Home Theater Forum and Systems banner

1 - 6 of 6 Posts

·
Registered
Joined
·
15 Posts
Discussion Starter #1
I have built a DIY microphone calibrator.

With that small construction, I can check the linearity of SPL microphones.



I have used that construction to create a calibration map of my SPL:

Left I had a Bruel&Kjaer class 1 SPL, right my cheapo SPL.

The pistonphone actuator is actually just a plain earphone. It is imperfect, but that does not matter for the comparison, since I sent for each frequency and each level just enough signal to get the required level on the Bruel&Kjaer class 1 SPL.
Now I have got a map of my own SPL. WOW! that one wasn't really linear at all.
But that does not matter either: I can inject the signals according to my map to recreate exactly what the professional SPL heard for each level an frequency.

That works well, but is not suitable to evaluate quickly a mic using white or pink noise.
For that, I would need a better earphone.

Who knows a good affordable earphone that is approximately linear?

Thank you for your input.
 

·
Plain ole user
Joined
·
11,121 Posts
I'll defer to our resident headphone guru, Audiocraver. You might look for some of his reviews.
 

·
Premium Member
Joined
·
1,562 Posts
I don't follow why that is needed.
For the same reason as you stated above, the only need is to able to record the differences at each freq between the reference SLM and the test Mic.

Use REW or other audio analysis program to plot the response of both units using a sweep or PinkPN signal. The difference in response is the cal info needed. REW and others include 'A / B' math.

Maybe the B&K SLM doesn't have a line out? The mic sure will.

Sorry if I misunderstood the problem.
 

·
Registered
Joined
·
15 Posts
Discussion Starter #4
I don't follow why that is needed.
For the same reason as you stated above, the only need is to able to record the differences at each freq between the reference SLM and the test Mic.

Use REW or other audio analysis program to plot the response of both units using a sweep or PinkPN signal. The difference in response is the cal info needed. REW and others include 'A / B' math.

Maybe the B&K SLM doesn't have a line out? The mic sure will.

Sorry if I misunderstood the problem.
No, you did not misunderstand the problem.
I just have got a lot of additional problems.

REW can fix some non-linearities but no real hard resonances of 40dB and more...

First of all my construction with the T-pipe was inadequate.
With white noise one could hear a lot of resonances, a bit like in a sewer pipe!.
So I had to use another construction and have the actuator directly against the microphone.

The second lesson is that earphones are inadequate as actuators as well.
Earphones are conceived to be psychoacoustically correct for most ears, and that is everything but not linear!

So I made another attempt with a Shure 1/2 inch dynamic microphone capsule.
The response was much cleaner, with no hard resonances, but it was not linear.

REW could fix it, but not above 5KHz.

Indeed at that frequency the wavelengths are so short, that the pistonphone construction becomes physically inadequate.

That means the only solution would be to go free field in an anechoic chamber, with a suitable mid-power actuator...
$$$ :sad:
 

·
Premium Member
Joined
·
1,562 Posts
No fixture is really needed. Most any inexpensive measurement mic is pretty flat from 50Hz to 1.5kHz or so. There is usually a rise of 3-6dB with the peak somewhere around 10kHz as you are probably aware.

It is easy to just use a dome TW with the mic at 2-5mm for the actuator. I did that using my SEAS TW to compare the response of 4 different calibrations of 3 mics (2 different calibration labs) to see how close they agreed. In my case I didn't have a reference mic so I was just looking to use the average so they all would agree in case my favorite mic failed. Now they all agree to a fraction of a dB. We can even do 90° calibrations accurately that way.

Since you have reference B&K SLM, you can use it to calibrate the other mic. The bass range is a little more difficult, but the same concept can be used there. The lower midrange is another story as the room can be an issue for high accuracy, I don't remember how well that worked, but again the measurement mics should all be pretty flat in that range. I only remember I didn't have an issue there.

There was several dB discrepancy of my mic calibrations at the low bass as that is the range that is difficult for most all calibration services. The good thing is that 3-4dB error in very low bass is not a sound quality issue.

This process takes a lot more work and time than the coupler used in calibration labs so it can't be used by those services. I find the results to be much more repeatable however so if you have a good reference mic and want to spend the time you can match that calibration very closely.
 

·
Registered
Joined
·
15 Posts
Discussion Starter #6 (Edited)
Up to 4 Khz the pistonphone technique (dynamic microphone as an actuator) at 1mm from test microphone connected by a flexible sleeve works well.
I just made a REW amplifier calibration with the combination above and can check the linearity of the microphones.

By the way I just found that the t.bone XLR to USB adapter with Phantom power is absolutely execrable:
That one: http://www.thomann.de/gb/the_tbone_micplug_usb.htm
aaaargh! http://www.cjoint.com/doc/15_11/EKCmokkSO5q_T-Bone-USB.jpg
Who could belive that one can sell THAT in the 21st century?

It was not the mic. The same mic on my mixer-table performs linear to 2 dB.

Now looking for a usable XLR to USB converter...
A mic with integrated USB like the UMIK is not an option: I have at least 20 meter cable in between. USB does not allow that length...
Using a mixer is too unconvenient.

P.S: I don't own a BK SPL, i just have a chap at the university, who can let me use one sometimes.
 
1 - 6 of 6 Posts
Top