10 years ago, it was pretty easy to hate MP3s. Even the software engineers who designed those early codecs would agree with that.
Over the last decade however, the quality of perceptual encoders has increased dramatically and today, most music fans have trouble telling even low-res 128kbps MP3s from full-resolution CDs.
I and some other die-hard audio nerds can still tell low-res 128kbps MP3s from high-res 320kbps MP3s just by hearing them. But we’re in a tiny minority of listeners, and even I admit that the differences aren’t huge. In fact, many self-described audiophiles routinely fail these tests.
Although I have yet to try a blind test between the current iTunes standard of 256kbps AAC and a full-resolution CD, I understand that several studies suggest no one has been able to reliably tell these types of files apart under controlled blind-listening conditions.
Because I know these things, and because it’s my job to listen to music critically, I was surprised recently when one of my musical heroes, Neil Young, made some pretty serious claims about audio quality, saying:
“We’re in the 21st century and we have the worst sound that we’ve ever had. It’s worse than a 78 [rpm record]. What happened?
“The MP3 only has 5 percent of the data present in the original recording… The convenience of the digital age has forced people to choose between quality and convenience, but they shouldn’t have to make that choice.”
“If you’re an artist and you created something and you knew the master was 100 percent great, but the consumer got 5 percent, would you be feeling good? “
Despite the fact that none of these statements are, you know, “actually true”, I absolutely love Neil Young and his music, and insist on giving him the benefit of the doubt.
These comments — coupled with his statement that “piracy is the new radio” — leads me to believe that there’s only one possible explanation for all this confusion:
You see, Neil Young has actually been living in 1973 for these past 39 years, and is only now coming to visit us here in the future.
Unfortunately, the time machine that he and a then-mobile Stephen Hawking built out of a rust-colored pickup truck blew a flux capacitor on the way here, stalling out in the year 2000 — an age when the frightful Napstasaurus still roamed the earth.
All this explains why Mr. Young has not yet heard of legal streaming services like Spotify, Rhapsody, YouTube, Google Music, Bandcamp, Soudcloud and “music blogs”, all of which actually are “the new radio”.
It also explains why he believes that MP3s are still encoded at 64kbps and have all the audio fidelity of someone shaking a box of Tic-Tacs in your ear.
Take Our Audio Test:
Sadly, I won’t be able to take this blind listening test myself, since I already know the answer. (Curses!)
We will however, be able to get a sense for whether large groups of people are able to correctly guess which file is a standard 256kbps AAC file created on iTunes, and which is the full-resolution 24-bit WAV master file.
If we’re to believe the software engineers who design these perceptual encoders, we should not expect our readers (who are some of the smartest, most attractive, and best-trained listeners on the planet) to tell the difference between these files.
If we’re to believe amazingly-talented singer/songwriter/time-traveler and living legend Neil Young, one of these files should sound like a shellac 78 of Robert Johnson singing Captain Beefheart’s greatest hits through an aluminum megaphone — and the other should sound something like properly-recorded music.
(Sorry, all images from that session were under copyright protection.)
About the Files
For this audio survey, we’ll be listening to two songs by two fantastic cult indie artists.
The first is a barely-processed track with minimal compression and extended dynamic range. The second is significantly more dense, and its production is more stylized.
You will find two versions of each song: A and B. One will be a full-resolution 24bit/44.1khz WAV file (an original master with even higher resolution than CD) and the other is a 256kbps AAC file (the iTunes standard).
Your job is to guess which is which.
The AAC version was up-sampled to WAV format so you wouldn’t be able to tell the two apart by file size or type. This process does not affect the sound quality further, or restore any lost resolution to the AAC.
Ready, Set, Go:
Clip 1 is a rough mix of a track called “Line ‘Em Up” by the roots-rock quintet, The Josh Dion Band.
(Coincidentally, the album version of this song was mixed by Neil Young engineer Niko Bolas.)
I tracked this album to sound great unmixed, with all the faders at zero, and when you listen to this clip, that’s exactly what you’ll hear.
There’s no EQ, and only minimal compression on the lead vocal, bass, and drum bus. There’s also a hint of reverb on the lead vocal and snare drum.
To help bring the track up to a reasonable reference level, a barely-there buss compressor and Massey limiter were set to reduce by about 1db at the loudest section.
(All of this is audio nerd speak for “these are good-quality raw tracks.”)
Listen here and then vote near the end of the post:
Clip 2 is from a reference mix of a track called “Kite of Love” by the electro-shoegaze band Soundpool. (I’m not sure if it’s the one we chose for the album or not.)
It was tracked by the artist at their humble home studio, and given to me to mix. This production is exactly the opposite of Song 1. Although the track is un-mastered, there are plenty of effects, compression and EQ.
Again, one version is AAC and one version is WAV:
Now vote below:
1) Many software engineers feel that the music industry will soon push them hard to adopt full-resolution audio. Some of them believe that this would be an outright waste of storage space and battery power — even with higher-resolution formats like the 192khz files championed by Neil Young.
You can read their arguments against high-definition home audio by clicking here.
2) Blind ABX tests are currently the best method we have for showing whether individual listeners can consistently tell two sound files apart on a conscious level.
Please note that the test you’ve just taken is not an “ABX” test, but a more simple blind “AB” test. That means it can not prove whether any one listener (you) can tell these files apart. Instead, it can only suggest whether broad groups tend to guess correctly.
With that said, I’m open to the argument that High-Definition audio might affect listeners on an unconscious level, like some advocates suggest. Can you think of some other types of studies that would help us test for that?
I have one in mind. Write in with your own proposal, and we may just share it in a follow-up next month.
Update 5/7/2012: The results are in! You can see them here: Results from our Audio Poll: Neil Young and High-Definition Sound