Bass thoughts.

Elsewhere in this issue you will find my review of the remarkable Trinaural Processor from Spread Spectrum Technologies, designed by the venerable James Bongiorno. Although I am extremely impressed by what the unit can do in terms of enhancing the enjoyment of music, I am less impressed by its attempt at being compatible with a subwoofer. Giving this matter some further thought, I have tried to put the whole matter of bass performance in music and A/V into perspective, because this is a subject I fear that too many audiophiles are not giving proper attention.

With regard to the fixed-corner 2nd order crossover in the Trinaural Processor, for example, I think it is next to useless, as are all such built-in fixed 2nd-order crossovers on A/V receivers starting at $200. Move up to the $400-$500 range and you get much more sophisticated approaches in some models. (Spread Spectrum will argue theirs is an all-analog crossover, but some models in the $400+ price range are offering fully analog bass management.)

As an absolute minimum, you need a 4th-order crossover (my own research work, soon to be published, shows even this is not adequate and is audible on double blind tests). Also as a minimum you need a crossover with a variable cut-off points since the main loudspeakers' low-end rolloff could be 30Hz or 150Hz. Even then you are going to have problems if you do not provide some sort of electronic driver compensation for the phase response changes that occur at 5 times the cutoff frequency of the woofer (details of this are in a paper that a student presented in March at an AES conference and will be printed in an AES journal in October). Phase changes are not directly audible in themselves, but in a crossover system the phase changes manifest themselves as significant changes in frequency response (the speakers are not pushing and pulling in synchronization over some part of the frequency band at which they are being crossed over).

Adding a subwoofer may get a frequency response extension below 40Hz for a full-range speaker, but nothing in classical music is down there except the organ. The 40Hz-80Hz range is totally messed up, with +6-30dB hills and valleys with a full range main speaker; if we are talking cubic foot boxes with 4-inch woofers, then 80Hz-200Hz may be sucked out completely. With 6-inch woofers in 1-cubic-foot boxes, we are somewhere in the middle, with +6/-30dB stuff happening out to 120Hz or so and with a big depression in the 70-90Hz region.

Ken Pohlmann gets it exactly correct in his article "Roll over Beethoven" in the April 2003 issue of Sound and Vision. The frequency aberrations and response holes introduced by a subwoofer/main channel interface in the 50-200Hz region "cuts out the balls of Beethoven, the warmth of Brahms, the power of Wagner. To put in bluntly classical music sounds bad on typical home theater speaker systems" ... "it can sound pretty good with pop music, which comprises vocals, a crisp high end and a deep bass beat." Pohlmann then goes on to argue that the poor reproduction of classical music by an A/V speaker system at a $1,000 price point (his number, not mine, although I am in agreement with the number) and below "will hasten the decline of classical music playback in the home." This is written not in some High High end magazine with a readership of 2000 but in the largest (by a factor of 10) A/V magazine by one of the most respected people in the field.

So you want to cross over two main speakers to a subwoofer? The main speakers are producing tones in the 40-200Hz region where the frequency response of the speaker is dominated by room placement, and the subwoofer's frequency response is also dominated by the room. Each speaker is in a different part of the room, with different room-related frequency aberrations. To make things even more difficult, you want to try to crossover a ported woofer to a subwoofer, where the subwoofer is at a significant distance in the horizontal and vertical plane from the main speakers. Add in the fact you fact you are summing to mono from two stereo speakers and you may come to understand that you are faced with the hardest crossover problem that exists in the Hi Fi world.

Note well this is not a problem that five degreed engineers at a major speaker manufacturer with large hardware and software facilities are going to solve. YOU get to solve it! Even if you purchase your subwoofer from the same manufacturer as the main channel speakers, you have no doubt also picked a receiver (or other external signal processing box such as the Trinaural processor) that is very unlikely to have the correct electronic crossover to create a flat response. Moreover, as Pohlmann points out, the agenda of the speaker manufacturer is to get it right for rock and car crashes. Flat response in the 50Hz-200Hz range is not even close to a top priority for most packaged 5.1 speaker designs.

Want to make the problem even harder? Well the subwoofer may or may not have an internal high pass filter, which often has a knob that claims to change the cutoff frequency but often also changes the Q of the roll off--and it may be badly miscalibrated to boot. YOU get to figure where to turn that knob for the flattest in-room frequency response, or whether you want to bypass it or the low pass crossover in the receiver (if this is an option, which it is not with the Trinaural processor).

Have you ever seen a midrange in a ported enclosure? Such a move could improve the efficiency of the midrange, but it also would introduce a 4th-order high pass response at the midrange's low end. Loudspeaker designers know just how difficult it is to make a smooth passive crossover between a ported midrange and a woofer, so they just are not willing to attempt a design with a ported midrange. Now what is the audiophile trying to do when he matches a subwoofer to his main channel speaker that inevitably has a ported woofer? He is trying to solve the problem that professional speaker designers run away from. It should be noted here that a couple of speaker manufacturers supply material to stuff the port of their speakers when a subwoofer is used in an attempt to create a compromise 2nd-order high pass rolloff from the main speakers. These manufacturers who offer such a port stuffer are tacitly admitting how difficult the subwoofer integration problem is.

Truth be told, all speakers need EQ in the region below 120Hz where the frequency response is dominated by the room. The Infinity RABOS system is an example. Even this very large-scale mass marketer realizes the problem is so significant that it must introduce a system that is at best complex and confusing to the average consumer. When a mono subwoofer enters the picture, the room's role in defining the frequency response is much more significant. To avoid this problem, Infinity sells most floorstanding speakers with integral passive or powered sub woofers (note no subwoofer is in the Kappa series).

Pioneer takes the equalization problem to the next step in its receivers by including a microphone and a DSP-based room equalization system. Such a system may be able to reduce the hills in the subwoofer/main speaker in-room frequency response, but it can do nothing about the wide and deep depressions.

Spread Spectrum has produced a product for music reproduction that greatly enhances how well a Hi Fi system accurately reproduces music produced by acoustic instruments. The processor is a very expensive piece of equipment that is well worth its price when it added to a Hi Fi system that has been purchased and assembled with its primary goal being the reproduction of music. However, the company has added in an inadequate subwoofer crossover section and a five-channel analog passthrough in an effort to state that the product is A/V compatible. To put it bluntly, these efforts are not successful. Furthermore, I restate the need to have a full-range speaker placed directly in the center for this thing to work at all. You cannot compromise this and thus you cannot have a successful combined A/V system and 3-channel audio system all in one.

In my opinion, subwoofers should not in most cases be used to attempt to enhance the accuracy of classical music or acoustic jazz reproduction. The frequency response aberrations introduced in the all-important 80-200Hz region far outweigh the enhanced low frequency reproduction below 50Hz. Is reproduction of a 32-foot organ pipe really more important to you than the accuracy of reproduction of the cello and trombone sections of a orchestra?

For accurate reproduction, you should use full-range speakers (very good ones are available starting at $1000/pair and excellent ones at $2000/pair). If you have chosen a bookshelf speaker, then you should be prepared to live with the bass restrictions of the system. Some powered and actively equalized speakers in this category, used in professional applications, are well under $1,000 and can reproduce 40Hz at 95dB SPL level at 1 meter with less than 10% distortion--good enough for a small room. --DAR