macdef wrote:
"Michael Hoffman" <[EMAIL PROTECTED]> wrote:
>>It is grossly irrelevant to assume 128 Kbps MP3 as the standard while
>>assuming 292 Kbps as the MD standard, then say MD sounds better than
>>MP3. This is all too obvious yet people insist on such obviously
>>unfair comparisons.
>
>I never do. I equate MDLP to 292k, MDLP2 to 192 or 256k, MDLP4 to 96k. At
>those levels, MD sounds better on each.
I've read that standard ATRAC is 292 Kbps. Why do you say "192 or 256k" for
MDLP2 - which rate is actually used for MDLP2?
>>The 256 Kbps MP3s I trade sound the same as standard 5:1 ATRAC,
>>because it's the same compression ratio and generally, all the
>>leading compression algorithms perform approximately the same as far
>>as fidelity vs. ratio.
>
>I would disagree. ATRAC is superior to even the best MP3 encoders. Even my
>friends who aren't "audiophiles" can tell the difference between standard MD
>and 256k MP3.
My blindfold headphone tests between WAV and MP3 CBR showed no audible
difference between 224 Kbps MP3 and the input .wav, with the encoder I used.
I have top-percentile hearing and know how to use the gear.
Because my ability to hear artifacts has rapidly improved, I bump up to 256
Kbps, though I consider 224 Kbps to be practically perfect even for critical
headphone listening in a quiet environment. Bitrates can be deceptive -- I've
heard 224 Kbps that sounds as bad as 128 Kbps (actually I suspect it was
re-compressed).
I back away from 320 Kbps because some of today's 1st-generation MP3 CD
players can't play 320 Kbps MP3s. My mind is not completely closed on this --
I fear I am cursed with good enough hearing so that I will learn to hear the
artifacts you claim even at 320 Kbps. I'm an audio fanatic but nowhere near
as much as some people.
>I also don't get the whole thing about titling.
Having to spend any time at all to title a copy is technologically absurd and
obstructs the ease of using MD. Why should I have to title my MD tracks when
the titles have already been entered and uploaded by someone else in CDDB?
You should also acknowledge that most people do not have the ability to plug
in a computer keyboard to type in titles. My setup is more typical -- I have
a $330 brand new top of the line Sony 900 portable, yet Sony expects me to use
the jog dial to spend 15 minutes laboriously titling tracks -- the included
Xitel "digital" interface is the perfect example of digital hype versus
digital reality that falls short of potential.
Why in the hell doesn't Xitel transfer the MP3 ID3 track titling? I am not
impressed. Xitel today offers no advantage over an analog connection (plus
the trackmarks malfunction sometimes). You are too forgiving -- have you
tried to trade 30 albums at a time on MD versus MP3 -- when including titles?
Titling is one of the biggest glaring flaws of MD versus MP3.
MD does titling in the most stupid, boneheaded, manual, tedious,
time-consuming way possible. MP3 titling is a godsend and puts MD titling to
shame. I found that it was easier to trade 30 titled MP3 albums (on CD-R) in
one shot than to trade just 5 titled albums on MD.
MDs are a dead-end for trades and each time you do copy an MD you lose the
titling and introduce another generation of lossy compression -- unlike MP3s.
With MP3s, once you have a copy of the album, you can trade it bit-for-bit
with much less effort than copying one MD to another, because the titles are
transferred inherently as the ID3 tag part of the MP3 file -- no lossy
compression, and no need to manually re-add the titles.
When considering copying albums, MD is a comparatively poor quality, dead end,
and manually intensive medium (as currently packaged). Once you have an MP3
computer with CD burner set up for MP3 CD-R trades, MP3 is bit-for-bit
accurate, ultra-fast, and intelligent medium. I've experienced glimpses of
the best potential for MP3s, and it's much greater than even most MP3
advocates claim.
MP3 technology is now in its 1st generation and has huge growth potential.
Sony knows this. MD is now in perhaps its 3rd-5th generation, and has little
growth potential that is as significant as adding some of the features that
MP3 introduced, such as "sticky titling" that is not left behind when you do a
file copy.
Sony deliberately crippled MD and restricted its potential features (such as
bit-for-bit MD cloning without another ATRAC generation) to appease the RIAA
(that group of industry lawyers specializing in ripping off musicians and
crippling technology). But now Sony's restriction of MD capabilities is
biting Sony in the ass, because MP3 technology does not deliberately cripple
itself like Sony did to MD. MP3 files are copied bit-for-bit, of course
including the titles.
Micronics MD titler
http://www.mironics.com/default.asp?get_Click=product&get_Pointer=3
>First of all, for me MP3 and
>MD are equally easy/a pain, because I can plug a PS/2 keyboard (which I
>bought for $2) up to my MD recorder and title songs. So it's just as easy.
>But second, I just don't get why people want to do this so much. After all,
>when you buy a CD, are the tracks labeled? Only if you have one of the very
>rare CDs with CD-TEXT, and one of the rare players that can display it. The
>only reason I could see for needing titling is if you have a hard drive full
>of songs. However, since that isn't possible with MD, the argument is
>largely irrelevant. If MD *was* able to store songs on a hard drive, but
>didn't allow you to title them, that would be a valid criticism.
I hate the way CDs only show track numbers instead of titles. CDs certainly
should have all had CD-text and bitmap displays on all players since 1983.
This is especially important when you have 25-100 tracks on one piece of
storage media. For example, my CD jukebox stores 200 albums but it is a
completely unusable disaster of a user interface, because "disk 183, track 17"
is utterly meaningless.
The people who are satisfied with today's approach to MD titling are those who
have not gotten into large collections of music where managing tracks and
albums becomes a major issue.
>> We should try to combine the strengths of MD and MP3.
>
>That is the key point of all this, and I agree wholeheartedly.
Yes, that needs to be re-emphasized. MP3 throws down a significant challenge
to MD -- add these features such as CDDB integration and *reliable* and
automatic trackmarks, or be left behind by some other solution that
incorporates these great features.
>Larry wrote:
>>Also, even MP3s are an improvement over vinyl, cassettes and FM.
I don't like that way of talking, because it lacks qualifiers. I would say
that the best MP3s have the potential to sound overall more high-fidelity than
ordinary vinyl-based systems.
>>>Mp3 compression generally is not as good as MD.
>>
>>Generally ? meaning what ? it obviously depends on where you get the
>>mp3. When downloading from kazaa and the like, the quality indeed
>>varies <greatly.BUT> all the mp3s i created myself are either 256 or
>>320 kbps cbr or lame vbr. This is equivalent to cd-quality
I agree, based on my blindfold tests with certain encoders. Feel free to slam
other people's MP3s, but I don't believe you could tell my 224 Kbps MP3s apart
from the CD in a critical headphone listening test. I have been using Music
Match for a couple years and they have changed the encoder and I might switch
to more dedicated and controllable rip/encode software. I need to test more
encoders.
>No offense, but 320k lame is *not* "CD-quality" -- anyone with a good system
>and decent ears can hear the difference even in a double-blind test.
I have not tested Lame, so no opinion here. I have top-percentile hearing so
I would be a good person to blindfold-test this encoder. If *I* can't
reliably tell the difference, then the difference is insignificant to the
general population. The way you put it, the difference is very blatant, so I
am open minded and would like to hear your tests myself.
> If you mean "close enough so that in a portable environment where there is
motion and background noise you won't know the difference," then I agree with
you.
>But in terms of actual sound quality, no.
I'm trying to express my version of "high but reasonable" standards. I do not
accept music that sounds compressed, not even when I'm on the bus. My music
must be indistinguishable from the CD when I hear it with good headphones in a
quiet room with blind A/B testing.
My previous tests showed this to occur at 224 Kbps -- but my darn ears keep
learning to detect artifacts, so I'm open-minded. Maybe 320 Kbps CBR with a
good encoder could sound bad to me, after a little more experience. I am
starting to hallucinate lossy-compression artifacts on my prerecorded CDs,
like an audiophile at a symphony concert who hears clipping.
>As for MP3 vs. MD, standard-bitrate ATRAC is superior to even lame encoding
>at 320k. I've compared it myself using good equipment.
I have not done an A/B test of ATRAC vs. 320 Kbps MP3.
>Again, maybe this doesn't matter to people using MD/MP3 in a portable, or on
a computer where the sound isn't high-quality anyway,
I reject the idea that computers sound inferior to a home stereo. I've had no
problem getting high fidelity out of computers. I'm not into ultra-fidelity,
but there are certain important things -- (reasonably) full frequency
response, no distortion or background noise. I simply don't want to hear
audible background noise or muted highs, or audible distortion, when listening
in headphones or with half-decent speakers.
On a scale of 1-10 where 1 is the person who calls 128 Kbps "CD-quality" and
10 is the extremist audiphile, I'm an 8. This says a lot: most 192 Kbps CBR
files are unlistenable for me, because the artifacts are so strong they make
all the music sound blatantly overcompressed. I continue to move toward
headphones rather than speakers, even in a quiet room, so that implies a
certain high standard for fidelity is required. If there are any *obvious*
artifacts, I will hear them.
My critical listening to MP3s has revealed many flaws in the CD itself, and
the studio mastering, that I would not have noticed. This is important: I now
have greater fidelity problems with the studio recording and mastering on the
CD than with my own playback equipment. For example, I have to change my
bass/mid/treble from one song to the next on a compilation album.
I've started hearing more flaws in the sound on the pressed CDs themselves,
whether pops or muted lead-ins or other glitches -- like the CDs used to say,
"Because digital is so perfect, flaws in the original analog recording may be
apparent."
>but in terms of absolute sound, I just get tired of the term "CD-quality"
used so loosely. Neither MD nor MP3 is "CD-quality." They are both ways of
compressing sound so that you don't need as much space to store it. They are
both lossy, and audibly so.
Audibly so? Here is where we can have a true debate. Here is a way we can
sort ourselves into groups. Who here says that MD is audibly inferior to CD?
Not me, not many others. That's why I coined this expression, "MP3 has the
potential to be MD-quality" -- a very solid and technically correct
comparison.
If you are aiming for CD quality and consider MD to sound inferior to CD, then
I have no discussion with you (so to speak). My goal is humbler than yours: I
consider MD to sound as good as CD, and my goal is simply to make MP3 sound
"MD-quality" rather than "CD-quality".
To do this, the most practical approach is to simply encode MP3 at the same
rate as MD's ATRAC. Any improvements beyond the sound quality of MD, I'm not
terribly concerned with. This helps keep my goal versus your goal in
perspective.
I respect the third goal some other people have: to find a good quality/size
tradeoff to use with today's 1st-generation half-baked MP3 solid-state players
or 56 K modems. My attitude toward MP3 is an MD/CD approach, emphatically
*not* a modem/64 MB Rio approach. The battle I'm fighting is to divide MP3
into two very different camps.
There are two opposite approaches to MP3:
A. The first wave wants to trade music free via slow modem and play it back on
tiny 32 MB (!) solid-state players. For this, highly compressed, low-fidelity
MP3 compression was needed.
B. The second wave, like me, wants to trade music via cable modem, or actually
CD-R, and play it back on a 256 MB or greater player, with the impressive
fidelity of a portable CD player. For this, the MD 292 Kbps 4.3:1 (correct?)
amount of compression is ideal -- that amounts to 256 Kbps or 320 Kbps CBR MP3
encoding.
My main problem debating MP3 versus MD is that MD people mistake me for a
first-wave lo-fi 128 Kbps MP3 person -- I'm not. I battle against 128 Kbps
MP3 and I'm trying to free the medium to realize its more bloated potential,
basically 256 Kbps, using larger storage capacity.
I ordered my computer with two 30 GB drives, which used to be considered a
huge amount of storage space. Today I would order two 80 GB drives. I hope
to switch from CDR to DVD for trades, when the DVD standards settle, and
perhaps for portables switch to the higher-than-CD density optical storage
such as Dataplay (MD is too bulky, anyway).
It was unfortunate that the marketroids hyped 128 Kbps MP3s as "digital music
with CD quality". I simply cannot stand listening to music compressed so
much; it's an insult to the ears and to the music.
Pet peeve: Rio-type box copy that equates 128 MB RAM with "an hour of
CD-quality music". For the high-fidelity type of MP3 approach, 128 MB RAM
equates to only one hour of music (2 MB RAM for 1 minute of music, when you
assume 256 Kbps encoding).
Look at this drivel that plagues today's audio magazines every time they
review MP3 -- are they out to kill it?
http://www.computeraudio.co.uk/caw/cawfeatureshtml/cawwhatismp3.html - "One of
MP3's most popular bit rates is 128kbps because it's a fine trade-off between
disk space, encoding speed and sound quality, but you can go down to save
space. In fact, MP3 offers a whole host of different quality options: 8kBit
mono, 8kHz; 16kBit mono, 16kHz; 24 kBit mono, 22kHz; 32 kBbit mono, 22kHz; 56
kBit stereo, 22kHz; 64 kBit stereo, 22kHz; 64 kBit mono, 44kHz; 96 kBit
stereo, 44kHz; 112 kBit stereo, 44kHz; 128 kBit stereo, 44kHz."
He doesn't even *mention* the existence of rates higher than 128 Kbps, though
he does say "high-bitrate" without stating numbers. He implies that 128 is
"high".
"high bitrate MP3 can sound surprisingly good. MiniDisc is a good yardstick.
First generation ATRAC MDs sounded poor - flat, uninvolving and processed.
Third generation ATRAC added musicality and pep, whereas fourth and fifth
generation MD are impressively natural. By comparison, 128kBit MP3 lies close
to third generation ATRAC MD."
I have enjoyed this magazine: Hi-Fi World & Computer Audio. July 2001.
http://www.computeraudio.co.uk
The magazine Computer Audio World is inside the magazine. Page 55:
Sony On The Move -- Sony's new Digital Relay series... a Discman and a
portable CD writer. Wired remote w/ LCD, 2 hours of CD recording, headphones.
http://www.oreillynet.com/pub/a/mac/2001/06/27/sony_digital_relay.html
It's a portable CD burner packaged like a portable CD player, with headphones
and MP3 as well as CDDA playback ability. Now imagine it with Mini CD-R. Now
imagine it with Mini CD-RW. Now imagine it as small as a round equivalent of
the square little Sony 900. Look out MD!
Sony has to plan and think in the 2-year timeframe and they know that MD is
dead meat in that time frame if MP3 and high-density CDRW reaches its
potential, which we can expect it to. Some upstart company, or many
companies, might very well engineer a fully MP3-based perfect MD-workalike.
The MD community should be extremely interested in this near-future
possibility.
You guys might not see it, but Sony, inventor of CD and MD, certainly sees it:
the entire concept of MD must be redesigned from the ground up with full
ability to be computer-integrated (as well as being as useful as a standalone
portable MD recorder). Don't live in today; live mentally 2 years in the
future, as Sony must always do.
http://computers.cnet.com/hardware/0,10000,0-1095-404-4920979,00.html?dir -
Sony Digital Relay
Editor's choice. CNET rating 9/10. "Sony's Digital Relay Drive gives a whole
new meaning to burning on the fly. It's the first portable burner designed
expressly with today's music enthusiast in mind, and it can play and record
both audio and MP3 discs. If you're in the market for a CD-RW drive and an MP3
player, Sony's Digital Relay Drive offers a terrific all-in-one solution. Why
settle for just an MP3 player when you can have an MP3 burner, an excellent
backup and recovery solution for your data, and a portable CD player in one
device? The Sony Digital Relay isn't cheap [$399, compare $330 MZ900 MD
recorder], but Sony paid attention to all the details, making this a reliable,
useful drive."
Why doesn't the Sony Digital Relay use ATRAC rather than MP3? How long before
ATRAC and MP3 fully collide within the Sony product line?
-- Michael Hoffman
http://www.amptone.com/audio
-----------------------------------------------------------------
To stop getting this list send a message containing just the word
"unsubscribe" to [EMAIL PROTECTED]
