Sony CDP307ESD Test Results via Audio Precision System One
Andrew Karanicolas
present on a CD player. I want to solicit information on
several questions that arise. The CD player is a Sony CDP307ESD
purchased in 1988 from (guess who!) Q-Audio.
I purchased a test CD from MCM Electronics (80-505) for $7.95. It
has sine waves recorded on different tracks from 20Hz to 20kHz, -10dB.
It has one track at 1kHz, 0dB. It has other stuff too.
We have an Audio Precision System One in lab so I decided that I was
going to measure my CD player out of curiousity. For those of you
familiar with AP, I ran the CD-THD program on the AP for this test. I
selected a 22kHz bandwidth for the THD+N evaluation. I start up the
CD player with the test CD and evaluate the THD+N. The results of the
THD+N are less than spectacular:
20Hz 0.70%
100Hz 0.15%
200Hz 0.09%
1kHz 0.06%
2KHz 0.06%
10kHz 0.04%
20kHz 0.05%
In fact, these results are *drastically* worse than the published
specification of 0.005% from 20Hz to 20kHz in the owner's manual. The
THD+N at 1kHz, 0dB only improved marginally, less than 1dB, over 1kHz, -10dB.
I do not believe that the THD+N is dominated by noise. The THD+N only
improves by 2dB with the 1kHz, 0dB case vs. the 1kHz, -10dB case.
I do not expect the THD+N to improve significantly at other
frequencies. In fact, the THD+N may degrade at 0dB compared to -10dB
for different frequencies. I thought the RCA cables might come into
the picture. They make no difference. I connected the CD player
directly to the AP, without any cables, no difference at any
frequency. So, the cables are not involved.
I measured the magnitude and phase response as well. The magnitude
response deviation was within +/- 0.5dB over audio bandwidth and the phase
response was within +/- 2.5 deg. The actual responses were usually
well within these limits.
I am thinking of getting the Denon or JAS test CDs. At least one of
those have a sine wave set at 0dB. I am curious to see how the THD+N
measures for this case. The test CD I am using might itself have
harmonic distortion in the recording, but I doubt it. I cannot
believe it would have tenths of percent of THD+N.
I next tried some other CD players, 2 portables. One of them was a
recent Sony with a 1-bit delta-sigma oversampled DAC. It showed a
THD+N of about 0.30% over most of the audio range, and was about 0.8%
at frequency extremes. The magnitude and phase response looked pretty
good, not much worse than my own CD player. I tested another Sony
portable. This unit is from the 1987-88 era, similar to my own
player. It measured out pretty bad, about 2.5% THD+N over entire
audio bandwidth.
The really nice feature of the AP is the real time THD+N measurement.
I noticed that drumming my fingers on the CD player would cause the THD+N to
increase slightly at mid-band frequencies before the track skipped.
I could tell when the track skipped because the CD player counter
jumps at this point. At 20kHz, the THD+N would degrade 6dB to 8dB!
At 20Hz, it hardly changed at all. I'm wondering if cleaning laser
will improve matters. I don't feel like sinking $25 into some laser
cleaner unless this is really contributing significant THD+N in some
way.
I'm wondering if there is some strange weighting curve for the THD+N
for CD players. As far as I know, there is no FTC regulation on the
THD+N of CD players, unlike amplifiers. One regularly sees the THD+N
published for amplifiers even in newspaper advertisements. When was
the last time you saw the THD+N for a CD player in print? Also, I'm
beginning to wonder if Stereo Review and Audio Magazine get tweaked up
CD players. They also use the AP for their measurements. Most
people, however, do not have a $15,000 AP to test the CD player they
bought from Circuit City or wherever.
So, who here has information that can shed light? If you have
experience in evaluating CD players, I am keen to hear how you do it
and with what instrumentation.
If you want to see the data I got, I can e-mail you the data. The
THD+N, MAG and PHASE plots are each only 5kbytes of postscript data.
*******************************************************************************
NOTICE:
I am fully aware that measurements have limited utility in the context
of the listenining experience. However, measurements are extremely
important for actually figuring out what the hell is going on, among
other uses. So, if you have some opinion about measurement vs.
listening, I don't care.
*******************************************************************************
If you have information that is useful, please send e-mail. I don't
regularly read wreckage.audio. I also posted on rec.audio.high-end.
OK, thanks.
--
Andrew Karanicolas
MIT Microsystems Technology Laboratory
ankl...@mtl.mit.edu
Keen Giap Law exp 1/94~
Keen Giap Law Intel, Corp.
Mailstop : CH11-91 5000 W. Chandler Blvd.
Tel : (602)554-4299 Chandler, AZ 85226
Fax : (602)554-7388
Chao I Wu
Hmmm...There seems to be a long thread on this one...here's my 2 cents.
It seems to me that the problem with quality in Sony isn't that its just
plain old bad, but it is the quality CONTROL. Too many faulty machines
enter the market. In my experience with Sony, if you can make it through
the first 6 months without a problem, it usually will last. But this is
only MY experience, and I am sure people have dozens of complaints that
don't fit this pattern.
If you take a serious look at Sony repairs, you will find that most of them
occur within the first couple of months. Too bad Sony hasn't learned that
quanitity isn't quality. They are quick to put out products, but they just
plain old break.
,,,
(o o)
+----------------------oOOo-(_)-oOOo------------------------------------------+
| Chao I (Bobby) Wu | Computer Services Division (CSD) |
| INet Address: ch...@csd4.csd.uwm.edu | University of Wisconsin - Milwaukee|
+-----------------------------------------------------------------------------+
j...@cmkrnl.com
>
> Hmmm...There seems to be a long thread on this one...here's my 2 cents.
>
> It seems to me that the problem with quality in Sony isn't that its just
> plain old bad, but it is the quality CONTROL. Too many faulty machines
> enter the market. In my experience with Sony, if you can make it through
> the first 6 months without a problem, it usually will last. But this is
> only MY experience, and I am sure people have dozens of complaints that
> don't fit this pattern.
>
> If you take a serious look at Sony repairs, you will find that most of them
> occur within the first couple of months. Too bad Sony hasn't learned that
> quanitity isn't quality. They are quick to put out products, but they just
> plain old break.
I'm not certain that it's just QC. Sony receivers, for instance, have a rep
for running a lot warmer than other units in their power range. This may
indicate a tendency to design other than conservatively. If they cut corners
on air flow, what else have they cut corners on? And of course all that heat
doesn't do the parts any good...
--- Jamie Hanrahan, Kernel Mode Systems, San Diego CA
Internet: j...@cmkrnl.com (JH645) Uucp: uunet!cmkrnl!jeh CIS: 74140,2055