🚀 go-pugleaf

kpwong@psy.cuhk.edu.hk (Kit-pui Wong) wrote:

What rate on the mp3? you could have 320kbps 48khz, or 128kbps 44khz
(which is considered cd-quality), so mp3's can have very good sound.
Afaik atrac is proprietary, and I don't know of any compressor for it
(other than md recorders)


>ATRAC (for MD) and MP3 are both irreversible data "compression"
>algorithms. 
>Which one loses more data during "compression" ? 
>Which one gives worse audio result, theoretically and/or 
>practically ??    
>
>Thanks in advance !

Kit-pui Wong wrote in message <77bvje$q3o$3@justice.csc.cuhk.edu.hk>...
>ATRAC (for MD) and MP3 are both irreversible data "compression"
>algorithms.
>Which one loses more data during "compression" ?

MP3 as typically used does about 12:1 compression, but you have some control
over this. With good compression programs, this can be cut to 6:1 with a
parameter.

>Which one gives worse audio result, theoretically and/or practically ??

This is controversial, but MP3, representing later technology and typlically
using more processing power to code, is considered by many to be more
accurate, in its best implementations.

One problem with MP3 is that people judge it based on files they download from
pirate web sites that are illegally distributing copyrighted music. How these
files are made varies all over the map. Some are obviously taken off FM
radios, others are taken off vinyl (ticks and pops), and many are from CD's
that are converted to analog and then back to digital, or digitally processed
under less-than-optimal conditions.

Some programs that are used to compress MP3 files have known and readily
measurable technical problems, like frequency response aberrations in the
10-12 kHz range, which is readily audible. As a rule, MP3 files have little or
no content above 15 or 17 kHz, but this is a less severe problem, audbily.

Pascal <pdurocher@usa.net> wrote in message
news:369A4971.D050DBD3@usa.net...
>Well... ATRAC is known to suck real badly with it's 4:1 compression (is
>this the one?). It is not using a psychoacoustic model like the MPEG
>Layer 3 standard. But honestly, technically speaking, I don't know much
>about the ATRAC standard other than it sucks. :)

I wouldn't go that far...but I will state for the record one of the main
reasons I sold my MD deck was the ATRAC artifacting. I found it particularly
noticeable and irritating in the upper range of the audible frequency
spectrum. My postulation is that the ATRAC coding scheme is fundamentally
flawed because it allocates a fixed number of bits in the lower (bass) range
of audible frequencies and varies the number of bits devoted to the upper
(mids/treble) range. Since the human ear is most sensitive to distortions in
the mids and treble, you'd think it would make more sense to allocate a
fixed number of bits for encoding that information. On the other hand, I
think if ATRAC were configured in that manner the compression ratio would be
less than 4:1.

>About the MP3 standard, well... you have total control over how much
>compression you get. The "internet standard" of 128kbps will sound okay
>but it is known not to sound right on a good system (I couldn't try it
>myself). You can compress at 160kbps, 192, 256 and 384kbps (or lower
>than 128kbps) for example. At 256kbps experts CAN'T distinguish between
>the original signal and the compressed one.
>
>I know I am going to get BLASTED for that but... imagine this :
>
>You copy 256kbps .mp3's on a CD-R and play them back through your
>hyper-noise PC house and through the S/PDIF output of your Sound Blaster
>card... to your external DAC... and you get equal or better results than
>your CD Transport with alumnium CDs to your DAC...
>Take that ! :)

I agree. What I notice most about MP3 playback on my PC through my stereo is
the vast improvement in dynamics and bass defintion and control. I do trade
an iota of smoothness, but that is probably explained by the fact it's an
MP3. My theory is that the output stage of my AWE32 card provides marginally
higher current than the output stage of my CD player. Or I may not have them
properly level matched.

Adam

>Kit-pui Wong wrote:
>
>> ATRAC (for MD) and MP3 are both irreversible data "compression"
>> algorithms.
>> Which one loses more data during "compression" ?
>> Which one gives worse audio result, theoretically and/or
>> practically ??
>>
>> Thanks in advance !
>

Adam Fuller wrote:
> 
> Pascal <pdurocher@usa.net> wrote in message
> news:369A4971.D050DBD3@usa.net...
> >Well... ATRAC is known to suck real badly with it's 4:1 compression (is
> >this the one?). It is not using a psychoacoustic model like the MPEG
> >Layer 3 standard. But honestly, technically speaking, I don't know much
> >about the ATRAC standard other than it sucks. :)
> 
> I wouldn't go that far...but I will state for the record one of the main
> reasons I sold my MD deck was the ATRAC artifacting. I found it particularly
> noticeable and irritating in the upper range of the audible frequency
> spectrum. 

Mmm... could I ask which particular MD deck you used to use ?  Any idea
of the ATRAC version employed ?  

Without wishing to give the impression that I'll pounce on any anti-MD
rhetoric with all the evangelical zeal of a recent convert, I must say I
find your comments (and those of Pascal) quite surprising.  I have a
relatively modest Sharp 702 portable and I'm getting quite exceptionally
good results with digital recordings.  The relatively subtle differences
that I can ascertain between the MD copies and the CD originals I tend
to put down to the (surely) superior DAC/filter/output circuitry in my
Copland, rather than any shortcomings in the compression algorithm.

But then, I haven't spent *that* long making A/B comparisons as yet...
perhaps my enthusiasm will wane.  

I have yet to venture into the world of MP3, so I can't comment there...

Mike.

What about MP4? Is this better than MP3?

Also, are there any plans for hardware that records
MP3 (or MP4) files on the MD medium? This would be
great! Around twice the length of an MD, with all the
advantages (and possibly better sound).

--
John
busenitz@earthlink.net

Under two kind are bad this conditions, Mp3 maybe better than Atrac in the
record time with walking music.

Still waiting for recorder in mp3 format with mini
dise......................
Kit-pui Wong wrote in message

<77bvje$q3o$3@justice.csc.cuhk.edu.hk>...
>ATRAC (for MD) and MP3 are both irreversible data "compression"
>algorithms.
>Which one loses more data during "compression" ?
>Which one gives worse audio result, theoretically and/or
>practically ??
>
>Thanks in advance !
>

In rec.audio.misc Michael Jones <mikewjones@mcmail.com> wrote:
[snip]
> But then, I haven't spent *that* long making A/B comparisons as yet...
> perhaps my enthusiasm will wane.  
I HAVE spend long hours doing both Blind and sighted A/B comparisons with
my Sony MZ-R50 portable recorder.  Im HUGLY impressed with it.  I couldnt
be happier with the recording quality, which to me IS identical sonically
to CD. 

> I have yet to venture into the world of MP3, so I can't comment there...
I agree totally with Mr Kruguer, in that I think MP3 has been given a very
bad rap.  Ive done quite a bit of compressing with it, and I really think
that when its setup correctly, MP3 can be every bit as good as ALTRAC, and
at higher compression too, I would say.  Still, I will continue to covert
my R50 until a portable mp3 recorder that uses a cheap blank medium, (like
MD) hits the market.

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
~ 	Timothy O'Connor, RMIT. All opinions expressed are my own. 	~
~ You can find yourself a lover, you can make yourself a home, you can  ~
~ want no other ever, but its never to late to be alone...[Del Amitri]  ~
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 

  

Michael Jones <mikewjones@mcmail.com> wrote in message
news:369A8E31.D11C02CE@mcmail.com...
>Adam Fuller wrote:
>>
>> Pascal <pdurocher@usa.net> wrote in message
>> news:369A4971.D050DBD3@usa.net...
>> >Well... ATRAC is known to suck real badly with it's 4:1 compression (is
>> >this the one?). It is not using a psychoacoustic model like the MPEG
>> >Layer 3 standard. But honestly, technically speaking, I don't know much
>> >about the ATRAC standard other than it sucks. :)
>>
>> I wouldn't go that far...but I will state for the record one of the main
>> reasons I sold my MD deck was the ATRAC artifacting. I found it
particularly
>> noticeable and irritating in the upper range of the audible frequency
>> spectrum.
>
>Mmm... could I ask which particular MD deck you used to use ?  Any idea
>of the ATRAC version employed ?

Yes, I was the former owner of a Sony MDS-JE510 and an MZ-E40. Both employed
Sony ATRAC version 4.0. I am debating jumping back in with an MDS-JB920 or
MDS-JA20ES since both employ improved construction, electronics, and the
JA20ES has the variable coefficient digital filter which may be able to
address some of the shortcomings my ears detect with ATRAC. On the other
hand, I really have no good use for a digital recording format of any kind
other than as a conversation piece or occasional experimental tool.

>Without wishing to give the impression that I'll pounce on any anti-MD
>rhetoric with all the evangelical zeal of a recent convert, I must say I
>find your comments (and those of Pascal) quite surprising.

I too was extremely taken with MD until I actually sat and focused on
comparing it to CD. I still think it kills any other format on a portability
vs. sound quality/durability/easy editing basis. My main beefs with the
artifacting

>I have a
>relatively modest Sharp 702 portable and I'm getting quite exceptionally
>good results with digital recordings.  The relatively subtle differences
>that I can ascertain between the MD copies and the CD originals I tend
>to put down to the (surely) superior DAC/filter/output circuitry in my
>Copland, rather than any shortcomings in the compression algorithm.

See, that's what I don't understand. My CD player is extremely modest (Sony
CDP-XA1ES, MSRP $350 USD, paid $200 used) and the MDS-JE510 has a 20-bit
ADAC whereas the XA1ES employs a single-bit DAC, albeit with 45-bit internal
processing and 20-bit outputs according to Sony's website.
The bottom line is that the DAC in my CD player is not leagues better than
that in the MD deck, so I expected the sound to be more comparable than it
was. However, I compared the XA1ES with the Cambridge Audio CD6 in my system
over the course of a weekend and found it acquitted itself well. I would
spend no more than $50 more for the CD6 after that listening experience. In
the end, the three main ATRAC artifacts I found myself noticing were:
sizzling cymbals, brittle mids, esp. on vocals, and a general graininess in
the material. On the other hand, MD quality murders any tape I have ever
heard, even when made and played back on the Dolby S deck I also used to
own.

>But then, I haven't spent *that* long making A/B comparisons as yet...
>perhaps my enthusiasm will wane.
>
>I have yet to venture into the world of MP3, so I can't comment there...

MP3 is extremely good for it's compression abilities. I find well done MP3s
almost indistinguishable from CD material.

Adam

In article <43Bm2.1585$Ge3.8385363@news.itd.umich.edu>,
  "Adam Fuller" <adamf@engin.umich.edu> wrote:

> I really have no good use for a digital recording format of any kind
> other than as a conversation piece or occasional experimental tool.

Funny that.  It's not as if I was aware of any great need for such a device
in my life either (it was a Christmas present), but now, after a couple of
weeks of shuffling, titling, splitting and combining, I look at the teetering
pile of assorted ferrics and chromes next to my Nak (all labelled something
helpful like "Radio 3, March '97") and wonder how I coped at all.

> See, that's what I don't understand. My CD player is extremely modest (Sony
> CDP-XA1ES, MSRP $350 USD, paid $200 used) and the MDS-JE510 has a 20-bit
> ADAC

Well, in that case, I guess you're in a better position than I to make an
assessment of ATRAC's failings (better MD unit, theoretically inferior CD
source).  Thing is, whenever I compare an MD copy to the original I'm reminded
of the shortcomings of my Goodmans CD portable; the Sharp is better but has
traces of that same flatness and lack of life.  Which I suppose is why I have
difficulty putting it down to the ATRAC system - I automatically associate it
with the fact that I'm listening to a 'mere' portable.

> sizzling cymbals, brittle mids, esp. on vocals, and a general graininess in
> the material.

After what I would euphemistically describe as an 'uneven' rendering of Arvo
Part's 'Tabula Rasa' last night, I may be coming round to your point of view.
Looks like the honeymoon might be over...

> On the other hand, MD quality murders any tape I have ever
> heard, even when made and played back on the Dolby S deck I also used to
> own.

... but the marriage has good prospects.

> MP3 is extremely good for it's compression abilities. I find well done MP3s
> almost indistinguishable from CD material.

I really must start using my PC for things other than e-mail ;)

Mike.

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Pascal heeft geschreven in bericht <369A4971.D050DBD3@usa.net>...
>Well... ATRAC is known to suck real badly with it's 4:1 compression (is
>this the one?).

Yes, it's the one (4.9-1 to be exact). Where do you base your opinion on?

>It is not using a psychoacoustic model like the MPEG
>Layer 3 standard.

Isn't it? ATRAC stand for Adaptive TRansform Acoustic Coding. In an article
that expains the workings of ATRAC it is said: ATRAC uses, just like PASC, a
form of acoustic masking. Now the funny thing is that PASC actually is Mpeg
Layer 1 and it can be compared to ATRAC. So it is using a psychoacoustic
model.

>But honestly, technically speaking, I don't know much
>about the ATRAC standard other than it sucks. :)


What recorder did you use and how long ago is it. What did you record and
what was the version of the ATRAC encoder used. I can't hear any difference
between a DAT recording and a MD recording most of the time. Sometimes the
MD recording sounds even better, other times the DAT sounds better.
Differences are small to say the least; using a cheap MDS-JE510 and a pretty
expensive DAT.

>About the MP3 standard, well... you have total control over how much
>compression you get. The "internet standard" of 128kbps will sound okay
>but it is known not to sound right on a good system (I couldn't try it
>myself). You can compress at 160kbps, 192, 256 and 384kbps (or lower
>than 128kbps) for example. At 256kbps experts CAN'T distinguish between
>the original signal and the compressed one.


Same goes for MD. You probably have bad experience with one of the first MD
players (they did sound awfully harsh).

Andr� Huisman
New Line licht & geluid
newline@usa.net
http://www.xs4all.nl/~newline/
--- pardon my French, I'm Dutch ---