It’s been a busy few weeks for me lately, but I managed to sneak a few hours away from work to attend a local presentation by CHORD Electronics’ head honcho John Franks and lead designer Rob Watts.
By now, most people fixated on the high-end of portable audio have heard of the Hugo. When it first unveiled early this year, the Hugo was touted as a technological revolution in the world of digital signal processing. By using Watts’ own bespoke circuit designs, shoehorned into six different Xilinx Spartan-6 FPGAs, CHORD Electronics claimed phenomenal benchmarks in both measurable performance and subjective perception. Never before had there been a DAC, at only about the size of a portable hard drive, that was able to resolve over 120 dB DNR and drive headphones at <0.001% THD+N. It blew peoples’ expectations away.
Thus, when I learned that both John Franks and Rob Watts were to appear in my neck of the woods at a press junket held by local Taiwanese distributor ACE Audio, I knew I had to attend.
The talk was mostly straightforward and nothing most people interested in the Hugo didn’t already know. Whereas John Franks is an aerospace engineer, Rob Watts is the one well-versed in semiconductor design, and thus it was Watts that led most of the talk.
The Hugo revolves around its FPGA cores, carrying out sampling interpolation with a very long-tap WTA filter, before passing onto the fifth-order sigma-delta modulator (with a -200 dB noise floor with little to no modulation; you couldn’t make this stuff up) and finally to the Pulse Array DAC (based around a concept of a high, but constant switching rate to allow for predictable and therefore cancel-able noise levels). The closet electrical engineer in me swooned. But I wanted to get to know the person and not just the technology.
After the talk, while most attendees were upstairs munching on scones and drinking tea, I managed to sit down with Rob Watts for an in-depth talk. While the discussion revolved around lots of technical details relating to why he prefers using FPGAs and how there are so many more minute technical hurdles to overcome with ASICs (i.e. mass-produced, purpose-built IC chips), we did eventually get to his philosophy on sound and how it translates into his design approach.
Watts is [ ] essentially a one-man fab-less design firm who happens to care deeply about audio…
What I learned from my conversation with Rob Watts, was that he was a man who did not dissociate the engineer from his inner audiophile, and instead embraced both sides. Couple that with truly in-depth technical knowledge of semiconductor design that extends far beyond what is necessary for audio equipment design, and you have a man who comes as close to audio mastery as a person can get. Technical mastery is of paramount importance to the design of a DAC, but only because it’s so important for the technical details to become a non-issue when it comes to audio.
Despite this, Watts refreshingly takes on a very forthcoming attitude to his technical approach in the design of the Hugo‘s WTA interpolation filter and Pulse Array DAC. If anyone has ever ventured into the lengthy CHORD Hugo thread on the head-fi forums, they’d find out how willing Watts is at disclosing how the Hugo performs and why he made the design choices he did (such as increasing tap length to >26,000 coefficients, eating up precious computing power within the FPGAs). During our conversation, I’d even asked him a pointed question about having to implement sample decimation in DSD music from 2.8/5.6 MHz, down to 352.8 kHz (usually, this is a bad thing, as the high sampling rate/timing advantage of DSD is effectively abrogated with decimation). He immediately admitted that he’d originally been worried about the performance of DSD in that manner, but mentioned that his comparative testing between the Hugo and their QuteHD yielded no discernable differences.
There are some who say that this lone-wolf approach can’t possibly be better than the collective hive-mind of countless electrical engineers at large IC firms. How can a one-man army be good enough to trump the efforts of so many notable design firms (e.g. Cirrus Logic/Wolfson Microelectronics, ESSTech, Asahi-Kasei, Texas-Instruments, Analog Devices, etc.) that cater to the audio world? The truth is that, while large companies have the resources to scale their productions on a grander scale, their design teams for high-end DAC products are relatively small, often boiling down to a couple of people. The numbers, then, are more equal than one might think. Couple that with the fact that large resources are only necessary if a company intends on mass-production (hence requiring miniaturization, silicon tape-out, optimization, etc.), and the numbers really equal out. Consider the fact that the general Sabre DAC design from ESSTech was also essentially majority-designed by one man, Dustin Forman, and you can tell that it really doesn’t take millions of dollars to build a successful chip — unless you’re simultaneously planning on making millions of them at a time.
That’s why extremely high-end DAC companies insist on their own DAC designs, often using FPGAs as the base, just as Watts has done with the Hugo (and all other CHORD digital products, for that matter). Field-Programmable Gate Arrays, or FPGAs, are essentially blank slates. They can be programmed to do anything and everything you want in the computing world. You just need the knowledge to know how to program them to run the correct routines you want. In the world of high end audio, volumes don’t nearly reach the threshold required for silicon wafers to become affordable (versus non-recurring engineering costs), so FPGAs are actually an ideal.
When a scientist and engineer begins to apply his talents to a very humanist pursuit, good things happen.
Watts is, then, essentially a one-man fab-less design firm who happens to care deeply about audio, and that’s great for our prospects as end users. There’s a sense of sincerity to the manner with which Rob Watts describes his pursuit for ultimate sound quality. It goes beyond the marketing speak and transcends the esoteric jargon of digital signal processing. In the half-hour that I spent with him alone, I could easily tell he genuinely cares to achieve the best sound he could possibly manage out of his designs. That type of care is immutably coupled with the immense repository of knowledge at his disposal.
Whilst his day job is that of a semiconductor consultant dealing out IPs on DSP designs, Watts’ audiophile juices come out in his capacity with CHORD. He not only attempts to correlate how technical changes lead to changes in the listening experience, he also studies the human auditory system in his own time, even asking for my assistance (given my neuroscience and medical background) in identifying notable papers on hearing. When a scientist and engineer begins to apply his talents to a very humanist pursuit, good things happen.
Unfortunately, I didn’t get to spend a lot of quality time with the Hugo, and haven’t had time to return to ACE Audio since, but in the short time that I spent with the Hugo, out of an HD800 in a somewhat loud environment, I could already tell that the Hugo exhibited extremely smooth attack and decay characteristics, but still resolved a ton of detail.
Hopefully, I can find another time in the near future to get a deeper listen.
Now and again, a device comes along in audio that turns heads from every walk of life. Sometimes, those devices turn heads for the wrong reasons. Not so for the CHORD Hugo. It’s clear that both Franks and Watts are extremely proud of the Hugo; they gush with personal pride when speaking about it, and are aware of the market impact such a device has, despite its relatively high sticker price in the portable segment.
At nearly $2400 USD, the Hugo is not necessarily for the everyman. It is, however, the highest performance DAC ever, footprint-for-footprint. Measurements back up that statement, and the subjective experience backs up that statement for a variety of people. The only question now is how to get the Hugo (or at least its trickle-down technology) into the hands of more people.