Hydrogenaudio Forums > Sample for aotuv3

Hydrogenaudio Forums > Hydrogenaudio Forum > Uploads

kotrtim

Nov 29 2004, 22:22

This sample is uploaded for aotuvb3.

The Cranberries - Zombie

.......
fisrt time musepack --standard fail so badly
I'm a person who can't ABX aotuvb2 -q 4 can get 10/10 for this sample encoded with muspeack

I've also tried WMA Pro 9.1 -q75 (~193kbps)
again, piece of cake although I Can't get 10/10, I still get 9/10

also tried Nero 2.6.2.2 AAC Normal, 10/10

aotuvb3 -q 4 , 8/10

iTunes 4.5 AAC 192kbps, 6/10 -----------> this is the 3rd last sample I test, exhausted......not sure

Vorbis 1.1 -q 5 193kbps, 4/10---------> 2nd last sample....exhausted, not sure

iTunes 4.5 224kbps, 5/10 ---------->last sample

can someone verify this......guruboolezz?

shadowking

Nov 30 2004, 03:38

You are talking about position 2-3.5 secs ???

Aotuv b3 Q4: pos 2-3.5secs : 8/8 - noticeable distorsions

MPC 1.15s Q5: pos 2-3.5 : 3/7 - nothing interesting

kotrtim

Nov 30 2004, 10:35

position 2-3.5 secs....yeah

mpc, have serious problem, the version i used is 1.14
mpc encode that particular part "longer" and "louder" than original......
it sounds like someone is spitting!.....I cannot describe any better..

AFAIK there is no quality tuning from 1.14 -> 1.15, just bug fixes

I'll try with 1.15s later

music_man_mpc

Nov 30 2004, 13:44

QUOTE(kotrtim @ Nov 30 2004, 08:35 AM)

AFAIK there is no quality tuning from 1.14 -> 1.15, just bug fixes

QUOTE(muespack.net description of 1.15s)

This is the latest StreamVersion7 encoder, it is needed to encode Wave files to MPC. It has less problems with highly tonal music compared to 1.14.

I think that you are thinking of 1.15r -> 1.15s not 1.14 ->1.15 where there certainly was quality tuning.

ProtectYaNeck36

Nov 30 2004, 14:30

but wasn't the quality tuning done between 1.14 and 1.15r in reference to artifacts that Frank himself couldn't even hear? don't quote me on this, however I believe I heard this somewhere around the HA community.

shadowking

Nov 30 2004, 17:16

QUOTE(kotrtim @ Nov 30 2004, 08:35 AM)

AFAIK there is no quality tuning from 1.14 -> 1.15, just bug fixes

I'll try with 1.15s later

I am using foobar abx, all tracks are replaygained. Vorbis has an artifact when the 'swoosh' drum hits . No volume differences for me, nothing grabs my attention with mpc on this track (just normal rock).. Any chance you used Q4 'radio' profile??

music_man_mpc

Nov 30 2004, 17:32

QUOTE(ProtectYaNeck36 @ Nov 30 2004, 12:30 PM)

I have no idea, but its different in any case, so its definately worth trying both.

shadowking

Nov 30 2004, 18:12

1.15 was made to improve samples like 'amnesia'

kotrtim

Dec 1 2004, 09:30

QUOTE

used Q4 'radio' profile??

no standard, maybe this flaw is only in version 1.14
maybe its solved in 1.15?, need to test 1.15 b4 making any conclution

honestly all the other part is transparent to me, only the "swoosh" part is not transparent, that's for 1.14

guruboolez

Dec 2 2004, 01:40

My results (ABC/HR, with mpc 1.14 --standard and aoTuV beta 3 -q 5,5):

CODE

ABC/HR Version 0.9b, 30 August 2002
Testname:

1R = D:\Temp\Cranberries___zombie_partial2.aoTuVb3 -q5.5.wav
2R = D:\Temp\Cranberries___zombie_partial2.mpc5.wav

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

---------------------------------------
1R File: D:\Temp\Cranberries___zombie_partial2.aoTuVb3 -q5.5.wav
1R Rating: 3.5
1R Comment: drums are noisy, coarse. Pre-echo is audible. Overall sound *seems* to be a bit confused (consequence of this initial problem?). Cymbals are also flat, shhhhhhhhhhhhh.... Typical of Vorbis.
---------------------------------------
ABX Results:
Original vs D:\Temp\Cranberries___zombie_partial2.aoTuVb3 -q5.5.wav
8 out of 8, pval = 0.004
Original vs D:\Temp\Cranberries___zombie_partial2.mpc5.wav
3 out of 8, pval = 0.855

MPC was transparent to me. Vorbis suffered from usual coarsness and probably other problems (see comments).

kotrtim

Dec 11 2004, 03:20

QUOTE

foo_abx v1.2 report
foobar2000 v0.8.3
2004/12/11 16:44:19

File A: file://F:\Cranberries___zombie_partial2.mpc
File B: file://F:\Cranberries___zombie_partial2.wav

16:44:20 : Test started.
16:45:06 : 01/01 50.0%
16:45:47 : 01/02 75.0%
16:46:07 : 02/03 50.0%
16:47:36 : 02/04 68.8%
16:48:41 : 03/05 50.0%
16:49:25 : 04/06 34.4%
16:50:20 : 05/07 22.7%
16:51:29 : 06/08 14.5%
16:51:51 : 06/09 25.4%
16:54:39 : 07/10 17.2%
16:56:30 : 08/11 11.3%
16:57:36 : 08/12 19.4%
16:58:09 : Test finished.

----------
Total: 8/12 (19.4%)

===================================================
foo_abx v1.2 report
foobar2000 v0.8.3
2004/12/11 16:59:09

File A: file://F:\Cranberries___zombie_partial2.wav
File B: file://F:\Cranberries___zombie_partial2.ogg

16:59:10 : Test started.
17:00:32 : 01/01 50.0%
17:02:31 : 01/02 75.0%
17:03:01 : 02/03 50.0%
17:03:26 : 03/04 31.3%
17:04:01 : 04/05 18.8%
17:05:31 : 05/06 10.9%
17:07:01 : 06/07 6.3%
17:10:16 : 06/08 14.5%
17:10:21 : Test finished.

----------
Total: 6/8 (14.5%)

8/12 (19.4%) for mpc 1.15s
6/8 (14.5%) for aoTuVb3 -q 5.68

I thought i was imagining thing! This time is harder, i cannot get 100% for mpc...It seem there is quality tunings from 1.14 to 1.15s

there is something very wrong with 2-3 sec, Vorbis and mpc seem to encode it differently from original
All the other parts is transparent to me, i tried hard but i still can't notice any cymbals flatness that guruboolez mentioned

QuantumKnot

Dec 11 2004, 06:44

It would probably be helpful if you posted your mpc encoded file (the one you could abx 100%) so that others can try it. That would eliminate some possibilities

kotrtim

Dec 11 2004, 07:39

Good Idea...Uploaded

kotrtim

Dec 11 2004, 07:48

Good Idea...Uploaded

Cranberries___zombie.mpc is the one i get 10/10
its encoded with version 1.14

Cranberries___zombie_partial2.mpc is encoded wiith 1.15s and i can get 8/12

This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.