IPB

Welcome Guest ( Log In | Register )

 
Reply to this topicStart new topic
Atrac3plus vs others at 64 kbps
Latexxx
post Aug 2 2004, 23:25
Post #1


A/V Moderator


Group: Members
Posts: 858
Joined: 12-May 03
From: Finland
Member No.: 6557



I did a quick test. I compared Sony's Atrac3plus to other formats at 64 kbps. Only one sample used. All files were normalized before listening.

The original sample used:
http://www.kolumbus.fi/hahne/atrac3plus/sample.flac
-Genre: Heavy Metal
-Artist: Nightwish
-Track: Nemo

Used encoders:

Lame 3.97 a 3 --preset 128 as anchor
Lame 3.97 a 3 --preset 64
Sonic Stage 2.1 for Atrac3plus 64 kbps (+ Total Recorder for wave conversion)
WMA std 9 (dBpowerAMP) 64 kbps
Vorbis q0 (oggdropXPd 1.7.11 - libvorbis 1.1-RC1 compiled Jul 8 2006)

The results:

CODE
ABC/HR Version 1.0, 6 May 2004
Testname: atrac3plus vs others

1R = lame128.wav
2R = vorbisq0.wav
3L = wma64.wav
4R = lame64.wav
5R = a3p64.wav

---------------------------------------
General Comments:

---------------------------------------
1R File: lame128.wav
1R Rating: 4.0
1R Comment: perceptible, but not annoying
---------------------------------------
2R File: vorbisq0.wav
2R Rating: 2.0
2R Comment: weird stereo problems
---------------------------------------
3L File: wma64.wav
3L Rating: 2.2
3L Comment: metallic sound
---------------------------------------
4R File: lame64.wav
4R Rating: 1.0
4R Comment:
---------------------------------------
5R File: a3p64.wav
5R Rating: 3.0
5R Comment:
---------------------------------------


Conclusion:
Atrac3plus is surprisingly good with this sample @ 62 kbps. Don't know about other samples though.

Edit: ABX results removed because they might make you misinterprete my main idea.
Go to the top of the page
+Quote Post
ff123
post Aug 2 2004, 23:37
Post #2


ABC/HR developer, ff123.net admin


Group: Developer (Donating)
Posts: 1396
Joined: 24-September 01
Member No.: 12



QUOTE (Latexxx @ Aug 2 2004, 02:25 PM)
I did a quick test. I compared Sony's Atrac3plus to other formats at 64 kbps. Only one sample used. All files were normalized before listening.

The original sample used:
http://www.kolumbus.fi/hahne/atrac3plus/sample.flac
-Genre: Heavy Metal
-Artist: Nightwish
-Track: Nemo


If plain Atrac3 is any indication, Atrac3+ should have its biggest problems with music containing lots of transients.

ff123
Go to the top of the page
+Quote Post
Latexxx
post Aug 2 2004, 23:44
Post #3


A/V Moderator


Group: Members
Posts: 858
Joined: 12-May 03
From: Finland
Member No.: 6557



QUOTE (ff123 @ Aug 3 2004, 12:37 AM)
If plain Atrac3 is any indication, Atrac3+ should have its biggest problems with music containing lots of transients.

ff123
*

I just took the first sample I found. smile.gif You are free to do your own tests.
Go to the top of the page
+Quote Post
ff123
post Aug 3 2004, 00:20
Post #4


ABC/HR developer, ff123.net admin


Group: Developer (Donating)
Posts: 1396
Joined: 24-September 01
Member No.: 12



QUOTE (Latexxx @ Aug 2 2004, 02:44 PM)
QUOTE (ff123 @ Aug 3 2004, 12:37 AM)
If plain Atrac3 is any indication, Atrac3+ should have its biggest problems with music containing lots of transients.

ff123
*

I just took the first sample I found. smile.gif You are free to do your own tests.
*



I did some at 48 kbps (with Roberto's help) and posted the results to Usenet here:

http://groups.google.com/groups?hl=en&lr=&....com%26rnum%3D1

However, this was just a test to show somebody that 48 kbps atrac3+ cannot equal the quality of mp3 at 128 kbps, so there isn't a good comparison with other formats at similar bitrates. The sample "trust" seemed to be a pretty difficult one.

ff123

This post has been edited by ff123: Aug 3 2004, 00:22
Go to the top of the page
+Quote Post
guruboolez
post Aug 3 2004, 00:33
Post #5





Group: Members (Donating)
Posts: 3474
Joined: 7-November 01
From: Strasbourg (France)
Member No.: 420



I've tested atrac3+ at 64 kbps with ~20 samples when sonicstage 1.5 was released, ~18 months ago: it's maybe more pleasant than old atrac3 at the same bitrate (higher lowpass, and maybe less artifacts), but it also suffers on transients, as ff123 pointed it out:
http://membres.lycos.fr/guruboolez/AUDIO/s...grincements.zip (copy and paste the link in the adress bar).

I've also tested atrac3+ at 64 kbps on some samples when SonicStage 2.0 was released, and it's really far from a lame --preset CBR 128 encoding in my opinion. But these tests were not far from the MP3 Roberto's test, including a lot of poor mp3 encoders. I'm not shoked by the idea of similarities between atrac3+ at 64 kbps and fastenc encoding at 128 kbps. I think it's possible. But difference is huge with a better MP3 encoder (ProducerPro or lame).

After some days, SonicStage stopped to work, even with uninstall/reinstall. I couldn't investigate further, and can't contribute anymore to listening tests
Go to the top of the page
+Quote Post
Cygnus X1
post Aug 3 2004, 01:40
Post #6





Group: Members (Donating)
Posts: 676
Joined: 5-June 02
From: New York
Member No.: 2224



Before I switched from Win XP to OS X last year, I did some informal tests with ATRAC3plus and came to the conclusion that I personally preferred the duller, squashed stereo image sound of ATRAC3 at 66kbps compared to the increased HF content (and increased artifacts) of ATRAC3plus at 64kbps. I guess it's a matter of personal taste at these low bitrates....obviously, no encoder will be transparent at 64kbps or lower, so it's a matter of what sounds "best." Having grown up with tapes, I prefer a more severe lowpass over more HF content and digital artifacts.

Interestingly, Sony seems to make the case that ATRAC3plus @ 64kbps=132kbps ATRAC3=292kbps ATRAC. (See the "hi-md FAQ at minidisc.org for a chart). Not sure if I buy that one laugh.gif
Go to the top of the page
+Quote Post
Cygnus X1
post Aug 3 2004, 02:30
Post #7





Group: Members (Donating)
Posts: 676
Joined: 5-June 02
From: New York
Member No.: 2224



Ok, I found that chart that I was talking about:



Now come on, people, if ATRAC3 @ 132kbps is worse than LAME -V5 @ 128kbps, and ATRAC Type-R at 292kbps s probably comparable to MP3 at 192kbps, how can they claim that 64kbps is equal? It's not even a logical statement. Now, if they are comparing ATRAC's performance in 1992 to ATRAC3 in 2000, there might be some merit there, but it is very misleading, nonetheless. It can be taken as "ATRAC=ATRAC3=ATRAC3plus" if you don't take codec changes (over time) into account. Let's not even get into the 256kbps ATRAC3plus files being compared to "CD Quality"....I already found one sample where it was easily ABX'able from not only the original, but the older ATRAC Type-R.

Gotta love marketing....the sad part is, there are actually people who might believe it.

This post has been edited by Cygnus X1: Aug 3 2004, 02:39
Go to the top of the page
+Quote Post
Mystic_Image
post Aug 22 2004, 04:57
Post #8





Group: Members
Posts: 9
Joined: 9-July 04
Member No.: 15193



Yeah... it's pretty clear from the chart they're talking about the original 1992 codec. As I recall, Sony claimed a doubling in coding efficiency when they moved to ATRAC3, but 132 * 2 does not = 292... so as such I wouldn't expect it to be a linear relationship as they seem to describe and ATRAC3 should still be lower in quality... I don't see how 64kbps ATRAC3Plus could at all be compared to 132 ATRAC3 though.

But really, again, I'd have to say that Sony's intent is really to compare with the constant 128kbps FH CODEC that about 99% of people listen to. No one really pays attention to anything beyond numbers.

Marketing does indeed suck... but the average consumer isn't going to particularly care either.
Go to the top of the page
+Quote Post

Reply to this topicStart new topic
1 User(s) are reading this topic (1 Guests and 0 Anonymous Users)
0 Members:

 



RSS Lo-Fi Version Time is now: 17th April 2014 - 14:16