how to take db measurement
- Thread starter sonu
- Start date
The answer would take a book. In fact, it has: Joe d'Appolito's "Testing Loudpeakers."
A pink noise measurement will give you a mixture of direct and reflected sound, unless you do it in an anechoic chamber. If you just want the speaker's response, you'd do better by using some sort of impulse or MLS testing, where you can truncate the time domain response before the reflections. In-room response will do a fairly good job of telling you what the basic spectral balance is, but it's very easy to misinterpret.
Subs can best be measured (for anechoic response) near-field, but if it's a bass reflex, combining the port and driver outputs is a little tricky. d'Appolito explains the procedure quite well.
A pink noise measurement will give you a mixture of direct and reflected sound, unless you do it in an anechoic chamber. If you just want the speaker's response, you'd do better by using some sort of impulse or MLS testing, where you can truncate the time domain response before the reflections. In-room response will do a fairly good job of telling you what the basic spectral balance is, but it's very easy to misinterpret.
Subs can best be measured (for anechoic response) near-field, but if it's a bass reflex, combining the port and driver outputs is a little tricky. d'Appolito explains the procedure quite well.
Measuring DB? Feel Free To Correct Me Here...
Well....
Maybe this related topic will help?
For measuring the response of my microphones, I use a reference mic of which I already know the response.
I can then tune my preamps and reference speakers to the exact intensity I need for correct measurement with other microphones assuming I use exactly the same mic placement.
Perhaps you could use a reference mic with known response characteristics to set up a testing area with mic stands in a fixed place, etc.
Even if you don't have an anechoic chamber, you can use the frequency response of the microphone to test signals and the charted frequency response of your speakers to determine what effect your room is having on the testing procedure.
It is not an exact science, but it might afford you an approximation of speaker performance until you can afford to build your ISO-9002 certified acoustics lab.
Food for thought maybe?
Well....
Maybe this related topic will help?
For measuring the response of my microphones, I use a reference mic of which I already know the response.
I can then tune my preamps and reference speakers to the exact intensity I need for correct measurement with other microphones assuming I use exactly the same mic placement.
Perhaps you could use a reference mic with known response characteristics to set up a testing area with mic stands in a fixed place, etc.
Even if you don't have an anechoic chamber, you can use the frequency response of the microphone to test signals and the charted frequency response of your speakers to determine what effect your room is having on the testing procedure.
It is not an exact science, but it might afford you an approximation of speaker performance until you can afford to build your ISO-9002 certified acoustics lab.
Food for thought maybe?
better hop in on this discussion while I'm at it. 
I have also ordered a digital SPLmeter (from Germany), and the confirmation email that my bank transfer has been recieved should be arriving after this weekend. Then it's waiting for the delivery.
My goal would be to measure the SPL (of course), but also to use Goldwave's sine generator on this PC to send certain frequencies to my to-be-built subwoofer. Then I could compare the backyard SPLcurve to the in-room curve and determine the room gain.
The predicted anechoic curve should be -3dB at 30Hz, so if I don't see this in my backyard, maybe I could use this info to make correction curves for the SPLmeter.
(I'll ask a friend of mine who studies civil engineering, if he can't help me with some equipment in a lab or so.)
Well ... how about measuring the cuve of my speakers?
- As long as there is no room gain, the reflected soundwaves shouldn't interfere too much with the signal.
- Phase differences between orig en reflect sound are more important with lower freq, so high freq should still be accurate enough to measure.
- If one would go for the pulsed-response, the difference between "fast" and "slow" peak-SPL-measurement would be insufficient to filter out the echo. (Unless one has a room longer than, ... lets say 340 meters? )
- If I can get a good correction curve with the help of my friend and the university lab, then I could calibrate my el-cheapo microphone and use a freeware measurement program on my PC to generate tones/pulses and use the pulse-technique to eliminate the echo's and reflections. This last one would be the best option I think.
Anyone other idea's?
I have also ordered a digital SPLmeter (from Germany), and the confirmation email that my bank transfer has been recieved should be arriving after this weekend. Then it's waiting for the delivery.
My goal would be to measure the SPL (of course), but also to use Goldwave's sine generator on this PC to send certain frequencies to my to-be-built subwoofer. Then I could compare the backyard SPLcurve to the in-room curve and determine the room gain.
The predicted anechoic curve should be -3dB at 30Hz, so if I don't see this in my backyard, maybe I could use this info to make correction curves for the SPLmeter.
(I'll ask a friend of mine who studies civil engineering, if he can't help me with some equipment in a lab or so.)
Well ... how about measuring the cuve of my speakers?
- As long as there is no room gain, the reflected soundwaves shouldn't interfere too much with the signal.
- Phase differences between orig en reflect sound are more important with lower freq, so high freq should still be accurate enough to measure.
- If one would go for the pulsed-response, the difference between "fast" and "slow" peak-SPL-measurement would be insufficient to filter out the echo. (Unless one has a room longer than, ... lets say 340 meters?
)- If I can get a good correction curve with the help of my friend and the university lab, then I could calibrate my el-cheapo microphone and use a freeware measurement program on my PC to generate tones/pulses and use the pulse-technique to eliminate the echo's and reflections. This last one would be the best option I think.
Anyone other idea's?