Last time I found some rather odd variations between my pair of RS SPL meters and my Galaxy 140. I wanted some new nearfield comparisons so took this opportunity to do a completely new set of REW measurements. I also included my new ECM8000 microphone and Xenyx 802 mic preamp in the tests. It soon became clear that the REW Cal file for my two RS meters was seriously out of step with the Galaxy and ECM8000. So I deleted the 33-2050 cal. file and downloaded the 33-2055-4050 instead. The latter file gave curves much closer to the Galaxy and Behringer microphone. Now I could progress with my testing.
All devices were tested twice without being touched to exclude possible environmental variations. (wind, distant traffic, people moving around in the house, doors opening and closing, etc)
The red trace below is the ECM8000 and those above this line are the Galaxy and then the two RS meters respectively. All traces have the BFD out of circuit. All devices used the matching REW calibration file freshly downloaded from the HT Shack forum. The CX2310 active crossover remained in circuit with a 24dB/octave cut at around 70-80Hz. It is difficult to read 80Hz accurately on the frequency setting dial. The speakers were muted in all cases at the crossover. One of the advantages of having all those push buttons.
These nearfield curves are all quite smooth and each pair of tests per device match each other closely enough to give me confidence that there is nothing untoward in my test methods this time.
A cynic might argue that the ECM8000 is reading low in the deep bass since it does not match the three other devices which all match each other closely. However, the high cost of professional microphone calibration will not have me following that route in a hurry. For my subwoofer testing purposes they all do well enough. The Behringer mic and Galaxy SPL meter are better for fullrange speaker testing at higher frequencies. These two are also better at reading deep infrasonic output accurately since the correction factors are very much smaller than the cheaper SPL meters and the cal file extends down to 5Hz in the case of the Galaxy 140.
It should be noted that some of my other graphs show much closer correlation between all these measuring devices. I have convinced myself that any of these four devices is suitable for measuring most domestic subwoofers. Only when speakers or IBs are to be measured should the Galaxy 140 or the ECM8000/Xenyx combination be seriously considered.
All the devices were laid horizontally and individually on a cushion on the floor at the mouth of my IB for these tests. The small variations seen could well be down to slight changes in position or minor resonance in the bodies of the SPL meters. The very solid, metal-bodied ECM8000 is probably immune to this but compressed the cushion much more than the other devices because of its comparatively small cross section and considerable weight.
From Left to Right: RS 33-2050, 33-4050, Galaxy 140 SPL meter and ECM8000 calibration microphone. The ECM8000 needs a mic preamp with phantom power adding considerably to the cost of this inexpensive microphone. I believe some computer sound cards offer phantom power so that may be worth exploring.
I chose to buy a Behringer Xenyx 802 to go with the microphone because this was the unit suggested by the HT Shack experts. This offered a proven device with settings already worked out for the new user. Here I have plugged the microphone into the input XLR socket. Normally an XLR cable will fit in here and the microphone would be mounted on a tripod for testing. Note the positions of the control knobs on the Xenyx for REW testing. The HT Shack has a better image for confirming these settings.
The background noise level reading on the Galaxy in this shot is due to the proximity of my computer and the C-weighting option being selected. A-weighting background noise in my room is around 31.5dB(A) at the listening position or 43dB(C-weighting) . Both readings were taken with the Galaxy resting on my listening chair with the rather noisy computer running about twelve feet away.
The nearfield curves in the REW graph show little or no room gain. At the listening position the lower frequencies are boosted considerably by room gain. The nearfield curves are reasonably flat from 20Hz upwards and speak volumes about the quality of these AEIB15 drivers.
Later reading suggests that there is some spread to the frequency response of the inexpensive Behringer ECM8000 microphone. Those who wish to enlighten themselves or have their microphones properly calibrated could do worse than check the HT Shack threads on this subject.
Remember that you can click on all images for a larger version. Back click to return to the blog. .