I probably won't get two mics, but thanks for the tip - good to know for the future
Project update:
I'm going to be too busy over the next couple of weeks (at least) to have enough time at once to move everything I need outside and back on top of doing all the measurements.
I did just take a couple preliminary measurements of the woofer's off axis performance though, to verify viability. They're a bit crude- I used a spectrum analyzer app on my phone. But since each frequency is a separate measurement the microphone's non-linear frequency response isn't a problem. The only potential problem is reflected sounds altering performance (making things look better than they are), but I did my best to minimize this with placement (room and microphone). I didn't measure the tweeter because it is omnidirectional at the considered crossover frequencies and it's known from theory discussed earlier in this thread that the only thing that will affect its frequency response is baffle interaction. As frequency decreases from about 1400hz, attenuation increases until finally a maximum of 6db is lost on axis, probably at ~700hz. Measuring the tweeter in the future outside should give useful information.
Woofer:
These rough measurements I took were at 0 and 45 degrees. I'm just posting the decibel decrease because I didn't bypass the crossover, or use a flat mic. It might look like I went excessively high with the measurements, and for my purpose I did, but this woofer is a topic of discussion online, and its measurements aren't available, so I'm making them.
I'm not sure if I should subtract 1db from all of the following measurements, because in theory, this speaker should be the same level at 0 and 45 degrees at 800hz. I think I should because I did the measurements very close nearfield, about 8 inches away, and when I turned the speaker on the lazy susan the distance increased. Not by a lot, but since the overall distance is small to begin with, I think the increase had a large enough effect to cause a 1db dip. Thoughts?
ATC SB75-150SL midwoofer measurements - decibels down at 45 degrees off-axis. First result measured, second corrected
800hz 1db 0db
1000hz 1db 0db
1200hz 1db 0db
1400hz 2db 1db
1600hz 2db 1db
1800hz 2.5db 1.5db
2000hz 3db 2db
2200hz 5db 4db
2400hz 8db 7db
2600hz 10db 9db
2800hz 8db 7db
3000hz 9db 8db
3200hz 12db 11db
3400hz 12.5db 11.5db
3600hz 10db 9db
3800hz 12db 11db
4000hz 12db 11db
4200hz 11db 10db
4400hz 12db 11db
4600hz 12db 11db
4800hz 11db 10db
5000hz 11db 10db
5500hz 12db 11db
6000hz 14db 13db
6500hz 12db 11db
7000hz 12db 11db
8000hz 16db 15db
10000hz 14db 13db
From this, performance can be estimated for 15, 30, 60, and 75 degrees off axis at different frequencies, and an approximate optimal crossover point for widest dispersion with default 9.84" and possible 14.5" wide baffle, discovered.
It's very interesting how from 3200 to ~7500hz, the -12db point of the woofer remains 45 degrees. I believe the reason for this is at the center of the woofer's cone is the 3 inch dome midrange ATC is famous for, and between it and the rubber surround is a thick, treated, heavily damped paper cone, shaped exactly the same as the waveguide surrounding ATC's mid-dome in their speakers. ATC's dome midrange is always crossed at 380hz and 3.5khz in their designs. It's safe to say that even with steep slopes, a flat output with controlled dispersion is required up to at least 4500hz for good integration with tweeters. The size of the waveguide tells us it focuses sound between 2500 and 4500hz (does nothing above and below). Problems contributing to the dome requiring a waveguide include: having 3 inch voice coil, and having a 3 inch dome. The large voice coil lowers output at higher frequencies much more than smaller coils. 3 inch domes beam beginning around 2500hz. This becomes a problem (narrowing past 45 degrees) around 4500hz.
Better woofer measurements in the future will provide more information.
I found that the actual crossover point of the included passive crossover is ~1.975khz, not 2.1 as published in the manual.
It looks like the woofer's -3db point at 45 degrees is 2200hz. Is it safe to assume then, that for the absolute widest dispersion, using it up to only 1100hz is ideal?
Absolute widest dispersion isn't my target, I'm looking for a balance obviously. But to balance things, I'm trying to find maximums. The tweeter's interaction with the baffle is next - how much output is lost on axis, at what frequencies
I'm going to be too busy over the next couple of weeks (at least) to have enough time at once to move everything I need outside and back on top of doing all the measurements.
I did just take a couple preliminary measurements of the woofer's off axis performance though, to verify viability. They're a bit crude- I used a spectrum analyzer app on my phone. But since each frequency is a separate measurement the microphone's non-linear frequency response isn't a problem. The only potential problem is reflected sounds altering performance (making things look better than they are), but I did my best to minimize this with placement (room and microphone). I didn't measure the tweeter because it is omnidirectional at the considered crossover frequencies and it's known from theory discussed earlier in this thread that the only thing that will affect its frequency response is baffle interaction. As frequency decreases from about 1400hz, attenuation increases until finally a maximum of 6db is lost on axis, probably at ~700hz. Measuring the tweeter in the future outside should give useful information.
Woofer:
These rough measurements I took were at 0 and 45 degrees. I'm just posting the decibel decrease because I didn't bypass the crossover, or use a flat mic. It might look like I went excessively high with the measurements, and for my purpose I did, but this woofer is a topic of discussion online, and its measurements aren't available, so I'm making them.
I'm not sure if I should subtract 1db from all of the following measurements, because in theory, this speaker should be the same level at 0 and 45 degrees at 800hz. I think I should because I did the measurements very close nearfield, about 8 inches away, and when I turned the speaker on the lazy susan the distance increased. Not by a lot, but since the overall distance is small to begin with, I think the increase had a large enough effect to cause a 1db dip. Thoughts?
ATC SB75-150SL midwoofer measurements - decibels down at 45 degrees off-axis. First result measured, second corrected
800hz 1db 0db
1000hz 1db 0db
1200hz 1db 0db
1400hz 2db 1db
1600hz 2db 1db
1800hz 2.5db 1.5db
2000hz 3db 2db
2200hz 5db 4db
2400hz 8db 7db
2600hz 10db 9db
2800hz 8db 7db
3000hz 9db 8db
3200hz 12db 11db
3400hz 12.5db 11.5db
3600hz 10db 9db
3800hz 12db 11db
4000hz 12db 11db
4200hz 11db 10db
4400hz 12db 11db
4600hz 12db 11db
4800hz 11db 10db
5000hz 11db 10db
5500hz 12db 11db
6000hz 14db 13db
6500hz 12db 11db
7000hz 12db 11db
8000hz 16db 15db
10000hz 14db 13db
From this, performance can be estimated for 15, 30, 60, and 75 degrees off axis at different frequencies, and an approximate optimal crossover point for widest dispersion with default 9.84" and possible 14.5" wide baffle, discovered.
It's very interesting how from 3200 to ~7500hz, the -12db point of the woofer remains 45 degrees. I believe the reason for this is at the center of the woofer's cone is the 3 inch dome midrange ATC is famous for, and between it and the rubber surround is a thick, treated, heavily damped paper cone, shaped exactly the same as the waveguide surrounding ATC's mid-dome in their speakers. ATC's dome midrange is always crossed at 380hz and 3.5khz in their designs. It's safe to say that even with steep slopes, a flat output with controlled dispersion is required up to at least 4500hz for good integration with tweeters. The size of the waveguide tells us it focuses sound between 2500 and 4500hz (does nothing above and below). Problems contributing to the dome requiring a waveguide include: having 3 inch voice coil, and having a 3 inch dome. The large voice coil lowers output at higher frequencies much more than smaller coils. 3 inch domes beam beginning around 2500hz. This becomes a problem (narrowing past 45 degrees) around 4500hz.
Better woofer measurements in the future will provide more information.
I found that the actual crossover point of the included passive crossover is ~1.975khz, not 2.1 as published in the manual.
It looks like the woofer's -3db point at 45 degrees is 2200hz. Is it safe to assume then, that for the absolute widest dispersion, using it up to only 1100hz is ideal?
Absolute widest dispersion isn't my target, I'm looking for a balance obviously. But to balance things, I'm trying to find maximums. The tweeter's interaction with the baffle is next - how much output is lost on axis, at what frequencies
Last edited: