MP3 world shlock atrocity!

Posted by: Don Braid on 15 October 2000

I finally heard an MP3 audio clip played through a Naim system. A friend has his portable plugged directly into his Nait 2. We compared the CD version of Mark Knopfler's Golden Heart with the downloaded clip - both track 3, the Golden Heart song.
The MP3 was the most appallingly distorted, ear-threatening crap I've every heard from a good stereo system. The bass was so distorted it made the amp sound like it was on the verge of clipping (although all we could bear was low volumn) and Knopler's voice was so muffled and distant that I thought he had a sock in his mouth. Then we played the CD track and everything came back to life.
This MP3 garbage is the worst garbage in the history of audio. It makes 8-track tape sound like HiFi. Trash an MP3 clip today!

Don

Posted on: 15 October 2000 by Don Braid

Jonathan: You said

quote:
...END RANT.

Please don't stop ranting. If we don't oppose this atrocity, who will know? The world will slip into darkness. We're the audio monks illuminating manuscripts agains the new age of barbarism (olay, we're geeks but so were the monks, no?)
TRASH AN MP3 CLIP TODAY!

Don

Posted on: 15 October 2000 by David Dever

Don-

Though I'd never defend MP3 as an audio file format, most MP3 files are of poor quality due to poor attention to detail at the encoding stage.

128 kbps is the smallest bit rate which does not require sample rate down-conversion from 44.1 kHz. Oddly enough, there is not much improvement to be gained by using a higher constant bit rate or variable bit rates...

The choice of codec used to originally encode the files makes a huge difference on ultimate sound quality (this information is buried in the MP3 file header). IMHO, the Fraunhofer codec is really the ONLY option, as IIS had a significant role in the original formulation of the MPEG-1 layer 3 specification--thus you'd expect their algorithms to be the best.

(Institut Integrierte Schaltungen, http://www.iis.fhg.de/audio): thanks to Infopop for automatically converting this to a link...

I use the following settings:
constant 128 kBit/s bitrate
intensity stereo coding / allow downmix
high quality mode (as opposed to fast)

Best results are obtained if you allow as long as possible for the encoder to analyze the file. I've seen situations where a six-minute file took over 32 minutes to encode in "high-quality" mode, rather than 3 minutes in "faster" mode, with a huge variation between the two in sound quality, even at the same bitrate!

All this said, the only real way you can compare MP3's performance as a file format to the original is to burn a CD-R with both the original file (Sound Designer II / AIFF / WAV) and a file back-converted from MP3 to SDII / AIFF / WAV from within the same application that you encoded it.

Hope this makes sense,
Dave Dever, NANA

Posted on: 15 October 2000 by Don Braid

Hey Dave, how dare you muddy a good luddite campaign with facts!
Thanks for the information. Fascinating. For now, though, I still cry:
SAVE THE WORLD - TRASH AN MP3 FILE TODAY!

DON

Posted on: 15 October 2000 by richard goldsmith

Oh well, confession time, I actually own a portable Sony MD to use "on the move", and it's actually much better than I'd expected. I think it's the top portable recording model (MZ-R90), and no, comparing it to the source on playback through a real system shows up its limitations. But seriously, what would one expect, a Linn? However, it has to be better than the MP3 I've heard.

However, it's a great way to dump all my favourite LP's onto a tiny format and listen while walking to work, surface noise and all!