The "loudness war" and 24 bit/196 kHz audio
I just wrote something for school (go WPI!) and I figured people here might appreciate it as well. Some of this might seem obvious to the Verge audience, but it's written for the layman, so just bear with me. Also, criticism accepted and encouraged!
"Can You Hear That?":
Regarding the Limits of Human Hearing when Improving Audio Quality
Early next year, musician Neil Young will release a product called Pono, an audio player that plays sound files that are identical to the studio masters. In technical terms, that means 24 bit sound with a sample rate of 196 kHz (Flanary). Producer Dr. Dre has similar ambitions. But does that really make a difference? Research shows that Pono's level of quality may be beyond the limit of human hearing. However, Neil Young is correct in that there is a problem with the way current music is produced, but the resolution is not the issue. Instead, there is a "loudness war" going on amongst the audio technicians mastering records, and the result of this war is that the volumes of songs are being pushed to the maximum at the expense of quality. So, although 24 bit, 196 kHz audio cannot hurt, the real problem stems from the way the studio master is being mixed in the first place.
This entire issue of bit rate and compression started during the transition from analog to digital. Unlike digital, vinyl records theoretically have infinite resolution, as the smooth audio waves are transferred directly from the tape to the vinyl grooves. However, when the compact disc came out, audio technicians had to decide how to digitally represent these waves. All computers operate in bits, which are essentially tiny electrical switches with two states, either off (zero), or on (one). To express audio in bits, the compact disc specification was determined to have a bit depth of 16 bits and a sample rate of 44.1 kHz, meaning 16 bits describe the amplitude of the wave at a specific time, and that amplitude is being recorded 44,100 times per second. These numbers were chosen for scientific reasons. Research shows that 16 bit audio, with 2^16 (65,536) possible levels, is precise enough to produce accurate sound, and when most people are blindly presented with two audio files, one at 24 bit and one at 16, they are unable to distinguish between them. Also, file size was much more important in the late ‘70s and early ‘80s when personal computers were first beginning to blossom and storage was expensive and limited. Therefore, 16 bit was chosen as it produced files that were small enough to fit on a compact disc but precise enough to sound good. As for the sample rate, the frequency range of human hearing is 20 Hz to 20,000 Hz, and theoretically, to produce that whole sound range, one would only need a sample rate of twice the highest frequency possible. Thus, to produce the maximum 20 kHz noise, a sample rate of 40 kHz is necessary. For good measure, the compact disc specification went a bit above that with a sample rate of 44.1 kHz. 24 bit and 196 kHz audio goes far beyond the limits of human hearing, and therefore it is unclear whether a new, higher resolution audio format would make any audible difference.
However, what can make a difference is the way the audio is mastered. If you've ever made a playlist with old recordings mixed with new recordings, you might have noticed that the new recordings are much louder. This is a result of something audio enthusiasts have dubbed the "loudness war." The gist of it is that psychologically, humans tend to prefer sounds that are louder. This is the same reason why commercials are louder than the typical television program. To take advantage of this, especially on the radio, audio engineers mastering a record will pump up the volume on each track to maximize volume and therefore the appeal of the song. Each engineer is competing with the other to make their song louder, and that is where the term, "loudness war," originates. There is no inherent problem with being loud, except in an effort to one-up their competitors, engineers have been mastering audio louder than the specification can handle, which causes something called clipping. Clipping is the phenomenon of the amplitude of a wave being larger than can be recorded, so the peak of the wave is "clipped" off, distorting the sound. Clipping is rampant among modern pop singles, sometimes even produced intentionally, like in Kanye West's recent works. In addition to clipping, the loudness war has lowered the dynamic range of music. Dynamic range is the difference in amplitude between the loudest and quietest sounds in a piece of audio. Movies have high dynamic range, which is apparent when characters are talking quietly and then an extremely loud explosion happens. Historically, music was the same way. For hundreds of years, musicians have used dynamics artistically to great effect. More recently, however, music that would normally have dynamics is flattened in an attempt to maximize volume. This reduction in dynamic range, in conjunction with clipping, results in sounds that are flat and distorted.
So if the mastering is the issue, why is there such a fuss about the resolution? It all goes back to vinyl. Although there is no technical reason why vinyl records are superior to CDs, vinyl does have higher dynamic range and no clipping. However, the only reason the loudness war does not affect vinyl is actually due to its limitations. The reason there is no clipping in vinyl is because if there was, the rough grooves would pop the needle out of place, resulting in skips. This is the same reason that vinyl tends to have less bass. The more vigorous the low frequency grooves are, the more susceptible the record is to skips. Because of that, vinyl requires a different master: one with no clipping and higher dynamic range. That said, vinyl has other limitations, such as degradation over time, pops and skips due to dust and scratches, and inconsistency in speed. This leads to a strange conundrum. The faults that make vinyl inferior are the very reasons why they sound better. Technically, there is no reason why producers could not master their records with better dynamics or without clipping, but because of the market, there is no incentive to do so. Recently, there have been some counter-efforts, such as the new Nine Inch Nails album, Hesitation Marks (Seifert). The album was mastered twice: once for iTunes and CDs and another for "audiophiles." In some ways this is ideal, in that the album can be commercially friendly, but still have a high quality version for enthusiasts, but it also means fewer people will experience the higher quality master. On the other hand, Daft Punk released the popular and critically acclaimed Random Access Memories earlier this year, which featured a meticulously crafted, high quality master. The album was created using a mix of vintage, high fidelity equipment and state of the art digital equipment, with some songs having over 200 separate tracks, something that simply would not be possible with analog (Dombal). The result is a refreshingly clear, crisp record. These releases offer glimmers of hope in an otherwise flat, distorted musical landscape.
Unfortunately, there is not much the average person can do right now to encourage the music industry to move towards these higher quality masters. The music industry is a notoriously slow-moving, sales-driven industry, and unless there is a value proposition for producing higher quality masters, they will continue mastering for radio and low bitrate mp3s. That said, influential musicians such as Dr. Dre and Neil Young are bringing this issue to the public in a way that has not been done before. Even though their efforts are a little misguided, there is still a concerted effort to improve the situation. In addition to musicians, larger companies are beginning to support the higher resolution as well. LG's latest flagship smartphone, the G2, supports 24 bit audio, and Sony's latest range of Walkman personal media players support the format as well (Kastrenakes; Klug). Again, 24 bit, 196 kHz audio is great, but the added resolution is only worth it if the music is mastered well, which will likely become apparent when such products are released.
It is truly strange that despite all of the advancements in technology over the past few decades, recordings from the ‘70s are still far superior to most recordings released today. Yet, for the millions of people listening to pop music, fidelity is the least of their concerns. I see people every day who only get music by extracting it from Youtube or from one of the throngs of file hosting sites online. They could not care less about the sound, because even if they did have higher fidelity, they wound not be able to hear it through their popular iPhone earbuds. The fact of the matter is that sticklers like me are among the minority, and therefore we have a small say in how music labels operate. Therefore, all we can do is wait and try to speak with our wallets, rewarding those who produce high quality audio.
Dombal, Ryan. "Daft Punk: Cover Story Outtakes." Pitchfork. Pitchfork Media Inc., 15 May 2013. Web. 28 Sept. 2013.
Flanary, Patrick. "Neil Young Expands Pono Digital-to-Analog Music Service." Rolling Stone. Wenner Media, 27 Sept. 2012. Web. 28 Sept. 2013.
Kastrenakes, Jacob. "Spec Sheet: Sony's New Walkman Players Challenge the iPod with High-quality Audio." The Verge. Vox Media, 27 Sept. 2013. Web. 28 Sept. 2013.
Klug, Brian. "Hands On with the LG G2 - LG's Latest Flagship." AnandTech. AnandTech Inc., 7 Aug. 2013. Web. 28 Sept. 2013.
Monty. "24/192 Music Downloads." Xiph.org. Red Hat Inc., 1 Mar. 2012. Web. 28 Sept. 2013.
Seifert, Dan. "NIN enters the 'loudness war' with multiple versions of new album." The Verge. Vox Media, 28 Aug. 2013. Web. 28 Sept. 2013.