Hi all,
This is a repost of a USENET post I just submitted to alt.audio.minidisc. I
hope this doesn't break the rules of the FAQ by sounding "argumentative",
but this is my attempt to be as objective as possible describing my personal
experiences with both the Sharp and Sony MD recorders, and specifically
their different implementations of ATRAC compression. Comments, criticisms,
flames (ok, not flames!) welcome..
As a new MD owner, I searched all over the Web and USENET for any
information on differences in quality between the different ATRAC versions.
Sure, on minidisc.org, you can read about the improvements in Sony's ATRAC
implementation from version to version, and how well Sony MD compares to
128kbps Xing-encoded MP3 (please!), but I couldn't find *any* mention of how
well Sharp's encoder stacks up, other than vague comparisons like this one
(from http://members.tripod.com/~Civil_Disobedient/r50vs701.html)
"The result is that the Sharp and Sony units make recordings with slightly
different sound characteristics, but which one people prefer is more a
matter of personal taste than one necessarily being "better". I'm
unfortunately not in any way equipped to test the differences, and until
someone does, it's fairly safe to judge them the same and move on to
something that actually makes an audible difference."
So, assuming that one ATRAC's as good as another, I ordered the MP3 combo
kit, including the Sharp MD-MS722, from minidisco.com. The new sound card
with optical out worked as advertised, but I quickly discovered that my Aiwa
bookshelf stereo's digital out worked even better (no manually splitting up
tracks, and since I own the original CD's to the music I want to listen to,
I wouldn't think about copying from an MP3 unless I had to).
Everything seemed okay at first. The Sharp had a nasty tendency to cut off
the first second of each track when I used the Aiwa's program play feature,
so that sucked, but manually recording a track at a time (or the entire CD)
worked around that bug. But after a few days of listening to my tunes, I
came to the unhappy conclusion that something wasn't quite right. Some of
my music started giving me the impression of listening to a low-bitrate MP3:
some people may be happy with 128kbps MP3's, but believe me they're not CD
quality!
After conducting some painstaking listening tests a few months ago, I
discovered that 192kbps, or even 256kbps is necessary to make a "CD quality
MP3", or else one starts to hear a certain "fuzziness" or "shrouding" about
the music.. Cymbals seem sizzly, and the natural ambience of the room
disappears. Maybe this is what audiophiles mean when they talk about
digital lacking "warmth", but what we're really talking about are
compression artifacts! When you're throwing away 80% (or in the case of
128kbps, 90%) of the music information, you better throw away the *right*
80%, and that's why making a good encoder is so very difficult (by
comparison, the decoder's job is much easier, as it just has to play back
the bits that are left). This is especially apparent with acoustic
instruments like guitar and cymbals (which can have very complicated upper
harmonics that the brain has an entire listening history to compare to), and
with techno music, which often has very pronounced percussion and strange
electronic effects which can be very difficult to encode (try encoding the
first 30 seconds of "Kalifornia" by Fat Boy Slim with *any* MP3 or MD
encoder, listen to the result with some good headphones, and you'll hear
what I mean!).
Since I knew that MD was using 292kbps, I assumed it wouldn't have any
problems with making me happy.. it certainly shouldn't sound as bad as
128kbps Xing-encoded MP3's, but here it was. I ended up pressing a CD-R of
the original 48 test tracks I'd used to compare MP3 encoders (all culled
from the first 30-60 seconds of various songs I suspected would be difficult
to encode), and copying them to MD using the Sharp. Sure enough, quick A-B
testing confirmed some serious quality problems! Acoustic guitars sounded
slightly muffled and synthesized.. Pan flutes, pianos, and drums suffered a
similar fate. There was a most noticeable loss of higher harmonics, leading
a fuzziness to everything, most unpleasant. Crisp attacks, such as cymbal
crashes were rendered hollow, to the point of sounding more like a bad
Commodore 64 white noise synthesized "drum" than the real thing. It sounded
so bad I was beginning to regret my entire MD investment (and after I bought
the Sharp 722, I had splurged on an entire 40-pack of blank MD's, AND the
JVC MD/CD car unit.. yep I went a little crazy!).
My only hope was to put faith in Sony, hoping that their ATRAC
implementation wouldn't let me down (after all, all of the head-to-head DAT
vs. MD tests by musicians that I'd read online had used Sony, not Sharp). I
ordered the MXD-D3: 4x copying from CD to MD is a *sweet* feature, but only
if the quality is decent. However, the MXD-D3 came with Sony's
next-to-latest ATRAC (4.5).. I assumed the differences between that and the
high-end "R-type" on their ES models would not be a big deal.. besides,
there's no way I was going to fill up my 40 new MD's at 1x speed!
Long story short: the Sony sounds fine. I'm perfectly happy with my quick
listening test of the 48 sample tracks, encoded with the Sony and added to
the same disc I'd previously recorded with the Sharp. Doing a shuffle-play
with my eyes closed, I can't always tell which track was encoded with the
Sharp and which with the Sony, but I'll try to get my roommate to help me do
a proper A-B comparison, for scientific accuracy. But my ears have already
convinced me: Sony ATRAC is acceptable.. definitely comparable to a high
bitrate MP3.. Sharp ATRAC loses worse than Xing!! (and that's saying
something!)
Don't get me wrong: the 722's a great MD *player*.. it seems to have a fine
DAC, and a great set of features. Add a stereo mike and it makes an
excellent portable recording unit for concerts or lectures, too (where the
sound is going to suck, no matter what!). But the encoder is *clearly*
substandard, and as far as I can tell, I'm the first person on the Internet
to point this out, or actually take the time to listen and seriously compare
the Sharp and the Sony ATRAC implementations.
In summary, don't be fooled into believing that all ATRAC's are the same.
Don't be fooled into thinking Sharp is better because it's "version 6" and
Sony is "only" at 4.5. If I had only been able to hear the Sharp ATRAC, I
would have probably written off MD completely as a poor competitor to
128kbps MP3 (despite the great attractiveness of the MD recording format).
Thankfully, I had faith that Sony's encoder wouldn't suck, and it didn't.
Don't get burned! Use a real ATRAC when you make your mixes!
-Jake Hamby
-----------------------------------------------------------------
To stop getting this list send a message containing just the word
"unsubscribe" to [EMAIL PROTECTED]
