In this article we'd like to highlight an aspect of speaker design that is not widely discussed but has significant impact on what you hear; mid and high frequency diffraction. Remembering that sound is a series of pressure/rarefaction waves with low frequencies being the longest and high frequencies the shortest, diffraction is defined as the process by which a system of waves spreads out due to passing through a narrow aperture or across an edge. This is typically accompanied by interference between the original pressure waves and new waves formed as the original waves pass through the aperture or over the edge. If the wavelength is bigger than the size of the obstacle, it passes through it like it's not even there. So for woofers, diffraction is not much of an issue. For higher frequencies however, with their shorter wavelengths, diffraction can and will happen.
We can see in this drawn explanation of diffraction that when a driver is generating sound waves the waves expand in a hemispherical fashion. As the waves reach the edge of the typical speaker baffle, the pressure drops, because the waves expand from half space to full space*. This pressure difference creates another sound wave, which originates from the edge of the cabinet and moves forward. The newly created wave will have phase mismatches with the original wave, and it can cause interference with the original. Our hearing mechanism subconsciously perceives this and as such, when there is sufficient time delay of the diffracted sound, as occurs with wide baffles, this degrades imaging and helps contribute to a boxy sound. Since the edge runs all along the baffle, different distances from the drivers to the edge means different wavelengths will be affected. You will note that we mount our tweeters off center, which results in two different diffraction time coefficients, which helps smooth response and minimize diffraction issues.
Those of you who have been around audio for some time might remember the famous Dahlquist DQ-10 from the 1970's. Jon Dahlqvist was one of the first designers to seriously take into account diffraction issues. He utilized a very complex multiple baffle concept in order to achieve his goals. In fact, we have learned a lot since Jon's day and we are able to optimize diffraction issues in a much simpler and in many ways more effective way.
The GoldenEar Way
Speaker designers can deal with this in multiple ways, to greater or lesser degrees. Here's how we control and reduce diffraction to a negligible amount in GoldenEar Tritons and other models:
The powerful racetrack shape subwoofers in GoldenEar Tritons are custom made to our demanding standards. A major reason for their unusual shape is their narrow cross section that allows us to keep the front baffles of the Triton series exceptionally small. These front baffles are effectively no wider than the frame of the midrange drivers. The result is that the diffraction from the edges of the cabinet occurs with very little time delay relative to the initial radiation of the drivers, and so that our hearing mechanism does not perceive this diffraction separately from the primary radiation. Interestingly, one point that Jon Dahlquist made to Sandy, back in the day, was that we are very sensitive to the human voice, and that optimally baffle width would be similar to the width of a human head.
The faceplate of the HVFR tweeter is shaped in such a way that it dramatically reduces the potential for edge diffraction from the high frequencies it produces. Compensation is also made in the advanced crossover networks to eliminate any remaining artifacts. And, as we previously noted, the tweeters are located off-center on the baffles to further ameliorate diffraction issues.
The result of all this is exceptionally clean and clear mids and highs with superb boxes sound and wonderful three-dimensional imaging. And as many of you have experienced, we've achieved this level of performance while remaining true to our stated goals of performance versus price.
* An extremely simplified explanation of half and full space:
Full space would have no reflective surfaces anywhere near the speaker system, like suspended outdoors or even in an anechoic chamber. Half space would have one reflective surface adjacent to the speaker like a back wall or side wall or floor, etc.