Wednesday, March 26, 2008

The Importance of Xd

It wasn't long after getting into audio that I realized speakers are the weakest link in the audio playback chain.  At the same time it was clear they also are the key to creating the magic... occasionally.  Listening to the right recording, in the right room and in the right mood shows us that replicating the live performance is possible but only when the stars are aligned a certain way and all the mojo is working.

It makes no sense to me that enthusiasts will fuss over a tenth of a percentage point in amplifier THD, all the while listening through loudspeakers that at their very best produce 10 times more distortion.  These contradictions were the reason I got into the loudspeaker business.  There just had to be a better way.

Every decision made in passive (and most active) speaker designs involves compromise.  Acoustic designers have to prioritize and balance electrical and mechanical distortion, frequency and time domain response, not to mention guessing what the listening room will be like.  This is one reason why they all sound different, and why any speaker will sound different based where it is placed.  Drivers and crossovers are imprecise, clumsy devices and choosing the right combinations rely as much on intuition as science.  While our industry has made some improvements over the years, there is very little we can do to advance the craft as it is.

While NHT was working on analog technologies such as our Focused Image Geometry (a method for minimizing the effect of the room on sonic playback), on the other side of the planet one of early pioneers of DSP and a co-founder of Fairlight was working on digital solutions for improving loudspeaker performance.  We met each other some years later and Xd was born.

The belief that we held in common with our new partners at DEQX is that the speaker really must be corrected first, then room correction was feasible, at least in the low frequency range.  When you can control both frequency and time domain response, and provide steep digital filters through software, most of the compromises we've lived with for years disappear.  Driver design then has one real purpose; to minimize mechanical distortion.

Don't misunderstand.  It still takes a good acoustics engineer to achieve quality results.  The interesting thing about software-based correction is that it gives designers a whole new pallet of tools and possible directions.  It means we can build smaller, affordable systems that will outperform traditional products by huge margins and they can be placed almost anywhere in a room.  In addition, the amazing resolution is so revealing you can actually hear a difference using better quality components.  And while a sub-$1000 dollar DSP corrected system will  blow any existing passive system away, there will be room for higher end systems.  This leaves lots of room for multiple speaker brands and lots of potential choices for us listeners.

So Xd goes on the "shelf" for a while.  Our first effort proved what we wanted it to, but the hardware is outdated, and there are better, faster, cheaper chip sets out there that we need to work with.  Back to the design caves we go, hopefully others in the loudspeaker field will join the effort and advance the technology quickly.
 

Wednesday, March 5, 2008

Disco Daze

NHT is not particularly fast at product development but when we finally get it ready to produce, we typically have a product that really performs well for its intended purpose. 

With every new product, the important details are labored over every step of the way.  By the time it is finished, we have taken it home, listened in every conceivable scenario, corrected problems, argued about voicing, and listened some more.  When done, that new product has become a part of our brand psyche, and part of the family.

This is why I hate discontinuing products.  It feels to me like abandoning a favorite pet by the roadside. But if you have taken the time to read this blog you know that NHT is once again a small, independent company.  This has forced us to look at what we offer and reduce the models to a select line of our best and most successful speaker systems.  

The changing - and contracting shape of the traditional audio business is diminishing the supplier base; another reason to take a long hard look at what the consumer market considers relevant. For years NHT bought drivers from a number of suppliers located primarily in Japan and Europe.  While more expensive, we supported these companies because of their investment in innovation and parts consistency, two key elements for building great-sounding speakers.  There are lots of driver suppliers springing up in Asia, and many are quite competent, but they do little R&D and most lack consistent suppliers for their components.  This actually has less impact on our new product development as we can design around problems, but makes supporting older models increasingly challenging.

Lastly, the newest reason for discontinuing speaker models: technology.  Electronics makers have been dealing with this for years, but not us speaker guys.  Our fundamental technology has not changed since 1940... at least not until recently.  In NHT's case, it's called Xd.  Xd represents the advanced use of DSP in the correction of loudspeaker performance.  Even though we began shipping Xd two years ago, the underlying technology is still considered radical and new.  It is dependent on microprocessors and is expensive to implement even in its most basic form.

Our original model is still one of the most amazing sounding loudspeakers I have ever heard, and we firmly believe the technology represents the future, which is why we decided to put it on hold until a new design is complete.

This is a longer story - stay tuned for part 2.