by Interpolation and Muting) if flagged, will interpolate
between data points to correct a single error, and mutes for multible
errors.  It stores data points in a fifo buffer which also is used to
control the speed of the disk so that the buffer is optimally half full.
Left and right chanels are clocked out of the CIM into the digital filter.
The digital filter does the interpolation and filtering and SIMULTANEOUSLY
clocks out the data to the two DACs.
The DAC output goes into a 3-pole analog filter and buffer and out to world.

So, to answer your question more fully, there are buffers and latches
between the laser and the DACs.

   Mark - 
   	which Magnavox did you get, Treismans in NH has the 2040 for $167.
   and I think I now have no more excuses.  
   
   How did you like the unit before the mods?  I don't think I'll get around
   to doing anything for a while, the budget is still very tight.  also,
   how much did the mods cost so far?  If you rather, send me mail?
   
   				dave
   				----

Treismans?   What sort of place is this and where are they located?

	Eirikur H.

I got the 1040.  Which is basicly the same player except it is a few inches
narrower and lacks a headphone jack.  I paid $167.95 at Service Marchandise.
$167 is an excellent price!  New Sonys are supposed to be good but cost more.

The 2040 is a BETTER unit!  Because it is on a wider chasis, they did not
have to use surface mount components, and there is more room for modifications.
The 1040 requires soldering components on to the foils.
                           
The unit before the mods was OK, and probably better than anything near its
price.  I have probably spent $40 on modifications:  $12.50 for a service
manual, $20 for a pair of premium DC blocking capacitors, and a few dollars
for metal film resistors and odd capacitors from my junk box.  I think that
the most important change is replacing the electrolytic DC blocking caps
with films and can be done for less than the $20 that I spent.  I have on
order a 2041 manual that will show the layout on that unit.

The 1040, 2040, and 3040 are being replaced by the 1041, 2041, and 3041.
The new models offer audible search, and (I'm not sure) faster access times.
They arn't in stores yet and are otherwise identical to the current line.
Sometime next spring, Phillips will bring out their 2nd generation machines
that will use dual 16-bit DACs instead of the current 14 DACs, so you may
want to wait.

READ NOTE 179.

Since that time, Musical Concepts has announced a Super Mod.  The regular
mod costs $180 installed and the supermod costs $280 installed.  The supermod
improves the power and passive components around the DACs, and replaces
the cheap polysteyrene anti-aliasing capacitors with premium polypropelenes.

Also since I wrote note 179, Musical Concepts sent me their opamps, and
they do indeed sound better than the original.  Imaging is better, and
on recordings like Flimm and the BB's Trycicle, what sounds like clipping
is gone too.

The current sound of my 1040 is fairly close to my Linn (also new since note
179)  so I have been content with it and havn't continued modifications.
I've lost some interest in it because LPs cost half as much and thanks
to CDs, when Strawberries dumped their stock of direct-to-disk records,
I bought 21 of them for $1.99 and 2.99 each.  Want to buy my player?

   woa 1.99 and 2.99 each - good catch to say the least...
   
   re .-2
   
   Treismans is like Service Merchandise.  there's one in simenaeu (sp) plaza
   in Nashua NH.  (in the center of town)
   
   on the dual 16 bit DACs - I'd love to wait, and might, but it will be
   a long time till they are discounted like the 2040 is now...
   
   2040 lists at 299 is 199 typically at treismans and is on sale for 167.
   
   so we'll see. If my wife gets a job this month we will get the 2040.
   if not i'll be waiting for the 2041 or 3041..
   
   				dave     
   				----

RE: .4  (sequential layout and phase shift with 2 DACs)

Yes, the bits are laid down sequentially on the disk, one channel before
the other.  This doesn't cause a phase shift with two DACs because there
is a parallel buffer for each DAC that holds the samples until samples for
both channels have been read.  Then, the samples are clocked into the DACs
simultaneously.  Hence, no phase shift.

--PSW

Back to the original topic of this note; the dbx DX3. Here is the brief comment
that the $ensible Sound had to say about it (Summer 1985 #26): 

"HIGH END AUDIO FOR THE HEARING IMPARED? The dbx DX3 CD player is being marketed
as a CD player with sonic differences - benefits anyone can hear. The machine
features a compression circuit (to squash those nasty CD dynamics that you are
paying extra for), an impact recovery unit (if you can't make up your mind) and
an ambience circuit. An unheralded feature that we noticed was that all of these
neat circuits can be switched off."

Nuff said? Before you go off the deep end, this magazine DOES favorably review
CD players and CD recordings. They very much liked the Meridian MCD ($650), more
so than the more expensive Nakamichi OMS-5 ($995) and Revox B225 ($1150). They
also liked the Sony CDP-302 ($550 list but typically $400-450).