I call bullshit: MP3 encoder preferences

This column originally appeared on the now-defunct dbrmuz blog. Warning: These articles are written to be deliberately controversial and may contain satire.

There is often much bullshit to be found in audiophile circles because audio quality is inherently subjective. And of course, where differences in sound quality can be perceived, a market can be found. That’s capitalism folks. My master’s previous post compared a recording made with a good deck using Dolby C encoding and metal cassette to the original CD, I wanted to do something similar. Today we are focusing on the possible differences between MP3 encoders.

For those living under a rock, this is what we mean by MP3. Andreas Westfeld and Rainer Bohme from the Technical University of Dresden in Germany explain:

The invention of the ISO/MPEG 1 Audio Layer-3 (MP3) audio compression algorithm is probably one of the most remarkable and far-reaching develop- ments in the area of digital media processing. The MP3 format enables compression rates of about 1/10 of the size of uncompressed digital audio while degrading the audible quality only marginally. Together with the moderate complexity of the compression algorithm—software implementations of MP3 coders/decoders (codecs) with acceptable performance even on low budget home computers soon became available—the format simplified the interchange of music and resulted in worldwide popularity for its users and sleepless nights for the music industry.

The maniac tested a freeware MP3 encoder that was created in 1998 called Plugger. Those that populate the Hydrogen Audio forums speak highly of it… Here are some examples of their praise:

Plugger is a very old mp3 encoder. AFAIK it is based on the ISO code and sacrifices speed for quality. It suxx big time (sic).

Their benchmark for “poor quality” was Plugger. Just to put things in perspective. I had to download a copy for myself, just to hear… yup, it’s bad. Real bad.

However after trying [Plugger] @320kbps I’d say it doesn’t sound shockingly bad (for the average person of course).. well, except with the fatboy sample.

These three singles were used for the test.

The above singles were ripped to uncompressed wave format from the original CD albums using Audio Grabber. Peak level normalization was performed on the tracks because the older CDs (for example, the unremastered 1987 Billy Joel CD) are notoriously of low volume. Normalization is considered undesirable because it may increase distortion, but in this instance it was unavoidable. The tracks were then converted using Plugger to 160 kbps MP3 format. I recorded all 6 tracks (3 x MP3, 3 x wave) onto a blank audio CD using Nero. After a few moments it was ready and I threw the audio CD into into my Sony CDP-397 CD player to have a listen with headphones.

I didn’t encounter one nasty artifact. And being an analogue addict, I can pick my artificial sounds (artifacts) with ease, just like Kevin Rudd can pick a comeback and Malcolm Turnbull can practice humility. Oh…  Anyway, what what was I saying again? Like all MP3 to CD comparisons, obviously the wave files sounded richer, particularly the guitar solos. The bass was tighter.

Perhaps the songs selected were inappropriate, perhaps with classical music the flaws of the encoder become far more evident? However, the conclusion is simple. Based on these tracks, Plugger produces entirely satisfactory MP3 recordings at 160 kbps and does not deserve the criticism levelled at it.

The aforementioned academic paper (statistical analysis of MP3 encoders) can be found here.

Comments are closed.

%d bloggers like this: