I have used this track for years on must ABX tests and it is the track that made my realise that 192kbps J/S Mp3 was not transparent and also every lossy format chokes at it on 128kbps CBR + ABR on a personal ABX test i did a few years ago.
I think this is a killer sample on Nero AAC becuase it is easy to ABX at q 0.5 and faild to ABX it on LAME 3.97 -V3 --vbr-new
ABX results:
foo_abx 1.3.1 report
foobar2000 v0.9.4.3
2007/08/15 21:45:13
File A: E:\Misc\Black Sabbath - Iron Man.mp4
File B: E:\Misc\Wave\Black Sabbath - Iron Man.wav
21:45:13 : Test started.
21:46:06 : 01/01 50.0%
21:46:48 : 02/02 25.0%
21:47:57 : 03/03 12.5%
21:49:30 : 04/04 6.3%
21:51:04 : Test finished.
----------
Total: 4/4 (6.3%)
Iron Man sample:
http://www.sendspace.com/file/6t1vb1
Full Version: Possible Lossy Killer Sample
Which CD did you get this song off of? Was it the original release of "We Sold Our Souls For Rock N Roll" or the remastered version? What about the original release of "Paranoid" or the remastered version? Did you get it off of another greatest hits album?
QUOTE (kornchild2002 @ Aug 16 2007, 00:02) 
Which CD did you get this song off of? Was it the original release of "We Sold Our Souls For Rock N Roll" or the remastered version? What about the original release of "Paranoid" or the remastered version? Did you get it off of another greatest hits album?
It is from the 2004 Sanctuary remaster of Paranoid.
6.3 % doesn't pass as valid abx. Not saying you can't abx but can people at least try for 3 % or better ?
CODE
Duration : 0:30.000 (1323000 samples)
Sample Rate : 44100 Hz
Channels : 2
Bitrate : 175 kbps
Codec : AAC
Codec Profile : AAC LC
Encoding : lossy
Tool : Nero AAC codec / Aug 6 2007
foo_abx 1.3.1 report
foobar2000 v0.9.4.2
2007/08/16 02:38:25
File A: C:\temp\iron_man.flac
File B: C:\temp\iron_man.mp4
02:38:25 : Test started.
02:38:44 : 01/01 50.0%
02:38:49 : 02/02 25.0%
02:38:54 : 03/03 12.5%
02:38:56 : 04/04 6.3%
02:38:59 : 05/05 3.1%
02:39:01 : 06/06 1.6%
02:39:04 : 07/07 0.8%
02:39:06 : 08/08 0.4%
02:39:09 : 09/09 0.2%
02:39:11 : 10/10 0.1%
02:39:13 : 11/11 0.0%
02:39:19 : 12/12 0.0%
02:39:21 : Test finished.
----------
Total: 12/12 (0.0%)
ABX position: from 11.1 seconds
Artefact: additional noise on the artificial "fatboy" voice. Easy to spot
QUOTE (shadowking @ Aug 16 2007, 01:28)
6.3 % doesn't pass as valid abx. Not saying you can't abx but can people at least try for 3 % or better ?
Yeah sorry about that, I just only did a quick and and simple ABX on this sample
Taken into account. Will be checked. Thank you for finding this one.
p.s. For future to all people that help us find killer samples: can you please extract shorter parts that produce artifacts. Long samples are problematic because it might happen that it sounds perfect in 90% of duration and that there is only short artifact. This makes tuning of encoder hard. Maybe 5-10s. Thank you.
p.s. For future to all people that help us find killer samples: can you please extract shorter parts that produce artifacts. Long samples are problematic because it might happen that it sounds perfect in 90% of duration and that there is only short artifact. This makes tuning of encoder hard. Maybe 5-10s. Thank you.
This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.