AVS Forum banner
1 - 20 of 22 Posts

·
Registered
Joined
·
3 Posts
Discussion Starter · #1 ·
Hi,

I was wondering if anyone could name me industry standard (Or professional) microphone calibration techniques.

I am using an arduino and microphone and i have access to a SPL meter.

I aim just to play a 1 kHz tone out of my speakers but i cant find many documents explaining the method. Does anyone know where i can find it?
 

·
Registered
Joined
·
213 Posts
Hi,

I was wondering if anyone could name me industry standard (Or professional) microphone calibration techniques.

I am using an arduino and microphone and i have access to a SPL meter.

I aim just to play a 1 kHz tone out of my speakers but i cant find many documents explaining the method. Does anyone know where i can find it?
Are you trying to create a calibration file for your microphone to load into software like REW?

What microphone are you using?
 

·
Registered
Joined
·
3 Posts
Discussion Starter · #3 ·
Are you trying to create a calibration file for your microphone to load into software like REW?

What microphone are you using?
No,

I have an adafruit electret microphone. The arduino code currently outputs values from 0-1023 (aka 0v - 5v (2.5v = 512)).

All i want to do is compare them, e.g. 600 = 72dB, 700 = 90dB. etc.
 

·
Registered
Joined
·
6,162 Posts
I aim just to play a 1 kHz tone out of my speakers but i cant find many documents explaining the method.
You can't use tones to calibrate levels because moving the microphone even an inch or two changes the level. It's much better to use pink noise, and better still to use pink noise that contains only midrange frequencies. Hey, you're in luck! Zip file attached. Just loop the MP3 file in a media player to play repeatedly.

--Ethan
 

Attachments

·
Registered
Joined
·
3 Posts
Discussion Starter · #5 ·
I've read about calibration by comparison with a reference microphone. It says to sweep a sine wave?
 

·
Registered
Joined
·
2,630 Posts
I've read about calibration by comparison with a reference microphone. It says to sweep a sine wave?
You don't have a reference microphone. Use Ethan's pink noise and your SPL meter as a reference. You'll be within +/-2dB mid band at worst.
 

·
Registered
Joined
·
12,004 Posts
I have an adafruit electret microphone. The arduino code currently outputs values from 0-1023 (aka 0v - 5v (2.5v = 512)).
All i want to do is compare them, e.g. 600 = 72dB, 700 = 90dB. etc.
The math is not that simple. Your math will be all wrong if you do that, for the following reasons:

1) Most SPL meters are using a C or A-weighting.
2) The adafruit is what? Z-weighted?
3) What is the frequency profile of the adafruit? You have to compensate for that as well...
4) db's are on a log scale, not a linear scale. So you have to convert that as well in the software. http://www.sengpielaudio.com/calculator-db-volt.htm
Your computer software will have to use a weighting conversion-table and two (known) mic responses to accurately compare them; and a log scale for db's.

Research ANSI/IEC certified mics and NIST-traceable acoustical measurement laboratories.

Furthermore, mechanical systems are only predictable within their designed operating range.
5) If the frequencies that your are attempting to capture are WAY outside it's sensitivity/operating range, the noisefloor will corrupt the reading, even with the best software correction values.
6) Or conversely... if the SPL is beyond it's operating range it will saturate the input stage with 0dbFS digital clipping.
All things have their limits after all...

7) and as Ethan already said: the room and mic placement (and speaker placement) and any obstructions and acoustical-anomalies thereof, will completely corrupt the results.
The mics have to be within a few millimeters of each other for best results, and that's just getting started...

Trust me I know what I'm talking about! I've been a software developer since the 90's and I've been into audio for almost as long. You are underestimating the complexity of this task, and you are starting with a lot of unknown variables.
HAVE FUN WITH THAT! hehe ;)
 

·
Registered
Joined
·
12,004 Posts
No,

I have an adafruit electret microphone. The arduino code currently outputs values from 0-1023 (aka 0v - 5v (2.5v = 512)).

All i want to do is compare them, e.g. 600 = 72dB, 700 = 90dB. etc.
Most audio-software reads the raw bits off the sound driver into a buffer, and then applies a windowed FFT using the following formula:

https://en.wikipedia.org/wiki/Fast_Fourier_transform

Then, you can apply the weighting conversions as discussed etc etc etc.

Oh it's barrels of fun alright! God is sick dude... he loves math above all other living beings... so much so that I believe we DO live in a computer simulation. Such a cruel god. But that's another topic for another day... hehe :p
 

Attachments

·
Banned
Joined
·
966 Posts
Most audio-software reads the raw bits off the sound driver into a buffer, and then applies a windowed FFT using the following formula:

https://en.wikipedia.org/wiki/Fast_Fourier_transform

Then, you can apply the weighting conversions as discussed etc etc etc.

Oh it's barrels of fun alright! God is sick dude... he loves math above all other living beings... so much so that I believe we DO live in a computer simulation. Such a cruel god. But that's another topic for another day... hehe :p
I think you might be overwhelming this gent. Especially because none of your comments are relevant or true, pertaining to FFT implementation (or the other thing) in MOST audio software - simply not true - and regardless of its implementation, the linear/non-linear transfer function qualities of the microphone are not predicated upon it (FFT) whatsoever, it's a post acquisition implementation, as is weighting for that matter.

This gent is looking for SIMPLE reference signals to improve the accuracy of what he has, I applaud him for doing so. He has no intention of using REW or any such software at this time. Ethan's approach and file, will meet his current and future need. Which will in turn improve any post signal acquisition processing inclusive of FFT analysis, and applied weighting. However, he seems to want to simply, see the SPL at various frequencies, for his own undisclosed purposes. Here's a link to a website that will permit him to create and download waveforms files.

http://www.audiocheck.net/audiofrequencysignalgenerator_sinetone.php

To me, it seem that he's trying to calibrate by comparison, which certainly isn't unheard off, in any field, even the so-call professionals use this technique when forced to.
 

·
Registered
Joined
·
2,630 Posts
I think you might be overwhelming this gent.
He asked, "I was wondering if anyone could name me industry standard (Or professional) microphone calibration techniques." Sometimes the answers can be overwhelming, but that also prompts a restatement of the question.
Especially because none of your comments are relevant or true, pertaining to FFT implementation (or the other thing) in MOST audio software - simply not true - and regardless of its implementation, the linear/non-linear transfer function qualities of the microphone are not predicated upon it (FFT) whatsoever, it's a post acquisition implementation, as is weighting for that matter.
Hmmm....Let's see.
1) Most SPL meters are using a C or A-weighting.
True.
2) The adafruit is what? Z-weighted?
No reference in the data on that mic to any weighting at all, and those little capsules are usually pretty darn flat, so we can go with Z-weighted. Relevant and True.
3) What is the frequency profile of the adafruit? You have to compensate for that as well...
An extension of 2. A completely flat, unweighted mic will produce errors in SPL in many cases. Depending on the end application there might need to be a weighting filter in there somewhere. Relevant and True.
4) db's are on a log scale, not a linear scale. So you have to convert that as well in the software. Your computer software will have to use a weighting conversion-table and two (known) mic responses to accurately compare them; and a log scale for db's.
No argument there. Relevant and True.
5) If the frequencies that your are attempting to capture are WAY outside it's sensitivity/operating range, the noise floor will corrupt the reading, even with the best software correction values.
6) Or conversely... if the SPL is beyond it's operating range it will saturate the input stage with 0dbFS digital clipping.
All things have their limits after all...
Both Relevant and True, as we who have done system calibrations would know.
7) and as Ethan already said: the room and mic placement (and speaker placement) and any obstructions and acoustical-anomalies thereof, will completely corrupt the results.
The mics have to be within a few millimeters of each other for best results, and that's just getting started...
Yup, also Relevant and True.
This gent is looking for SIMPLE reference signals to improve the accuracy of what he has, I applaud him for doing so. He has no intention of using REW or any such software at this time. Ethan's approach and file, will meet his current and future need.
Yes, Ethan's file is band-limited pink, good for rough in-room calibration. But there's no predicting future need.

To me, it seem that he's trying to calibrate by comparison, which certainly isn't unheard off, in any field, even the so-call professionals use this technique when forced to.
Perhaps the so-called pros would do this, but real-pros can't afford to cal anything that way and then bill for it. Quick sanity checks, sure, but not real calibrations.
Trust me I know what I'm talking about!
It's pretty clear here who does.
 

·
Banned
Joined
·
966 Posts
He asked, "I was wondering if anyone could name me industry standard (Or professional) microphone calibration techniques." Sometimes the answers can be overwhelming, but that also prompts a restatement of the question.

Hmmm....Let's see.
True.
No reference in the data on that mic to any weighting at all, and those little capsules are usually pretty darn flat, so we can go with Z-weighted. Relevant and True.
An extension of 2. A completely flat, unweighted mic will produce errors in SPL in many cases. Depending on the end application there might need to be a weighting filter in there somewhere. Relevant and True.
No argument there. Relevant and True.
Both Relevant and True, as we who have done system calibrations would know.
Yup, also Relevant and True.
Yes, Ethan's file is band-limited pink, good for rough in-room calibration. But there's no predicting future need.


Perhaps the so-called pros would do this, but real-pros can't afford to cal anything that way and then bill for it. Quick sanity checks, sure, but not real calibrations.

It's pretty clear here who does.
Hello:

Interesting reply; however, the OP stated what he stated and asked what he asked. If you had any experience with the project devices that he mentioned, you'd know what he was after, and that none of the comments but my own are on point.

Making the rest a pile of hay, no matter how you spin them.:)
 

·
Registered
Joined
·
2,630 Posts
Hello:

Interesting reply; however, the OP stated what he stated and asked what he asked.
Which, you seem to forget, was: "I was wondering if anyone could name me industry standard (Or professional) microphone calibration techniques. (answered)

I am using an arduino and microphone and i have access to a SPL meter.

I aim just to play a 1 kHz tone out of my speakers but i cant find many documents explaining the method. Does anyone know where i can find it? (answered)
If you had any experience with the project devices that he mentioned, you'd know what he was after,
You mean an Arduino (yes) a microphone (yes, quite a bit) and an SPL meter (yes, quite a bit)
and that none of the comments but my own are on point.
As usual...nope. You claimed that comments by BassThatHz were somehow not relevant "none of your comments are relevant or true, " All I did was point out that your claims were wrong, line by line.

But suit yourself, if you want to be the only "right" one, fine. Stamp feet, be insulting, whatever. Knock yourself out. Maybe that's how it worked in Academia.
 

·
Banned
Joined
·
966 Posts
Which, you seem to forget, was: "I was wondering if anyone could name me industry standard (Or professional) microphone calibration techniques. (answered)

I am using an arduino and microphone and i have access to a SPL meter.

I aim just to play a 1 kHz tone out of my speakers but i cant find many documents explaining the method. Does anyone know where i can find it? (answered)
You mean an Arduino (yes) a microphone (yes, quite a bit) and an SPL meter (yes, quite a bit)

As usual...nope. You claimed that comments by BassThatHz were somehow not relevant "none of your comments are relevant or true, " All I did was point out that your claims were wrong, line by line.

But suit yourself, if you want to be the only "right" one, fine. Stamp feet, be insulting, whatever. Knock yourself out. Maybe that's how it worked in Academia.

Again, Interesting reply; however, I must restate that the OP stated what he stated and asked what he asked. If you had any experience with the project devices that he mentioned, you'd know what he was after, and that none of the comments but my own are on point.

Last I checked, only the likes of Rumpelstiltskin can spin straw into gold.:)
 

·
Registered
Joined
·
2,630 Posts
Again, Interesting reply; however, I must restate that the OP stated what he stated and asked what he asked. If you had any experience with the project devices that he mentioned, you'd know what he was after, and that none of the comments but my own are on point.
Repetition and circularity doesn't make it right. I've already outlined where your comments failed, and that I do have experience with everything listed. Apparently, more than you do.

Oh well, and so it goes. It's Halcyon .. and on...and on....
 

·
Banned
Joined
·
966 Posts
Repetition and circularity doesn't make it right. I've already outlined where your comments failed, and that I do have experience with everything listed. Apparently, more than you do.

Oh well, and so it goes. It's Halcyon .. and on...and on....
It doesn't make it wrong apparently either!:confused:

What you've done in here, is simply illustrated your tenacity towards handwaving.

And you have yet to successfully prove that you know more than me, on any topic that I have commented on (you have in fact conceded that you do not), however, I will concede that your efforts to do so, have been much greater than my efforts to hold the line.

But as you said, Halcyon...:)
 

·
Registered
Joined
·
5,896 Posts
Handwaiving. :) That one made me chuckle, thanks!
 

·
Registered
Joined
·
2,630 Posts
And you have yet to successfully prove that you know more than me, on any topic that I have commented on
What are you, 10? Or in high school practicing for the debate team? "I know more that you so I win." Seriously? So this is a p***ing match? And who's judging the contest, you?

Thanks for clearing that up, the insight was most enlightening.

For some reason I thought you worked for a university. Huh. Don't know why I thought that, guess I was wrong, sorry.
(you have in fact conceded that you do not)
I have conceded nothing. I have, on occasion, chosen not to continue to argue with someone whose rich fantasy world has them positioned in it as the omnipotent supreme being who creates his own rules and science.

A choice which, again, seems imperative.

Oh, BTW,that means "unsub" and "ignore".
 
  • Like
Reactions: Bigus

·
Banned
Joined
·
966 Posts
To all desiring to calibrate their microphones for general SPL measurements:

I have found this little gem online for a fantastic price, IMO/E

http://www.ebay.com/itm/Sound-Level...ion-94dB-114-dB-Mics-Microphone-/221882096371

No FFT implementation, just saying.:)
Well, for the price I thought I'd be a Ginny-pig and order one of these calibrators. Total price to my door, just under $160.00.



I have planed to make some before and after cal measurements of one of my mics. When I complete the task, I'll post the results.
 

Attachments

1 - 20 of 22 Posts
Top