Reading Michel Houellebecq’s latest novel Submission, I was struck by one of the statements he made and how it pertains to the future of audio. The book’s fictional protagonist, speaking about an author he has studied, states, “His masterpiece was a dead end–but isn’t that true of any masterpiece?”
High-end audio is a masterpiece. Traditional audio engineering has been perfected, at least in the sense that it’s been pushed about as far as it can go. Sure, new amps and DACs might sound slightly better than the ones we have now. The resolution of digital files can be raised to even higher levels. Improvements in mass production will bump up speaker quality slightly. Still, traditional analog and basic digital audio engineering are pretty much at a dead end. No improvement in amplifier, DAC, or passive speaker design is likely to result in a significant improvement in sound quality.
The good news is, we WILL see significant improvements in audio reproduction in 2016 and over the next few years. As I was walking around the show floor and attending presentations at the Audio Engineering Show in New York City in October, it was obvious to me that digital signal processing, or DSP, presents numerous possibilities for better sound in any system…and for better sound from smaller and less expensive products, too.
DSP is built into many Class D amplifier chips now, and it’s also available in easy-to-program modules, such as those from Danville Signal Processing. As high-end audio companies like Bowers & Wilkins, Dynaudio, MartinLogan, and others have begun to build active products–i.e., wireless speakers, soundbars, and subwoofers–they’ve been using DSP more and more. Many audiophiles, perhaps scarred by memories of terrible-sounding DSP modes in cheap AV receivers, react negatively to any mention of DSP. My suspicion, though, is that DSP will find its way from these lower-end products into more elite, higher-end products because the benefits of DSP are too powerful to ignore.
We often think of high-end manufacturers spending whatever amount of time it takes to fine-tune their products, but the reality is that development time is always a limited resource for any company, and no product is ever perfect. There’s always a time when the engineers have to say, “That’s good enough.” DSP allows engineers, within the development time they have, to experiment with many more possibilities in product tuning.
In traditional analog audio design, an engineer fine-tunes a product by physically changing one or more parts, such as a resistor or capacitor. With DSP, the engineer fine-tunes performance using a control interface running on a computer. I’ve included a screen shot (below) of the parametric EQ interface from a Quickfilter Technologies QF3DFX DSP to give you an idea. For any filter, the engineer specifies the center frequency, the Q (bandwidth), the amount of boost or cut, and the type of filter (high-pass, bandpass, low-pass, etc.). Any change takes just seconds. The engineer has time to experiment more, and to tweak a product to a higher level of performance than could be achieved in the analog domain.
DSP also allows a level of precision that analog circuitry cannot achieve affordably and practically. Using DSP, an engineer can tune speaker crossover filters to within fractions of a decibel; with analog circuitry, crossovers are typically designed in increments of 6 dB, so the engineer is limited to, say, a -12dB high-frequency roll-off on a woofer where a -14.5 dB roll-off is what’s actually best.
Filter frequencies can be specified down to fractions of a hertz with DSP. With analog, such precision is practically impossible because the capacitors and inductors used in analog circuits are typically manufactured to tolerances of 5 or 10 percent. On, for example, a high-pass filter for a midrange driver in a speaker, even a five percent tolerance in a capacitor would result in a range of error of roughly -25 to +30 Hz.
Notice that the QF3DFX interface offers 10 filter bands per channel. This lets the engineer tune out minor speaker driver and cabinet resonances and response flaws without increasing parts cost or circuit complexity. Doing this with analog filters would take longer, increase the parts cost considerably, and possibly impact sound quality.
This is just scratching the surface of DSP’s potential because I’m not even getting into the QF3DFX’s other capabilities. And big DSP chips from companies such as Analog Devices and Texas Instruments can do far more than the relatively low-cost QF3DFX can.
Of course, audiophiles may be concerned that DSP requires analog signals to be converted to digital, but the extremely subtle effects of converting an analog signal to digital and back again are several orders of magnitude less significant than the improvements in performance that DSP offers.
Bottom line: Speakers work better with DSP.
A hint of what DSP can do was visible at the AES show, where the Barefoot Audio booth attracted some of the biggest crowds. Not only does the company use DSP to tune its recording monitors (shown above) to near perfection–and to get far more bass output than their small cabinets would suggest–but it also uses DSP to create its MEME (Multi-Emphasis Monitor Emulation) technology. With the twist of a switch, MEME allows Barefoot monitors to mimic the sound of Yamaha’s legendary (and no longer manufactured) NS-10M monitor, the classic Auratone cube-shaped recording monitors, and a typical consumer hi-fi system.
Audiophiles might not want a switch on their speakers to emulate different sounds, but they might want one that fine-tunes the speaker to different acoustical environments…or provides some gentle, non-invasive tonal balance control. The AES show proved that DSP is becoming more powerful yet easier to use. It will be exciting to hear what audio product designers accomplish with it in 2016.
• Six AV Trends We’re Thankful for at HomeTheaterReview.com.
• CEDIA 2015 Show Report and Photo Slideshow at HomeTheaterReview.com.
• How to Choose a Subwoofer for Surround Sound or Stereo at HomeTheaterReview.com.