So I was talking about the wonderful event that I experienced in NYC last Wednesday evening sponsored by the NARAS Producers and Engineers wing and the Audio Board of the CEA (I’m a member of both organizations). Andrew Scheps, a prominent engineer, made a very strong case for better quality (at least between MP3 and CD quality) sound by demonstrating examples of the same track in different encodings played back one after the other.
He first played portions of “Strawberry Fields Forever” from Sgt. Peppers from YouTube and then as a 44.1 kHz/24-bits PCM file. It was immediately apparent that the PCM file was better. It’s true that the YouTube version as encoded at 128 kbps (vs. a CD quality files which requires 10 times more data bandwidth). He played other examples including a section from a Mahler symphony and a tune by Esperanza Spalding, who won the “Best New Artist” Grammy several years ago. Each time the difference between the heavily compressed ogg vorbis, MP3 or AAC files and the full CD resolution files was readily apparent.
So it seems reasonable that there is a noticeable difference between the fidelity of lossy compressed music files and a file at CD quality. The Harmon Kardon survey established that kids can tell the difference and the demonstration put on by Andrew reinforced that finding for a mixed crowd of press and audio engineers. That means it’s settled, right? It is as far as that comparison is concerned. What I’m interested in goes well beyond the MP3 vs. CD debate. There are lots of questions that I want answered in a new survey…one that I may pitch to the CEA’s Audio Board for funding.
Let’s explore an entirely new set of questions that focus on the state-of-the-art developments of the past 10-15 years. Let’s not focus on the fidelity of tracks that were recorded back in the 60s. Wouldn’t you like to know if there is a statistical preference for real HD-Audio over CD quality? That seems like a pretty realistic place to start a new inquiry. But it can’t be the kind of test that the Boston Audio Society trumpeted as the definitive study that established that their members couldn’t tell the difference between so-called HD resolution music releases and a re-sampled version at CD resolution (I wrote on that sham in a previous post, you can check it out here: The take away from the article was that its impossible to test for an attribute when none of the samples tested contain what you’re looking for!
The new survey would have to be very rigorous and fully double blind A | B | X in nature. Yes, it’s true that I have a horse in the race so my own personal beliefs and preferences could play any role in the outcome. However, I believe that any new survey would have to be structured in a new an open-minded approach. It won’t work if the traditional short examples are played and participants are simply asked to identify one sample over the other. It might involve some extended listening and then an exit survey that asks how you feel after an hour or more of CD resolution vs. a session at HD-Audio specifications.
I’ve just gotten started thinking about the questions and the wording of those questions that should be included in the new survey. Here’s a short list of some of the questions that I would like to get answers for:
1. Which is perceived as more musically accurate, natural and pleasing:
A. A commercially mastered recording
B. An unmastered/unprocessed selection of music.
2. Which format is preferable:
A. A data compressed (mp3, AAC etc) version of a track
B. A standard definition track
C. A high definition track.
3. Which mixing perspective is preferable:
A. A monophonic mix
B. A stereo mix
C. A 5.1 surround mix
4. Which type of surround mix is preferable:
A. A “stage” 5.1 surround perspective
B. An “audience” 5.1 surround perspective
5. Which headphone experience is preferable:
A. A traditional “inside your head” experience
B. A “room realized” virtual surround experience
6. Does the knowledge of a format’s frequency response and dynamic range make any difference in what you enjoy and purchase?
7. If you prefer the sound of one recording to another, does that mean it has higher fidelity specifications?
8. What is more important, the sound experience or the level of the performance? Why?
9. Would be prefer to be able to stream 10 low fidelity “playlists” or 1 high fidelity “playlist”?
10. What role do the producer and engineer have in the ultimate sound quality of the released track or album?
So these are just a few ideas. I’m open to any and all suggestions. I think it’s time to get some reality back into the whole audio recording and reproduction arena. It’s been drowning in its own techno-babble and touchy-feely cesspool for too long.
The “Lost In Translation” presentation was a very good warm up to the really important questions that loom large in the digital music marketplace but it really just opened up an entirely new set of questions.