Bevan said:
WmAx, could you say more about this? i dont quite understand what an abnormal amplitude vs. frequency response is, and why the crossover point or number of subs affects this. many thanks
b.
In order to have a flat response/linear integration of two merging acoustic signals, they must combine in relative phase rotation to each other according to the crossover type/target response that is desired. If there were only two signals[one main and one sub for example] it would be easy to delay the farther away source to match the same arrival time as the closer source, thus providing for optimal integration of the two signals. However, let's say you have two main channels and one subwoofer. If you locate the subwoofer in a position where it is the same distance from each main speaker relative to the listening position, then you can easily delay the subwoofer and accomplish an excellent integration. But it's not so easy to locate the subwoofer in this fashion, and even if you had the middle space between the mains available for placement, this is usually a very poor location for a subwoofer relative to room interaction(s). If the subwoofer is moved off to one side or in a corner, for example, then the sound sources can never have an ideal integration[
unless you are crossing at a very low frequency and with a very steep xover rate, because then the wavelengths become large relative to the room dimensions thus lessening or eliminating the integration problems]. You will have 3 sources in this example. If you have two sources[mains] and an assymetric subwoofer placement[relative to wavelength], you can only compensate for the time differential of one of the mains, not both. This will cause a frequency response error, potentially of large magnitude/cancellation, if at a sufficient placement to combine the signals at the listening position in or at close to phase inversion relative to each other. It is an incidental situation[you can calculate accurately the effect at a specific placment, but then you have to also be able to predict the best room placement acoustically for the subwoofer and have them match up]. You can try to compensate by placing the subwoofer at an assymetric distance where the 2nd arrival is in relative phase[1 cycle delayed, 360 degrees rotated] and also in phase with the other channel[
This would require placing the subwoofer in a position where it is 1 wavelength in distance from the farther channel mathematically after compensating for the time delay with the nearer channel. This will usually require a significant effort in room placement and may compromise the main channel placement.]. The easiest/best way to integrate the subwoofer is to use one at very low frequency crossover point with a steep crossover rate[4th order, 40-50Hz would be sufficient for average rooms/distances] or to use 2 subwoofers, one placed near each main speaker. So long as a steep xover rate is used in combination with a low frequency, a single subwoofer will work. However, in the event that a larger room/farther distances are concerned or a higher crossover frequency[remember that much higher frequencies than the xover frequency are still transmitted, dependant on the slope cutoff rate] or a more gradual crossover rate, two subwoofers[one placed near each main channel relative to crossover wavelength] are ideal. In the case of 6" or 7" 2 way speakers, it is most beneficial to crossover at around 80Hz or higher, in order to substantially reduce the cone excursion/incursion in order to reduce modulated phase distortion[a.k.a. Doppler distortion] on moderate to high SPL situations. All of these things must be considered in a particular situation to determine if 1 or 2 subwoofers are appropriate. It needs to be stated that both the subwoofer and mains MUST be crossed over with an accurate line level crossover if the full benefits/ideal integration is to be realized. For a large floor standing 3 way speaker with a 10" woofer, a 40 Hz crossover rate should be no problem. For a 2 way 6.5" woofer speaker, a 40Hz crossover rate would not be sufficient to remove the burden of cone movement from the midrange driver that will hinder performance at moderate to high SPLs.
-Chris
