HydrogenAudio

Hydrogenaudio Forum => Listening Tests => Topic started by: omikron23 on 2004-07-09 00:04:23

Title: ATRAC3 superior or inferior to mp3 ?
Post by: omikron23 on 2004-07-09 00:04:23
Hi there,

There are two interesting tests on ATRAC3

1:  http://www.sony.net/Products/ATRAC3/tech/lab/ (http://www.sony.net/Products/ATRAC3/tech/lab/)
      http://www.sony.net/Products/ATRAC3/tech/l...test_report.pdf (http://www.sony.net/Products/ATRAC3/tech/lab/ITS_test_report.pdf)

2:  http://www.rjamorim.com/test/multiformat128/results.html (http://www.rjamorim.com/test/multiformat128/results.html)

The signal analysis in the ITS report suggests ATRAC3 being far superior to mp3 and wma, whereas the subjective listening tests in test 2 suggest it is the worst.

Can signal analysis be relied on, when it comes to subjective perception of audio signals ?
Is it possible, that a bad codec colours the sound file in such a way, that it  actually sounds better but not authentic ?
Perhaps a good codec should not sound good, but reproduce in a transparent way.  What do you think ?
Title: ATRAC3 superior or inferior to mp3 ?
Post by: Dibrom on 2004-07-09 00:17:07
Quote
Can signal analysis be relied on, when it comes to subjective perception of audio signals ?
[a href="index.php?act=findpost&pid=224268"][{POST_SNAPBACK}][/a]


No, at least not without taking into account certain aspects of the faculties which color its perception.  We aren't testing for signal level accuracy, we are testing for how it appears to sound.  There is not a linear correlation between the two.

Quote
Is it possible, that a bad codec colours the sound file in such a way, that it  actually sounds better but not authentic ?


Sure.

Quote
Perhaps a good codec should not sound good, but reproduce in a transparent way.  What do you think ?


Well then all you're doing is redefining the term "good codec."  You can define it to mean whatever you want, whether that means objective accuracy, or subjective accuracy.  What the judgement hinges upon then is what is important to you.  Most people on this board will probably choose sound quality over pretty looking frequency sweeps, but then your mileage may very.

Note though that in your case "good codec" has nothing to do with perceived sound quality and is simply a rather arbitrary designation.

In my opinion, what is important is that it "sounds good" and "sounds transprent".  If these conditions are met then whether or not the encoded signal deviates more statistically from the original signal in comparison to another so-called "good codec" is not important (in fact, to favor that approach would be to miss the entire idea behind psychoacoustic lossy audio compression).
Title: ATRAC3 superior or inferior to mp3 ?
Post by: Pio2001 on 2004-07-09 00:42:21
Quote
Can signal analysis be relied on, when it comes to subjective perception of audio signals ?[{POST_SNAPBACK}][/a] (http://index.php?act=findpost&pid=224268")


This question is answered in the FAQ. Here's a direct link : [a href="http://www.hydrogenaudio.org/forums/index.php?showtopic=7834&st=0&p=77265&#entry77265]http://www.hydrogenaudio.org/forums/index....265&#entry77265[/url]

We can notice also, that their subjective tests both show ATRAC3 132 kbps superior to MP3 128 kbps, with proper blind listening methods, and with confidence intervals of 95 %.
If I'm not mistaken, MusicMatch Jukebox uses Fraunhofer encoder, like CoolEdit. The MP3 files they got being slightly inferior to ATRAC, it shows the overwhelming superiority of the Lame VBR mode used in Roberto's test over the CBR fraunhofer version they used.
Title: ATRAC3 superior or inferior to mp3 ?
Post by: Pio2001 on 2004-07-09 01:09:23
Oops, I missed a serious flaw in the two tests linked above : one of them was run with 4 musical samples given by Sony, and only 1 chosen by the lab, and in the other, 7 samples were chosen by Sony, while 3 were chosen by the lab.
It is likely, Sony having very big interests in the ATRAC success, that they gave musical samples especially chosen so as to advantage ATRAC over other codecs. Thus there might be a strong bias for ATRAC in both tests.
Title: ATRAC3 superior or inferior to mp3 ?
Post by: saratoga on 2004-07-09 01:44:09
Quote
The signal analysis in the ITS report suggests ATRAC3 being far superior to mp3 and wma, whereas the subjective listening tests in test 2 suggest it is the worst.


Test 1 does not use a relevent MP3 codec, so its results are themselves irrelevent (at least with regard to MP3).
Title: ATRAC3 superior or inferior to mp3 ?
Post by: rjamorim on 2004-07-09 02:14:28
Quote
Oops, I missed a serious flaw in the two tests linked above : one of them was run with 4 musical samples given by Sony, and only 1 chosen by the lab, and in the other, 7 samples were chosen by Sony, while 3 were chosen by the lab.
It is likely, Sony having very big interests in the ATRAC success, that they gave musical samples especially chosen so as to advantage ATRAC over other codecs. Thus there might be a strong bias for ATRAC in both tests.
[a href="index.php?act=findpost&pid=224287"][{POST_SNAPBACK}][/a]


heehee. How clever of them.

Besides, I don't believe in paid-for tests mentioned at the company's site. They could well have asked 20 labs to test their codec, and only choose to announce the results that hailed it as winner.
Title: ATRAC3 superior or inferior to mp3 ?
Post by: MugFunky on 2004-07-09 05:40:09
heh.  poppycock.  totally poppycock.

perhaps if microsoft ran these exact same tests there might have been different results?  or perhaps FhG? (or Xing...)

true, ATRAC3 sounds better to me than ATRAC, but it's still very much lacking as a serious codec, in my opinion.  (ATRAC is rather revolting to listen to, especially on human speech)

sony should stick to what they do best - making VCRs, and discmans that clip at half volume.
Title: ATRAC3 superior or inferior to mp3 ?
Post by: Mystic_Image on 2004-07-09 14:17:02
I've been lurking for awhile, but I think it's time to speak up a bit.

I think there has been a somewhat undeservedly negative view taken towards ATRAC3 AND MiniDisc in general, without a real understanding of why ATRAC3 is the way it is and what it is targeted to.

I felt that the multiformat128 test result was probably accurate in its results, but I don't feel that the test was particularly fair for ATRAC3, and I also feel that no real conclusions can be made directly regarding the performance of ATRAC3 (MDLP) on an actual MD player.

With regards to an actual MD player, assuming that the DSP on the minidisc player is in use, I'd like to note that:

1. ATRAC Type S, used in the newer lines of MD recorders is actually a different encoder (and decoder?) for MDLP/ATRAC3 compared to the one used originally when the Sony R900 series came out, and it purportedly improved quality. In addition, though it is compatible, I don't think it is the same as the software codec that is used by SonicStage on the PC, which more than likely uses something similar to the original algorithm used in the first generation of MDLP recorders. Unfortunately, this makes it difficult to truly compare performance when it comes to MD players versus MP3 players.

2. On a personal note, I've always found 132kbps ATRAC3/LP2 mode to be superior to CBR 128kbps MP3 (assume Frauenhofer). Perhaps, more equivalent to 160kbps. I don't have the ABX test to show that personally onhand, but I find that the main difference is that ATRAC3 does not suffer from upper-frequency artefacting nearly as much. It might be that ATRAC3 changes the sound from the original, but it certainly less grating to my ears compared to CBR MP3 at less than 160kbps.

3. Why is it NOT so fair to compare Minidisc ATRAC3 to VBR MP3?
The whole reason of ATRAC3's CBR has to do with backwards compatibility to a technology that is over 10 years old! I believe that because of CBR and its less strenuous requirements on hardware design and flexibility, Sony and other manufacturers have been able to achieve much better battery life and editing flexibility with MiniDisc players compared to competitors, along with much a cheaper storage medium (at this point) on which to store music. ATRAC3 had to conform to a preexisting standard and is aimed at a certain specific use that the other codecs don't really do.

Note that as it is, today, for most music you still find 128kbps CBR MP3 as the standard, so I wouldn't rap on the MD users for saying that they find it to be superior to MP3, because in general, for most of the music they've heard, it probably IS better. Yup, maybe they're a little less informed than the crew here.

I won't argue that on today's terms, that if you were going to choose the best-quality codec with which to store your music on a computer or hard disk at ~128kbps for some reason, that you should choose the ATRAC3 software encoder. Certainly, the other VBR codecs will do a better job on a computer. However, in use for a portable device, I think it's still quite competitive for what it can do.
Title: ATRAC3 superior or inferior to mp3 ?
Post by: Latexxx on 2004-07-09 15:23:54
Quote
Frauenhoefer

I'm not going to tell you anything about this funny spelling mistake.

Instead of it, I tell you that even Sony is pushing Sonic Stage 2 as the main Atrac encoder nowadays. First of all, I higly doubt that record companies use hardware encoders to encode their files for Connect instead of that they propably have some software solution which is based on the same software encoder as Sonic Stage. And! The new Hi-MD devices are all about transferring music to device from CDs using Sonic Stage. And the all mighty and definite answer: This thing was discussed about in the threads of the listening test! You should read them before posting.
Title: ATRAC3 superior or inferior to mp3 ?
Post by: Mystic_Image on 2004-07-09 15:51:05
Quote
Quote
Frauenhoefer

I'm not going to tell you anything about this funny spelling mistake.

Instead of it, I tell you that even Sony is pushing Sonic Stage 2 as the main Atrac encoder nowadays. First of all, I higly doubt that record companies use hardware encoders to encode their files for Connect instead of that they propably have some software solution which is based on the same software encoder as Sonic Stage. And! The new Hi-MD devices are all about transferring music to device from CDs using Sonic Stage. And the all mighty and definite answer: This thing was discussed about in the threads of the listening test! You should read them before posting.
[a href="index.php?act=findpost&pid=224392"][{POST_SNAPBACK}][/a]


Whoops... bad spelling mistake... corrected... :-)

I did read the threads of the listening test (though admittedly not in the past 2-3 weeks) but I felt that people were making comments that weren't particularly in context or fair regarding ATRAC and MiniDisc in particular.

When it comes down to Hi-MD we can't make a conclusion either as none of us in our right minds would be using ATRAC3 MDLP2 modes anymore. I don't particularly care about what Sony tells me to use - otherwise I'd be encoding everything at 64kbps in LP4 mode, believing that I've got 320 minutes of listenable music on my MD. We haven't done any testing on ATRAC3Plus at 256kbit modes so I don't think it's easy to conclude.
Title: ATRAC3 superior or inferior to mp3 ?
Post by: magic75 on 2004-07-09 15:53:51
1. Creating an high quality audio codec for hardware is much more difficult than for software, due to constraints as complexity, battery life etc. as you pointed out yourself under item 3. Why on earth would Sony cripple their software to be worse than this? Wouldn't they wanna showcase their codec at their best? It would be the worst marketing ever to actually reduce quality.

2. Well, claims without proof usually isn't accepted around here. Why not just perform an ABX test yourself quickly, and you will get taken much more seriously here. And personally I can't understand howcome a codec that (maybe) colors the sound (as you speculate yourself) can be better. To me the only thing important is staying as close to the original.

3. For me this test is as fair as it gets. Its about audio quality and nothing else. Maybe the MiniDisc players do have excellent battery life and cheap storage, but how is that relevant to audio quality? If we were to take such things into account in these kinds of tests, any ranking would be impossible. And in addition to portable use, the test was also aimed at finding out what online music stores uses the highest quality audio format, of which ATRAC is what is used in Sonys music store (don't remember the name).
Title: ATRAC3 superior or inferior to mp3 ?
Post by: saratoga on 2004-07-09 16:39:17
Quote
3. Why is it NOT so fair to compare Minidisc ATRAC3 to VBR MP3?
The whole reason of ATRAC3's CBR has to do with backwards compatibility to a technology that is over 10 years old! I believe that because of CBR and its less strenuous requirements on hardware design and flexibility, Sony and other manufacturers have been able to achieve much better battery life and editing flexibility with MiniDisc players compared to competitors, along with much a cheaper storage medium (at this point) on which to store music. ATRAC3 had to conform to a preexisting standard and is aimed at a certain specific use that the other codecs don't really do.

Note that as it is, today, for most music you still find 128kbps CBR MP3 as the standard, so I wouldn't rap on the MD users for saying that they find it to be superior to MP3, because in general, for most of the music they've heard, it probably IS better. Yup, maybe they're a little less informed than the crew here.


And MP3 is not an ancient spec    Last I checked it was also more then ten years old and filled with stupid limitations like the sfb21 problem.  How does this make it unfair to compare to ATRAC3?  If anything i think that ATARC looseing to such an outdated codec just make's ATRAC look less appealing, certainly not more as you seem to imply.

Quote
On a personal note, I've always found 132kbps ATRAC3/LP2 mode to be superior to CBR 128kbps MP3 (assume Frauenhofer). Perhaps, more equivalent to 160kbps. I don't have the ABX test to show that personally onhand, but I find that the main difference is that ATRAC3 does not suffer from upper-frequency artefacting nearly as much. It might be that ATRAC3 changes the sound from the original, but it certainly less grating to my ears compared to CBR MP3 at less than 160kbps.


Check out rule #8 from the TOS:

Quote
8. Any statement about sound quality must be supported by the author responsible for such statements by a double blind listening test demonstrating that he can hear a difference, together with a test sample.


Perhaps you could post results?  Afterall we already have a lot of double blind test results saying ATRAC is a lot worse then LAME.  Some fhg tests on the samples you're refering to would be interesting to look at.

Edit:  Opps missed something with a quote.
Title: ATRAC3 superior or inferior to mp3 ?
Post by: Mystic_Image on 2004-07-10 05:31:49
Quote
1. Creating an high quality audio codec for hardware is much more difficult than for software, due to constraints as complexity, battery life etc. as you pointed out yourself under item 3. Why on earth would Sony cripple their software to be worse than this? Wouldn't they wanna showcase their codec at their best? It would be the worst marketing ever to actually reduce quality.

2. Well, claims without proof usually isn't accepted around here. Why not just perform an ABX test yourself quickly, and you will get taken much more seriously here. And personally I can't understand howcome a codec that (maybe) colors the sound (as you speculate yourself) can be better. To me the only thing important is staying as close to the original.

3. For me this test is as fair as it gets. Its about audio quality and nothing else. Maybe the MiniDisc players do have excellent battery life and cheap storage, but how is that relevant to audio quality? If we were to take such things into account in these kinds of tests, any ranking would be impossible. And in addition to portable use, the test was also aimed at finding out what online music stores uses the highest quality audio format, of which ATRAC is what is used in Sonys music store (don't remember the name).
[a href="index.php?act=findpost&pid=224405"][{POST_SNAPBACK}][/a]


1. Yes, creating a hardware codec is certainly difficult... but Sony has two sides which have constantly been at each other's throat these days: one is the technology, and the other is the music industry, so really, you could argue that Sony's music / software side has no interest in improving the codec, as they don't want people just distributing their ATRAC files everywhere...

Obviously, on the other hand, they'd like to show their best hand (which hand?), but I think making it exclusive to the hardware can actually make it more attractive to purchase it.

Regardless of this speculation, the fact remains that they changed their ATRAC3 codec from the original hardware and we do not know whether the software version was updated at the same time.

2. An ABX test is something I will certainly try and set up when I have time. In regards to colouring sound, I mean that it is coloured in a way that is much less obvious/distracting compared to a Fraunhofer CBR codec at 128kbps, at least to me. Obviously, I can't make any real claims at this point according to the forum rules.

3. If the whole point is to compare quality (regardless of application), I would accept the results without reservation. On a quality basis, I do accept the results reached. The issue is that really, I don't think ATRAC3's design goals really fit the same profile as any of the others. You're right, this makes it very difficult if not impossible to compare, especially if it comes down to hardware. For portable use, I really do think other factors come into play and that quality must be balanced at times against them. Otherwise, we'd all be walking around with portable tube amps and battery backpacks.

On the side, if one of the points here was to compare online music store codecs, then I don't think that was achieved by including VBR LAME MP3 (never mind Vorbis or MPC even), as it is not being used by any of them (aside from allofmp3.com, possibly, though the legality of that service is in question).

As a result of this test, everyone has taken a liking to criticizing users of MDs (and of course, I'm one of those) because the sound quality is obviously inferior to some available solutions... but comparably speaking, I believe that most of these MD users are talking from a 128kbps Fraunhofer MP3 CBR (which is what the majority of MP3s have been encoded in) vs. ATRAC3 132kbps situation.

To Mike Giacomelli:

I see nothing wrong with comparing ATRAC3 to MP3... but I do think the settings and codec used are important. MP3 is old too, but it wasn't designed with particular hardware specifications in mind, and certainly it didn't really show up until much later than MiniDisc did. Note that LAME doesn't count in this case since it is hardly the age that MD is on those terms. Also note that MiniDisc took off in Japan/Asia in general when it was introduced, so that's why I'm saying that it had a presence compared to MP3, which wouldn't have been useable at the time at all because computers that were powerful enough to decode MP3 weren't readily available. Also, as mentioned in the listening results thread, ATRAC3 can't simply just be constantly developing, as it is tied to hardware, so to try to compare something that hasn't had the chance to develop...

Again, I'm NOT comparing ATRAC3 vs LAME... I'm saying that a more useful comparison is ATRAC3 vs. Fraunhofer, if we want to understand why people have argued that MiniDisc is superior, as opposed to just putting them down as unreasonable. LAME is better in many ways, but I'm trying to compare things from a more real-world applied point of view. I guess I should do LAME CBR as well as Fraunhofer. Something to put on the list.

Generally, I think most of this will be pretty moot as standards reach higher bitrates, 256kbps and above, as they all seem to be heading. The only reason NOT to go higher is so that the music industry can continue to think they are not just giving away all their money if they use higher bitrate songs, in which case this is their justification for limiting the quality of their songs distributed online.
Title: ATRAC3 superior or inferior to mp3 ?
Post by: spase on 2004-07-10 06:09:12
I also use MiniDisc on a fairly regular basis, although I have a Sharp unit.  This hasn't even been taken into account, as early on it was pretty much accepted, as far as I know, that the Sharp DSP (especially version 6 and higher) were better than the Sony DSP's available at the same time.  I haven't seen any proof of this however, just heresay, and as for today's DSP's, I have no idea which is better.

That set aside, I would be interested in some test sample of ATRAC3+ (the codec used in the new Hi-MD units) at 256kbit (perhaps someone with a Hi-MD can record it via optical and post a sample?).

Also I think it should be mentioned that no one seems to use ATRAC, ATRAC3 or ATRAC3+ as a codec for storing music on their computer (in general...), while MP3 is widely used.  I for one use my MiniDisc unit (and therefore use ATRAC) for running or other active outside uses, which introduces a lot of outside noise, which brings me to my final point....

I think it would be interesting to run some sort of ABX test in a noise outdoor environment to find out which codec really performs the best in this situation -- an active environment with a reasonable amount of ambient noise/background noise... maybe like on a train or bus (as I would imagine this is where most users use their portables...)

Maybe then we could see which is the best portable codec, or if they all end up sounding generally the same, due to the extraneous noise.
Title: ATRAC3 superior or inferior to mp3 ?
Post by: Dibrom on 2004-07-10 07:19:10
Quote
I think it would be interesting to run some sort of ABX test in a noise outdoor environment to find out which codec really performs the best in this situation -- an active environment with a reasonable amount of ambient noise/background noise... maybe like on a train or bus (as I would imagine this is where most users use their portables...)
[a href="index.php?act=findpost&pid=224547"][{POST_SNAPBACK}][/a]


All other conditions being relatively equal (bitrate, etc.), I think it should pretty much always be the case that the codec that performs best in a quiet environment is also going to sound the best in a noisy environment, unless something really strange is going on.

The noise might diminish ones ability to discern deviations in quality compared to the original sample, but I don't think it should fundamentally change the sound in a way which makes the kind of difference you're talking about.
Title: ATRAC3 superior or inferior to mp3 ?
Post by: MugFunky on 2004-07-11 18:27:14
it's conceivable that in a particular noise environment the spectral nature of the noise might mask some corresponding deficiencies in the codec.

for example the harmonic whine of a jet engine could well make ATRAC listenable...

btw, does anyone have some tech details for ATRAC/ATRAC3?  in particular where the scalefactor bands lie, and things like that.  from looking at spectrograms (please don't hit me), ATRAC3 seems derivative of mp3 in many respects, whereas ATRAC (SP mode) is a completely different beast.

of course that says nothing of the sound quality - i've found ATRAC3 to sound better than ATRAC in most cases, but i've only really dealt with human speech with MD sources.
Title: ATRAC3 superior or inferior to mp3 ?
Post by: Latexxx on 2004-07-11 18:52:58
Quote
btw, does anyone have some tech details for ATRAC/ATRAC3?[{POST_SNAPBACK}][/a] (http://index.php?act=findpost&pid=224936")


Minidisc.org has [a href="http://www.minidisc.org/aes_atrac.html]a nice paper on Atrac[/url]. They also have some patents (http://www.minidisc.org/patents/) listed. And some final links:
http://www.minidisc.org/mdlpfaq.html#r_q60 (http://www.minidisc.org/mdlpfaq.html#r_q60)
http://www.minidisc.org/mdlpfaq.html#r_q75 (http://www.minidisc.org/mdlpfaq.html#r_q75)
Title: ATRAC3 superior or inferior to mp3 ?
Post by: Gabriel on 2004-07-12 08:49:04
*ATRAC3 is NOT backward compatible with ATRAC
*MP3 standard was finalized in 1992
*ATRAC3 was introduced several years after
Title: ATRAC3 superior or inferior to mp3 ?
Post by: Mystic_Image on 2004-07-12 21:08:30
Quote
*ATRAC3 is NOT backward compatible with ATRAC
*MP3 standard was finalized in 1992
*ATRAC3 was introduced several years after
[a href="index.php?act=findpost&pid=225152"][{POST_SNAPBACK}][/a]


It is quite right in the sense of a CODEC, but not in the sense I was talking about. I was addressing the fact that ATRAC3 had to be useable on a hardware platform 10 years old, i.e. MiniDisc (which just happens to use ATRAC).

Even though ATRAC3 is new(er), it is restricted on a fundamental basis by its tie-in to the MiniDisc platform, which is relatively quite old.

One of the more obvious results of this restriction was that even though ATRAC3 should have been half the bitrate of MiniDisc (292kbps), it ended up being at 132 instead of 146, with 14kbps wasted as blank space in the name of compatibility with older MD players (so they wouldn't play noise).
Title: ATRAC3 superior or inferior to mp3 ?
Post by: Gabriel on 2004-07-12 21:54:19
Quote
It is quite right in the sense of a CODEC, but not in the sense I was talking about. I was addressing the fact that ATRAC3 had to be useable on a hardware platform 10 years old, i.e. MiniDisc

Wrong.
ATRAC3 is not usable on older minidisc players.
Title: ATRAC3 superior or inferior to mp3 ?
Post by: guruboolez on 2004-07-13 00:38:58
Quote
Oops, I missed a serious flaw in the two tests linked above : one of them was run with 4 musical samples given by Sony, and only 1 chosen by the lab, and in the other, 7 samples were chosen by Sony, while 3 were chosen by the lab.
It is likely, Sony having very big interests in the ATRAC success, that they gave musical samples especially chosen so as to advantage ATRAC over other codecs. Thus there might be a strong bias for ATRAC in both tests.
[a href="index.php?act=findpost&pid=224287"][{POST_SNAPBACK}][/a]


You've maybe missed this other important flaw:
Code: [Select]
1.3 - Listening test material

All Atrac3 and Atrac3plus files were encoded and converted by Sony Corporation. The Laboratory takes on trust that these files were encoded with good intention and at the correct bitrate.


Highly scientific approach 
Samples came from Sony, encoded material from Sony, headphones from Sony. I wonder if listeners are not also from Sony.

Anyway, they didn't precise the mp3 settings used. Are files in full-stereo like Atrac3 at the same bitrate? Could be possible.


Other important thing to precise: the frequency graphs at the end of the report (APPENDIX III) are showing a strange cut-off behaviour of atrac3plus:
- 20 Khz lowpass with atrac3plus@48 kbps
- 21 Khz lowpass with atrac3plus@64 kbps
- 22 Khz lowpass with atrac3@132 kbps
If you look on the lowpass used in commercial atrac3/atrac3plus encoders, these values are totally different
- ~18 Khz for atrac3@132
- not sure for the other, but it must be something like 16 KHz for 64kbps and 14 KHz for 48 kbps.

These graphs are very important: they are proving either that encoders used for this test are totally different from encoders available for MD users via SonicStage or they are proving that Sony didn't respected the bitrate asked for this test.
Title: ATRAC3 superior or inferior to mp3 ?
Post by: Mystic_Image on 2004-07-13 02:47:35
Quote
Quote
It is quite right in the sense of a CODEC, but not in the sense I was talking about. I was addressing the fact that ATRAC3 had to be useable on a hardware platform 10 years old, i.e. MiniDisc

Wrong.
ATRAC3 is not usable on older minidisc players.
[a href="index.php?act=findpost&pid=225365"][{POST_SNAPBACK}][/a]


I apologize for my use of the word 'useable' in a different manner, as ATRAC3 is indeed not possible to LISTEN to on older hardware, but if you read my argument, what I am trying to say is that ATRAC3 still has to fit the MiniDisc standard mold.

So even on an old player, an ATRAC3 track *WILL PLAY*, though it will be mute, as opposed to simply being dysfunctional and causing errors or noise.

You do understand what my point is, right?
Title: ATRAC3 superior or inferior to mp3 ?
Post by: Pio2001 on 2004-07-13 11:03:40
Quote
These graphs are very important: they are proving either that encoders used for this test are totally different from encoders available for MD users via SonicStage or they are proving that Sony didn't respected the bitrate asked for this test.
[a href="index.php?act=findpost&pid=225414"][{POST_SNAPBACK}][/a]


Wow ! I would never have noticed this ! Are you sure that ATRAC3 lowpasses everything at 18 kHz ? Couldn't it be an apparent lowpass caused by the ATH, like the apparent 16 kHz lowpass of MP3 APS -Y ?
Title: ATRAC3 superior or inferior to mp3 ?
Post by: guruboolez on 2004-07-13 12:07:57
Could only talk about music encoding: but I never seen any sample, any bit, any information beyond 18...19 KHz. I seriously doubt that dedicated signal tones could force the same encoder to reach higher values.
Even with APS -Y, it's easy to find short parts going with maximum lowpass on regular music.
Title: ATRAC3 superior or inferior to mp3 ?
Post by: Latexxx on 2004-07-13 15:15:39
Quote
Could only talk about music encoding: but I never seen any sample, any bit, any information beyond 18...19 KHz. I seriously doubt that dedicated signal tones could force the same encoder to reach higher values.[a href="index.php?act=findpost&pid=225518"][{POST_SNAPBACK}][/a]

Have you checked any hardware implementations? We can't say that they are spreading disinformation before we have checked all the possibilities. 
Title: ATRAC3 superior or inferior to mp3 ?
Post by: guruboolez on 2004-07-13 16:02:00
No, I don't have any MD player to compare software and hardware encoding. Anyway, there are serious chances that the encoder tested by the laboratory is different from consumers encoders. The potential of ATRAC3 is certainly superior to what we heard during Roberto's 128 test: SonicStage is optimized for high-speed transcoding, and is ~4 time faster than lame --preset abr/cbr on my computer, 2.5 time faster than mpc, and even faster than wma. I suppose that a slower encoder could be better (slightly or much better, I can't say).
Title: ATRAC3 superior or inferior to mp3 ?
Post by: Latexxx on 2004-07-13 16:05:25
We should get some samples encoded using some hardware encoders which do it in real-time instead of 100x.
Title: ATRAC3 superior or inferior to mp3 ?
Post by: saratoga on 2004-07-13 16:41:21
Quote
Quote
Quote
It is quite right in the sense of a CODEC, but not in the sense I was talking about. I was addressing the fact that ATRAC3 had to be useable on a hardware platform 10 years old, i.e. MiniDisc

Wrong.
ATRAC3 is not usable on older minidisc players.
[a href="index.php?act=findpost&pid=225365"][{POST_SNAPBACK}][/a]


I apologize for my use of the word 'useable' in a different manner, as ATRAC3 is indeed not possible to LISTEN to on older hardware, but if you read my argument, what I am trying to say is that ATRAC3 still has to fit the MiniDisc standard mold.

So even on an old player, an ATRAC3 track *WILL PLAY*, though it will be mute, as opposed to simply being dysfunctional and causing errors or noise.

You do understand what my point is, right?
[a href="index.php?act=findpost&pid=225445"][{POST_SNAPBACK}][/a]


Thats nothing.  MP3s have to play on the computer "mold".  So while ATRAC may be 10 years old, this obviously makes MP3 over 50 years old 

Heck they also play on electronics and devices with batteries.  Maybe they're over a hundred years old!
Title: ATRAC3 superior or inferior to mp3 ?
Post by: Mystic_Image on 2004-07-14 01:07:26
Clearly, someone hasn't extrapolated what I thought they would from what I argued... doh.

I obviously haven't communicated my ideas about engineering and design restrictions well enough.

You still seem to be saying something about MP3's being comparable to MiniDisc, claiming that I don't believe it should be done, when I've repeatedly said that such a thing is fine... under the right circumstances.

Quite simply, the design goals/requirements were quite different from most CODECs, and that means you have to take into account these differences in order to make a comparison that is more meaningful (i.e. where only one variable counts).

----

I should be able to get some samples of hardware encoded ATRAC and ATRAC3. I recorded a show I helped out with recently, and I used both modes on my N910 (with ATRAC Type-S). The only problem is trying to get it off the MiniDisc digitally as I don't have a standalone player with digital out.

One other issue we haven't covered yet is that there is apparently a noticeable difference between MD manufacturers... Sharp vs. Sony to be particular. I think they actually use different hardware encoders too...
Title: ATRAC3 superior or inferior to mp3 ?
Post by: guruboolez on 2004-07-14 01:11:38
Quote
The only problem is trying to get it off the MiniDisc digitally as I don't have a standalone player with digital out.
[a href="index.php?act=findpost&pid=225725"][{POST_SNAPBACK}][/a]

It's not a problem here. An analogic capture won't change the lowpass value (background noise will be easy to distinguish from the real musical informations). It's the only thing we need to check (in order to see if the frequency response (FR) showed in the laboratory test is comparable to the FR of the hardware encoder).
Title: ATRAC3 superior or inferior to mp3 ?
Post by: Cygnus X1 on 2004-07-15 06:36:54
A friend of mine who is quite dull in terms of cognitive functioning (to put it kindly) sold me a basically-unused MZ-N1 portable MD recorder for a mere pittance because he couldn't quite figure out how to use it  Although I bought it with the goal of doing some live recording and documenting my diagnostic assessments for transcription (part of my PhD requirements), it dawned on me that with such a high-quality unit, I can record and upload samples of hardware-encoded ATRAC and ATRAC3 files. Since the MZ-N1 is considered by many in the MD community to be one of the "best sounding" units of all time, samples recorded on the unit would carry a bit more objective validity amongst those types (ok, maybe).

Please, I would appreciate fellow HA members to recommend a collection of maybe 20 "killer samples" that I will locate and encode to MD. The MD samples can be recorded using my standalone CD deck or my PowerBook's audio input (which is very quiet). I will then compress them with FLAC. This way, HA can have a permanent database of MD samples, both software and hardware encoded, that will help answer people's questions regarding the quality of ATRAC and ATRAC3 in an unbiased, objective fashion. In the future, I'd possibly even be open to conducting a listening test (probably with Roberto's help  ), so suggestions are welcome!
Title: ATRAC3 superior or inferior to mp3 ?
Post by: Mystic_Image on 2004-07-15 06:38:44
Quote
Quote
The only problem is trying to get it off the MiniDisc digitally as I don't have a standalone player with digital out.
[a href="index.php?act=findpost&pid=225725"][{POST_SNAPBACK}][/a]

It's not a problem here. An analogic capture won't change the lowpass value (background noise will be easy to distinguish from the real musical informations). It's the only thing we need to check (in order to see if the frequency response (FR) showed in the laboratory test is comparable to the FR of the hardware encoder).
[a href="index.php?act=findpost&pid=225729"][{POST_SNAPBACK}][/a]



Ok, I've done some preliminary tests using a Sony N910 portable, a Sony JE440 deck, and a Sharp IM-DR80 portable, all using digital-line in from a Denon DVD player (playing a standard CD) as a source and then played back and recorded into my computer at 16 bit, 48KHz on an M-Audio Revolution.

There's definitely something fishy. For all 3 devices, on standard ATRAC modes, the frequency response reaches 20 - 21KHz. However, when using ATRAC3, there's a cutoff around 18KHz.

Looking on Sony's sites, I can't find any MiniDisc device that has a frequency response rated above 20KHz, regardless of whether it is ES (high end) series or not... so something is most likely wrong with the report. The ATRAC3 132kbps frequency response in the report is identical (maybe even BETTER) than the wave file one, and seems to exceed 20KHz by far.
Title: ATRAC3 superior or inferior to mp3 ?
Post by: Cygnus X1 on 2004-07-15 06:44:45
Quote
Ok, I've done some preliminary tests using a Sony N910 portable, a Sony JE440 deck, and a Sharp IM-DR80 portable, all using digital-line in from a Denon DVD player (playing a standard CD) as a source and then played back and recorded into my computer at 16 bit, 48KHz on an M-Audio Revolution.

There's definitely something fishy. For all 3 devices, on standard ATRAC modes, the frequency response reaches 20 - 21KHz. However, when using ATRAC3, there's a cutoff around 18KHz.

Looking on Sony's sites, I can't find any MiniDisc device that has a frequency response rated above 20KHz, regardless of whether it is ES (high end) series or not... so something is most likely wrong with the report. The ATRAC3 132kbps frequency response in the report is identical (maybe even BETTER) than the wave file one, and seems to exceed 20KHz by far.
[a href="index.php?act=findpost&pid=226096"][{POST_SNAPBACK}][/a]


I think I may have a theory on this. I remember reading a review on the MD Community forum in which somebody bubled about the quality of LP4, claiming that it had a "razor-flat" frequency response to 20Khz when using a test tone. However, we all know that when recording music, the upper response is probably around 14Khz at best (17.5Khz for LP2). This leads me to believe that  ATRAC3 might use some sort of adaptive lowpass, which probably could be tweaked by Sony to look really nice for an easy sample. Does this make any sense?
Title: ATRAC3 superior or inferior to mp3 ?
Post by: ssamadhi97 on 2004-07-15 06:54:06
Quote
Please, I would appreciate fellow HA members to recommend a collection of maybe 20 "killer samples" that I will locate and encode to MD.
[a href="index.php?act=findpost&pid=226095"][{POST_SNAPBACK}][/a]

Try Castanets for a good laugh. I've heard some MD encoded sample once, it was really pre-echo-ridden. Wonder if that's the best MD can do.. at least it was audibly worse than lame --aps
Title: ATRAC3 superior or inferior to mp3 ?
Post by: Cygnus X1 on 2004-07-15 07:09:51
Quote
Quote
Please, I would appreciate fellow HA members to recommend a collection of maybe 20 "killer samples" that I will locate and encode to MD.
[a href="index.php?act=findpost&pid=226095"][{POST_SNAPBACK}][/a]

Try Castanets for a good laugh. I've heard some MD encoded sample once, it was really pre-echo-ridden. Wonder if that's the best MD can do.. at least it was audibly worse than lame --aps
[a href="index.php?act=findpost&pid=226104"][{POST_SNAPBACK}][/a]


Yes, that was one of my first ideas...I had tried this a few years back with a ATRAC Type-R MD deck and the results were quite poor. Samples with trumpets or creaking doors or tolling bells also threw it off, IIRC.
Title: ATRAC3 superior or inferior to mp3 ?
Post by: Mystic_Image on 2004-07-15 18:29:41
Quote
Quote
Ok, I've done some preliminary tests using a Sony N910 portable, a Sony JE440 deck, and a Sharp IM-DR80 portable, all using digital-line in from a Denon DVD player (playing a standard CD) as a source and then played back and recorded into my computer at 16 bit, 48KHz on an M-Audio Revolution.

There's definitely something fishy. For all 3 devices, on standard ATRAC modes, the frequency response reaches 20 - 21KHz. However, when using ATRAC3, there's a cutoff around 18KHz.

Looking on Sony's sites, I can't find any MiniDisc device that has a frequency response rated above 20KHz, regardless of whether it is ES (high end) series or not... so something is most likely wrong with the report. The ATRAC3 132kbps frequency response in the report is identical (maybe even BETTER) than the wave file one, and seems to exceed 20KHz by far.
[a href="index.php?act=findpost&pid=226096"][{POST_SNAPBACK}][/a]


I think I may have a theory on this. I remember reading a review on the MD Community forum in which somebody bubled about the quality of LP4, claiming that it had a "razor-flat" frequency response to 20Khz when using a test tone. However, we all know that when recording music, the upper response is probably around 14Khz at best (17.5Khz for LP2). This leads me to believe that  ATRAC3 might use some sort of adaptive lowpass, which probably could be tweaked by Sony to look really nice for an easy sample. Does this make any sense?
[a href="index.php?act=findpost&pid=226100"][{POST_SNAPBACK}][/a]


It does makes sense. It shouldn't be too hard to detect extreme or limit cases where a low-pass should be changed. Still, I wouldn't expect it to go up to 22KHz when regular ATRAC won't do that either.

I'm generally interested in seeing comparisons between LAME, Fraunhofer and ATRAC3 on the same samples.
Title: ATRAC3 superior or inferior to mp3 ?
Post by: MugFunky on 2004-07-15 19:07:33
Quote
Wow ! I would never have noticed this ! Are you sure that ATRAC3 lowpasses everything at 18 kHz ? Couldn't it be an apparent lowpass caused by the ATH, like the apparent 16 kHz lowpass of MP3 APS -Y ?


from what i read in the above links, an apparent lowpass would only be observed at 11025 Hz if it were the same kind of lowpass seen in mp3 (ie. occuring on a major band boundary).  such lowpasses are seen in ATRAC with speech, but i haven't noticed them with ATRAC3 in most cases.

however, it's quite possible that pure tones would have a higher resolution in ATRAC/ATRAC3 as there is no transient variations to speak of, and a pure tone is an ideal situation for something like ATRAC.  whether sony tweaked their encoder to deal separately with each piece of music could probably only be determined by getting hold of the original samples (preferably not decoded to PCM) and doing some serious comparisons with commercially available encoders.

hmmm... after reading that paper above, it's no wonder speech comes out all grotty in ATRAC - the bit allocation described there made me sick to my stomach.
Title: ATRAC3 superior or inferior to mp3 ?
Post by: Latexxx on 2004-07-15 21:26:24
Quote
hmmm... after reading that paper above, it's no wonder speech comes out all grotty in ATRAC - the bit allocation described there made me sick to my stomach.
[a href="index.php?act=findpost&pid=226317"][{POST_SNAPBACK}][/a]

Hmm, the paper (http://www.minidisc.org/aes_atrac.html), which dates back to 1992, says: "ATRAC does not specify a bit allocation algorithm; any appropriate algorithm may be used." They might have figured out something during these 12 years (or not  ).
Title: ATRAC3 superior or inferior to mp3 ?
Post by: MugFunky on 2004-07-16 16:13:13
hehe.  from the sound of some ATRAC i've heard, and from reading the article, it seems bit-allocation is the culprit... i suppose a simple (bad) algo means cheaper manufacturing costs.