Are MP3s really responsible for a general decline in the quality of digital audio?
Here we go again. Another editorial has appeared based on the (mistaken, IMO) premise that digital audio (and video) quality is declining.
I actually agree with much of the article - "In defense of low-quality audio, video," published recently on DSP DesignLine - in which the author makes the case for why "low-quality" audio and video content (i.e., "MP3s" and "YouTube videos") is a good thing. However, at least when it comes to audio, the term "low quality" is at best a generalization, and at worst, a misrepresentation of current portable audio formats.
It's true that no one would argue that poorly encoded 128-kbps (or lower) MP3s are "high quality," or would ever be mistaken for true CD-quality sound. But today there are fewer excuses than ever for anyone having to put up with such sub-optimal encodings.
Not only are encoding schemes and codecs continually improving - resulting in increasingly more "transparent" lossy encoded files compared to the original recording - but cheaper and ever increasing storage space availability is reducing the need for draconian compression of audio recordings in the first place. And, in fact, anyone downloading music from iTunes is already receiving files in a format (AAC) that is considered to be better quality than MP3 at most bit rates.
So why the perception that audio quality is getting worse? Well, it could be in part due to audiophile reviewers, such as John Atkinson at Stereophile, who in a recent article titled "MP3 vs AAC vs FLAC vs CD" dismissed lossy file formats out of hand with the statement that "MP3s and their lossy-compressed ilk do not offer sufficient audio quality for serious music listening."
This isn't consistent with the results of many blind ABX listening tests (such as those conducted at the audio coding forum Hydrogenaudio), which suggest that in fact MP3 bitrates as low as 175 kbps can be transparent (indistinguishable from the original recording) to almost everyone. If so, then wouldn't it be suitable for most "serious listening" purposes, which I assume refers to listening to music and not subtle encoding artifacts?
Now, to be fair, I'm not going to archive my CDs using such a lossy MP3 format, nor do I listen to it at home - I use the lossless FLAC format. Not because I necessarily think I might hear a difference over using an MP3 format, but because I don't need to.
And this just further refutes the suggestion made by the Stereophile article (and other recent audio-related articles) that MP3 formats "and their ilk" are - like an audio technology version of Gresham's law - somehow a threat to the existence of "high quality" digital audio.
Comments, questions or suggestions? Email me at email@example.com.