$250 MTM Horizontally Oriented Measurements

The first speaker we’ll look at is a very common MTM design, available from a big box retailer where you can reportedly get some best buys. The picture shows the speaker horizontally placed on a tripod with angle markings so I can accurately measure the speaker’s off-axis response. The horn-loaded tweeter crosses over to the dual 5.25 inch midrange drivers at 2400 Hz. This means that while we will see the performance of the tweeter’s off-axis response, we will focus our analysis in the bandwidth of 80 to 2400 Hz.

In the chart below, the frequency response of this MTM center channel was mapped with 1/24 resolution, from 80 to 20,000 Hz in the x axis and from zero to 40 degrees angle in the z axis (front to back). The initial response on axis was normalized so the measurement you see in the chart is the deviation from the on-axis response as the angle of the speaker was changed. You can see a significant change in frequency response in the top octave, but we’re not here to pick on the off-axis response of this speaker’s horn tweeter, as terrible as it is. What we are here to pick on is the frequencies that the two “Ms” are reproducing, from 80 to 2400 Hz, which is shown with the yellow bracket. In this zone the measurements show peaks of over four dB from the additive effect of the two midranges and wave cancellation of over -10 dB.

The 1/24 octave chart above looks pretty messy. However, the peaks and troughs are less audible the narrower they are because your hearing will naturally smooth out the response. Plus or minus a couple dB also isn’t anything to lose sleep about, but what we can see from the chart is that the two midrange drivers exhibit significant wave cancellation from around 800 Hz to where the crossover kicks in at 2400 Hz. Some of the variation near the crossover point is from all three drivers playing the same frequencies. A higher-order crossover would reduce this problem, but just placing the tweeter above the midrange would fix it completely as we’ll see later on.

If we take a slice of one frequency from the 1/24 octave measurements we can plot out the frequency response of the loudspeaker and get a visual representation of lobing. The lobing effect is from the appearance on a polar map of the peaks and valleys of the frequency response. In a real polar graph, it looks like lobes, or flower petals. This chart is a bit simpler, as I only measured an 80 degree angle (assumed to be symmetrical) and am using a more traditional chart.

Conventional wisdom is that people typically can’t differentiate frequency changes finer than 1/3 octave, but I argue that is a bit too course a guideline to follow. In my experience, 1/6 octave smoothing is a more appropriate method to determine audibility of frequency responses. In order to better chart the audible effects of the wave interference, the chart below shows the same measurements but averaged at a 1/6 octave resolution. This chart indicates the significant, and audible, wave cancellation in the upper midrange. When the variation from on-axis response, from 80 to 2400 Hz is calculated using standard deviation, this speaker gets a score of 1.94. We’ll see later how that compares to other types of center channels.