The duration shown in the plot is approximately 200mS. The top trace,
in red, is that from the gym measurement; and, the three below are from
the middle microphone positions for each of the three distance conditions.
The front wall behind the speakers has had several acoustic treatments applied to it in past decades. There is a 1" thick layer of acoustic cork glued to the drywall, and that is now covered by a layer of fiber 'sea mat'. The floor is hardwood, and covered over much of it's area by thin area rugs. The metal 'industrial' rack behind the acoustic foam contains the amplifier(s) and a television. The primary audio control and signal sources are contained in a rack in the L-section on the opposite side of the room. There are line level cables and video cables installed under the floor for signal transfer from one side of the room to the other.
The acoustic foam seen between the speakers is augmented by two sections of the same foam (Markertek 2")suspended from the top of a 6.5' tall Torchere halogen lamp behind each speaker. This places the foam mid-way between the wall and the speaker. The foam pieces are roughly trapezoid, 12" wide at the top and 18" wide at the bottom (and are the salvaged sections previously on the rear of the gym version of DynaPleat arrays).
I give equal weight to both Home Theater and Music reproduction since I must use the same room for both functions. When playing a video source the 54" square piece of acoustic foam hanging from the equipment rack is folded over the top of the rack to expose the television. The position of the foam, up or down, creates very large differences in the perceived sound of the system.
Subjective descriptions of those differences are not my particular strength, but in numerous demonstrations of the effect of the foam, up or down, all have noted that sonic imaging becomes much more "solid" and sounds seem "clearer" went the foam is down. This is not to disparage the sonic environment when the foam is up for Home Theater, but I always lower the foam for critical music listening.
When the foam is up for Home Theater purposes there is no sense of non-localized voices, but the sound is much more diffuse - bigger if you will. Often the 'bigger' sound is just what is desired when an asteroid is crashing into the planet, a fusion bomb is exploding over L.A., or a large-toothed creature is chasing one around and about.
Let me be quite clear here. That (now) $20 piece of foam between the two speakers makes for two VERY different sound fields in the room when either up or down. That difference is so great as to imply two different loudspeaker systems - which is obviously not the case. The importance of room treatment effects cannot be overstated in my opinion. I am in process of acquiring more acoustic foam with which to fabricate a complete floor-to-speaker top foam insert (82" high) to replace the hanging 54" square section that is shown in the photo. I anticipate further improvement in the room sound field from that change - and an even stranger 'look' to system.
The foam sheet was in the down position as shown in the photograph during this study. The left speaker was measured at three distances from the front wall, as follows:
1. At 22" from the front wall to the inner edge of the speaker, and towed-in such that the front plane of the speaker was perpendicular to a vector drawn to the central seating postion. At 22" from the front wall this placed the speaker's inner edge flush with the foam, thereby presumably increasing to some extent the dipole baffle's effective width.
2. At 33" as above.
3. At 44" as above.
In each condition the inner edge of the dipole baffle was kept on line with the initial orientation relative to the hanging foam sheet. That is, the baffle was pulled directly out from the wall, there was no lateral shifting of it's position.
Five measurements were made for each experimental condition. The couch at the listening position was moved against the back wall, and five microphone positions were defined across the 6' width of the seating area, each being 18" laterally separated from the others. The microphone was 11.5' from the front wall. Facing the front wall, there was an extreme right position, a right, a middle, a left, and an extreme left microphone position. The middle microphone position is centered in the 14' span created by the 2' extension of the 12' floor width by the bay window. The microphone height was at 37", which is the nominal listening height from the couch.
Given the three baffle positions and the five microphone positions there were a range of distances involved. Distances to the middle microphone position from the left speaker were approximately 8 (2.4m) to 10 (3.0m) feet over the three experimental conditions.
Prior to presenting the frequency response results it will be worthwhile to consider the impulse responses from the gym and listening room measurements as an additional aid in understanding in-room low frequency response evaluations. Below is an overlay plot of four impulse responses.
The duration shown in the plot is approximately 200mS. The top trace,
in red, is that from the gym measurement; and, the three below are from
the middle microphone positions for each of the three distance conditions.
The gym impulse was acquired at a higher input level than the listening
room impulses, but what should be immediately obvious is the large gap
between the impulse and the following room reflections. The time duration
between the beginning of the impulse and the beginning of the reflections
is approximately 63.5mS. That gap is what makes a 'semi-anechoic'
measurement possible.
What of the listening room impulses? Where is the gap? We see from the plots that the reverberations in the listening room are just about dying out by the time the reveberations in the gymnasium arrive at the microphone position.
There is no gap between the impulse and the reflections! There is no way to tell what is speaker response and what is room response - and this is the delimma for a speaker designer who wishes to know the true low frequency response of a loudspeaker at listening distances. One can never know from measurements made in small rooms.
This is where the value of comparative studies comes into play. The frequency response data will be presented in several different ways, including direct comparison with the semi-anechoic data from the gym. A reminder that visual analysis of such data is facilitated by loading the GIF files into a slideshow program and rapidly alternating between the plots.
Below are three plot overlays of the the three different conditions. Each has five plots representing the responses obtained from each of the microphone positions. These plots are 1/12 octave smoothed, but there is no FFT windowing utilized. The vertical amplitude scale is 12dB per division, which allows for the entire extent of the frequency response to be displayed.
Plot overlay for the 22" from front wall condition.
Plot overlay for the 33" from front wall condition.
Plot overlay for the 44" from front wall condition.
Page 1
Page 2
Page 4 - Next Page
Acoustic Line Source Research - Table of Contents.
