Autumn 2006 Listening Test: http://www.hydrogenaudio.org/forums/index....showtopic=47313

Click to view attachment
l3enc 1.0 from raw PCM:
Click to view attachment
l3enc 1.0 from wave:
Click to view attachment

- l3enc encoded the "PCM version" correctly.
- the "wave version" demonstrates this problem: http://www.hydrogenaudio.org/forums/index....st&p=420778

Edit

Removed the samples to make room.

ha! now my killer pre-echo sample comes! here (preecho2.flac)

and the old one (also for preecho, focus on fourth "clap" or snare or whatever): here (scooter_fixed.flac)

both are by no means transparent even at -V 2

J.M.

Maybe it's not important what samples to use. There are hundred of samples already from previous tests.
More important thing will be choose samples of various music styles, speech with possible background music, artificial sounds. But there are already hundred of this kind of samples.

I didn't try to propose my sample to be used as a test sample.

I just wanted to show that the other contenders except LAME are quite bad at 128 kbps even with this kind of simple old fashioned hard rock track.

Hopefully someone actually tries this sample with some of the discussed MP3 encoders.

- The first about three seconds are quite easy to ABX
- Use a relatively high volume level, the original CD is old and not very loud or compressed.

MrQuestionMan reports:

- RealPlayer 10.5 MP3 @ VBR 128 kbps
- GoGo 3.13 MP3 @ -a -b 135 -q 2

Edit

Removed the attachment to make room.

Thanks for the results Alex.

Reuploaded the AC/DC sample.

The related forum post:
http://www.hydrogenaudio.org/forums/index....st&p=514962

My encoding speed & bitrate test data for the 128 kbps MP3 test.

Edit

the related forum post: http://www.hydrogenaudio.org/forums/index....st&p=515303