The Speaker Guru speaks: Dave Dal Farra

Speaker Designer Questions

DIY speaker design nut for over 25 years, worked 10+ years professionally in telecom as a senior audio design engineer: designing sound rooms, running blind psychoacoustic experiments, designing audio electronics, DSP audio algorithms, audio measurement methodologies for standards bodies, mic/speaker integration etc. Designed numerous home speakers over the years since the late 70s, on and off.

1. Without giving the standard "weakest link" answer, how important would you rate speakers as component?
Loudspeakers are much more complex devices to specify than electronics due to their inherent interactions with the room, and the nature of spatial perception, both of which are inadequately studied in the context of high-end audio. This leaves speaker design much more open to personal interpretation and “expression”. This, and their obvious non-linearity limitations, makes them by far the richest areas to mine for improvement.

2. How have speakers changed in the past 25 years?
Technology convergences (home theatre, computer, iPods etc) and much greater increases in personal wealth in the west have created opportunity. The result has been a great drive towards smaller size, lower price, and ultimately better quality control through improved manufacturing processes. The trickle down to DIY has been enormous choice, lower cost of entry, production tolerance improvements, and much more linear long throw (i.e. smaller) drivers. 25 yrs ago, you were stuck with a few Vifa drivers and a handful of Peerless and Philips parts, if you were lucky. How many back-chambered tweeters (never mind ribbons) were available in 1980? Times are good.

3. Do we have to spend a lot on woofers and tweeters to get good sound?
While they are far from perfect, playing with a few $15 Tang Bands answers that question.

4. What are the top three design parameters you use? (or the top three things you worry about most)
#1. Frequency response
#2. Directivity
#3. Dynamic linearity
Frequency response aberrations are easily the most audible degradations. Directivity is merely a second order application of frequency response. Mind you, there is no one right frequency response target, as bandwidth (especially), distortion characteristics, box resonances etc all determine what the target should be for a particular design. I do not believe in an all-purpose ruler-flat school of design, though it’s the starting point. Dynamic linearity is important as it expresses wide dynamic range and low distortion all in one parameter. This often sets crossover frequency and order, box alignment, and driver choice and sizes.

5. Do you design on measurement or sound?
Design on measurement, hunt and peck and finalize on sound, using the measurements to enable the changes your ears are driving. Twenty years ago, understanding the physics/electronics, and being the lucky rare person to have simulation ability really differentiated design acumen. Now, the proliferation of low cost tools have freed up the DIY community to focus on what is really important: learning the art of the recording engineer/producer, identifying what to adjust in the sound, guided by listening, to attain neutrality.

6. What to you think is that "special something" is that makes some speakers sound so good?
Have to agree with the other responses, an open midrange. When it sounds right, the analytical side of the brain reflexively turns off. That’s when you know the design is done. You can tweak the frequency extremes through placement, the mids you have to get right out of the box.

7. What are your speaker audition strong points and weak points, i.e., what problems do you find most difficult to pick up in listening tests?
Due to the difficulties in attaining good measurement resolution in the low frequencies without an anechoic chamber or a good outdoor environment, the most difficult aspect for me would have to be balancing the response below 800Hz and controlling any aberrations in that range.

8. Do you have a theory on crossovers or use whatever seems to work?
Where possible, go small: smaller component count, lower order and Q’s (lower sensitivity to driver parameter changes with excursion and heating), smaller caps…. IME, it increases openness. Every design seems to have a minimum acceptable crossover topology, finding it is half the challenge.

9. Do we overstress the bottom octave and area above 20K Hz?
Most of us listen far enough off axis that we’d be shocked at what the tweeter is doing above 20 kHz. While little music has content below 40 Hz, moving air in that range is important to provide the physical cues of live music, when needed.

10. What one piece of advice would you give every new speaker builder?
Take up golf!

11. When you get cornered at a party by a speaker builder, what is the question they ask most often?
“My mom wants to buy new speakers, what should I get her?”

12. Will we ever find the Holy Grail in sound reproduction?
Ever is a long time. Having played professionally with personalized auralization (i.e. measuring and simulating your own head’s diffraction) over headphones, I would lean in that direction. It removes the room from the equation. Cheap and user-friendly DSP will be the enabler. The challenge will be factoring user head movement into the illusion.

13. So, which is it, tubes or SS? Looking to date a Supermodel?
Tubes. Looking for a wife? Solid State. Tubes can sound wonderful, and get an undeserved bad rap due to some poor boutique implementations, but I can more easily live with the deficiencies of really good solid state. SS also seems to be more speaker tolerant, requiring less matching. This is a big advantage for the DIY speaker designer.

14. What does the future hold for speaker design?
DSP, real DSP where you write, or at least configure, the algorithms vs. twiddling prefab settings. Get ready for it, its fantastic. There are numerous applications well beyond crossovers. I’ve been speaking with an old acquaintance that is now operating an audio signal processing company. The single ended non-linear algorithms the industry is perfecting today for barrel (echo) suppression will shake things up in high-end room correction.

Questions for the driver designer:

A. What was the best part of working for a large speaker manufacturer?
I’d like to answer this from the telecom side. Toys, and more toys, access to some of the brightest minds in the audio industry, and cutting edge research. It’s been said on the net before, and I can support it: the most advanced audio work is being done where there is the most money, sometimes in the seemingly inconspicuous places such as telecom (Bell Labs was proof enough of that).

speaker icon

Copyright Peter Jay Smith 2005 Return to