Posted by Ivan Beaver of Danley Sound Labs.
"For those playing along at home, here is a run down of the setup used and some of the test procedures.
In the case of the powered loudspeakers, no particular input voltage was used, just a general level to get in the ball park-level wise. And no attempt was made to match levels. So any difference in the level of the powered loudspeakers was simply due to where the level control was set.
In the case of the nonpowered loudspeakers, almost all of them were run off of one half of a Crest 8002 amp. The exception would be Ricci’s speakers which were wired to be bridged and they were run off a Crest 8002 mono Bridged.
The only times an amp was changed was with the Worx and the Danley TH50 subs. A Crown I Tech 8000 was shipped in to be used with the Worx loudspeaker. So after the measurements were done, the amp was swapped out to provide the drive for that cabinet for the duration of the listening.
On the Danley TH50, the measurements were done on the same Crest 8002, but then it was moved to a larger amp, a QSC PL9. It tuned out that we didn’t need the extra power at all-as the amp was barely tickling the -20dB LED’s.
This resulted in a 6dB gain (because of double the input voltage) being applied to them during the testing phase. So when looking at the sensitivities of the different loudspeakers, those graphs would be actually 6 dB down in level. Those are noted in the names on the graph.
We also made no attempt to account for impedance/ wattage on the input levels. They all received the same input voltage. So the drivers that were of a lower impedance would be “getting” more power. Since the impedance of loudspeakers is NOT a constant (and varies with freq), I feel this is a fine way to test. It is important however to understand what the input actually was and what was done.
The basic procedure that we used was as follows. I did my tests and then Ricci did his. We approached them very differently and the results are sometimes VERY different. It is not a matter of one being “right” or “better” than the other, just different approaches and positions in the room.
I am sure Ricci will tell of his setup, but mine was a matched Earthworks M30 mic placed in a “ground plane” measurement-but actually located in the middle of the largest wall in the room, which was right behind the couch, where most of the people sat during the day. I could not get to a corner easily. My measurement software of the day was EASRA SYSTUNE.
The idea being to attempt to remove one of the room modes. And to try and get a more consistent measurement.
Ricci used the 1Meter (OK actually 3’-we used a yardstick), method. He placed the mic 3’ from the front of the loudspeaker-also in a ground plane position. This actual physical position (in the room) moved throughout the day, due to the different depths of the various cabinets being used, but the distance from the loudspeaker remained constant. You will see different room modes affecting the response in his measurements.
We did do a close up measurement of the Shiva and it was almost textbook perfect-with no room interactions-but this was measured right up against the cone and right at the exit of the port.
So anyway-after we measured (which probably took about 1 minute total) we put in the THX intro track. It was the most brutal track used and we ran each loudspeaker up to what we felt would be safe. This was very unscientific and was solely based on our judgments. The particular point of level adjust was made on the input to a mixer/splitter that Ricci provided. This kept the measurement levels unaffected by the change.
Remember this was intended to be a LISTENING event-so not a lot (almost none) of time was spent trying to “dial” in each sub-as we would not have time for that, and most people in attendance would have gotten bored quickly. I am sure we will get lots of “flack” for that-but those who want to throw up the flack can have their own event and see what other “issues” come up because of that. HA-HA We wanted to listen to a good bit of music and watch movie clips and go through a good number of loudspeakers-so some things just can’t be done in the time allowed.
The level of the mains and surrounds did not change at all during the day. We also used the internal 100Hz low pass (that was as low as it would go) inside his receiver. High passes were varied a bit during the day-as will be noted later.
We then proceeded to listen to a variety of music clips and styles, with one brutal track with lots of 16 Hz notes in it-Bass I love you.
Then it was on to the movies.
We used the same music and video tracks for all loudspeakers, so it was easier to compare one to the other.
The methods used for setting the highpass filters were as follows. On the powered subs-no filters were used. It was assumed that they had an appropriate HP inside the box. On Ricci’s sealed boxes we used a Rane unit that was set around 15Hz-Ricci correct me on that one.
(We used no filter at all on the sealed cabs, the Rane was only used for Brandon's big ported cab, 15hz/Ricci)
On all of the other boxes (except the Danleys), Ricci’s Behringer DCX2496 processor was used to set the highpass filters. Ricci and I looked the measured responses of the unprocessed boxes and made a determination based on that. I used a 24dB Butterworth filter to keep the “garbage” out of the drivers. Due to the design, I was unable to set a highpass lower than 20Hz..
On the Danleys, the Danley DSP was used as I had already programmed the highpass filters in it for the different boxes-I just used different outputs. There was no eq used for the Danleys except for a little “trick” that I played to extend the highpass filter a little lower than the 20hz limit for the THSPUD. I put a 3dB (BW.3) bump at 20Hz to give a flat response to19Hz, because the filter would normally start rolling off above 20hz, and this would flatten it out to a lower freq.
The only time a low pass filter was introduced was with the TH50. I knew the mains would not be able to keep up and I did not want the extra level we would be sending to the TH50 to “muddy up” the upper bass range. I think I lowpassed it at 80Hz.
The max levels that were recorded were by various SPL meters brought by others and placed in the middle of the room-on the floor. I would say on an average the distance was 6-10; away from the loudspeaker, but the actual distance would be varying-due to the different depths of the various cabinets. There was a bit of discrepancy between them-as much as 4-5dB. But as long as the same one was being read (I didn’t do this), the comparative results should be the same.
The peak reading of the day was 128.7dB unweighted peak, by my Audio Toolbox which was sitting in front of the screen on top of Ricci’s rack. Maybe 6’ or so off the side of the subs. This was with the TH50 and it could have done a good bit more, but I don’t think the room could have taken it, without physical damage.
Brandon stopped the listening test when the 16Hz music track started knocking stuff around."