Short test: Vangelis - Alpha, Original vs. ND vs. LAME vs. AoTuV Options

[Sebastian Mares]

Here is a small test with the sample I just posted called "Alpha" (by Vangelis).

Nero Digital 3.1.0.2, new LC encoder, Streaming, ABR:

CODE

foo_abx 1.3 report
foobar2000 v0.9 beta 13
2005/12/28 11:41:17

File A: D:\Vangelis - Alpha.mp4
File B: D:\Vangelis - Alpha.wv

11:41:17 : Test started.
11:41:56 : 01/01 50.0%
11:42:01 : 02/02 25.0%
11:42:22 : 03/03 12.5%
11:42:27 : 04/04 6.3%
11:42:33 : 05/05 3.1%
11:42:44 : 06/06 1.6%
11:42:49 : 07/07 0.8%
11:43:03 : 08/08 0.4%
11:43:06 : Test finished.

----------
Total: 8/8 (0.4%)

Some pre-echo, but not very disturbing - I would give it a 4.

LAME 3.97b2, -V5 --vbrnew:

CODE

foo_abx 1.3 report
foobar2000 v0.9 beta 13
2005/12/28 11:47:13

File A: D:\Vangelis - Alpha.wv
File B: D:\Vangelis - Alpha.mp3

11:47:13 : Test started.
11:48:13 : 01/01 50.0%
11:48:15 : 02/02 25.0%
11:48:18 : 03/03 12.5%
11:48:20 : 04/04 6.3%
11:48:22 : 05/05 3.1%
11:48:24 : 06/06 1.6%
11:48:26 : 07/07 0.8%
11:48:28 : 08/08 0.4%
11:48:29 : Test finished.

----------
Total: 8/8 (0.4%)

Pre-echo, but performs slightly better than Nero IMO. I would give it a 4.3.

AoTuV 4.51, -q 4:

CODE

foo_abx 1.3 report
foobar2000 v0.9 beta 13
2005/12/28 11:48:52

File A: D:\Vangelis - Alpha.wv
File B: D:\Vangelis - Alpha.ogg

11:48:52 : Test started.
11:49:25 : 01/01 50.0%
11:49:37 : 02/02 25.0%
11:49:41 : 03/03 12.5%
11:49:51 : 04/04 6.3%
11:50:07 : 05/05 3.1%
11:50:10 : 06/06 1.6%
11:50:16 : 07/07 0.8%
11:50:23 : 08/08 0.4%
11:50:29 : Test finished.

----------
Total: 8/8 (0.4%)

Pretty bad - smeared sound, HF boost. Score: 3

This post has been edited by Sebastian Mares: Dec 28 2005, 12:01

[Undesirable]

Interesting that AoTuV seemed to fare so badly.

What I'm really wondering, though, is why your "Multiformat 128 kbps Listening Test" thread had over 44,000 hits, yet nobody at all has replied here, and your previous thread was closed without explanation.

[sh1leshk4]

The previous thread of his was open up until the listening test is considered done, which is why it is closed (and it's by his own request IIRC), and a new result thread is opened.
This thread is for that particular sample (vangelis - alpha), in which shows that AoTuV doesn't work quite well.

[Sebastian Mares]

The previous thread of his was open up until the listening test is considered done, which is why it is closed (and it's by his own request IIRC), and a new result thread is opened.
This thread is for that particular sample (vangelis - alpha), in which shows that AoTuV doesn't work quite well.

Come on, don't blame the moderators or administratos for everything; it was me who asked Garf to close the thread.

[sh1leshk4]

Come on, don't blame the moderators or administratos for everything; it was me who asked Garf to close the thread.

I didn't blame anyone!

... (and it's by his own request IIRC), ...

That 'his' refers to you, Sebastian...

[Sebastian Mares]

OK

"was closed without explanation" sounded a bit suspicious so I didn't want people to think that Garf closed my thread because he's evil. Also, I wanted to clarify that it was indeed me who wanted the thread to be closed (since you wrote "IIRC").

Anyways, what do you think about including this sample in the 48 kbps listening test? It's pretty difficult for every encoder (unlike some other killer samples that are aimed at a single encoder).

[sh1leshk4]

It's a pretty nice sample.
Despite the $20 Philips SHE 9500 canalphones, I could pretty much ABX AoTuV; LAME was harder for me.
Didn't try Nero though; I only have Nero 6 installed here.

Will be very interesting to include this one for the 48kbps listening test.