precise analysis of microphones? at DVinfo.net

Dennis Stevens · January 11th, 2007, 08:48 AM

I got my AT4053a mic the other day, and I thought I'd see how different it sounded from my AT835b shotgun when recording indoors.

Yes, of course, it sounded different, it's a different mic! But I wanted to see how much, as it was a good amount of $ for me.

Anyway, I recorded myself reading a page from a book into both mics. I did with twice with each mic, with the filter off and the filter on. I had the both mics above, pointing down at a point in front of my mouth.

Well, I heard the difference, but then I opened up the .avi files in Adobe Audition 2.0. I looked at it in spectral analsyis to see if I could see a big difference there. I did, but I also adjusted the volume between mics. So maybe that's all I'm seeing. (I'm no sound engineer, obviously).

Just wondering is there a way to analyze the recording in Audition and sort of see how and where they are different? Would I see different/more frequencies being used?

All kind of academic, but I thought it would be cool to do. Plus it helps to justify to my gang of fellow film fiends that it was a worthwhile purchase.

Just curious.........

Abe Dolinger · January 11th, 2007, 09:22 AM

It's going to be tough to get a scientific reading using that particular method. I can't claim to know the best way, but what I would do is generate a 0hz to 25khz sine wave sweep in Audition over a minute or so (a fairly long period of time). Then mount the mic in question on a stand, pointing directly at the center of the speaker, about 2 feet away. Comparing the spectral and waveform data should yield some small differences - you'd want to check the waveform to see which one has a flatter frequency response (constant dB over all frequencies). Then you could change the direction of the mics to 90 degrees and 180 degrees away from the speaker to check off-axis rejection. Then if you wanted to be completely anal about it, you could do these tests again at different distances.

I wonder if anyone's made a program that would compare the results . . I could make a really clumsy one that would only work with Audition, but it would probably take some time. I'm a terrible programmer.

edit: The best test is probably going to be your ears, though!

Bill Ravens · January 11th, 2007, 10:06 AM

Two comments here:
1-Microphones are like musical instruments. Even tho the manufacturer shows optimistic frequency response curves for a good microphone, and claims the response is flat, there's a TON of harmonics and resonances that make each microphone unique. To make things even more confusing, the room ambience, electronic path and speakers used all add viariables to the analysis of a mic. the most "scientific" way to evaluate a mic is to look at a known signal on a waveform monitor. an even more scientific method is to display the waveform as a histogram...that is doing a fast fourier transform on the waveform and looking at the statistical occurrences of frequency, taking time out of the display. OK, if you're with me so far, what I'm saying is you have to look at a reference signal, say a 1000KZ pure sine wave played thru the mic, on a histogram. The histogram will show a big spike at the 1000Hz reference position, and it will show a whole bunch of smaller spikes at other frequencies, even tho' the input signal was a pure 1000 Hz. These other small spikes are what makes a mic unique.

2-I've been using software in my audio studio called Harbal (www.har-bal.com) Harbal will do the FFT for you and show you the histogram or EQ curve. It's the SHAPE of this curve that you want to be interested in. By recording a pure sine wave in your recording environment, played thru your system speakers, then looking at the EQ curve in HARBAL, you can get a real good idea of the acoustics in your soundroom, from end to end. The original pure sine wave is NEVER a pure sinewave by the time it gets to your ears. The mic, speakers, and room acoustics all add their own signature to the mix. hehehehe, no pun intended.

January 11th, 2007, 08:48 AM	#1
Dennis Stevens Major Player Join Date: May 2006 Location: Boston, Massachusetts Posts: 425	precise analysis of microphones? I got my AT4053a mic the other day, and I thought I'd see how different it sounded from my AT835b shotgun when recording indoors. Yes, of course, it sounded different, it's a different mic! But I wanted to see how much, as it was a good amount of $ for me. Anyway, I recorded myself reading a page from a book into both mics. I did with twice with each mic, with the filter off and the filter on. I had the both mics above, pointing down at a point in front of my mouth. Well, I heard the difference, but then I opened up the .avi files in Adobe Audition 2.0. I looked at it in spectral analsyis to see if I could see a big difference there. I did, but I also adjusted the volume between mics. So maybe that's all I'm seeing. (I'm no sound engineer, obviously). Just wondering is there a way to analyze the recording in Audition and sort of see how and where they are different? Would I see different/more frequencies being used? All kind of academic, but I thought it would be cool to do. Plus it helps to justify to my gang of fellow film fiends that it was a worthwhile purchase. Just curious.........

January 11th, 2007, 09:22 AM	#2
Abe Dolinger Regular Crew Join Date: Apr 2006 Location: Brooklyn, NY Posts: 111	It's going to be tough to get a scientific reading using that particular method. I can't claim to know the best way, but what I would do is generate a 0hz to 25khz sine wave sweep in Audition over a minute or so (a fairly long period of time). Then mount the mic in question on a stand, pointing directly at the center of the speaker, about 2 feet away. Comparing the spectral and waveform data should yield some small differences - you'd want to check the waveform to see which one has a flatter frequency response (constant dB over all frequencies). Then you could change the direction of the mics to 90 degrees and 180 degrees away from the speaker to check off-axis rejection. Then if you wanted to be completely anal about it, you could do these tests again at different distances. I wonder if anyone's made a program that would compare the results . . I could make a really clumsy one that would only work with Audition, but it would probably take some time. I'm a terrible programmer. edit: The best test is probably going to be your ears, though!

January 11th, 2007, 10:06 AM	#3
Bill Ravens Join Date: Nov 2001 Location: santa fe, nm Posts: 3,264 Images: 10	Two comments here: 1-Microphones are like musical instruments. Even tho the manufacturer shows optimistic frequency response curves for a good microphone, and claims the response is flat, there's a TON of harmonics and resonances that make each microphone unique. To make things even more confusing, the room ambience, electronic path and speakers used all add viariables to the analysis of a mic. the most "scientific" way to evaluate a mic is to look at a known signal on a waveform monitor. an even more scientific method is to display the waveform as a histogram...that is doing a fast fourier transform on the waveform and looking at the statistical occurrences of frequency, taking time out of the display. OK, if you're with me so far, what I'm saying is you have to look at a reference signal, say a 1000KZ pure sine wave played thru the mic, on a histogram. The histogram will show a big spike at the 1000Hz reference position, and it will show a whole bunch of smaller spikes at other frequencies, even tho' the input signal was a pure 1000 Hz. These other small spikes are what makes a mic unique. 2-I've been using software in my audio studio called Harbal (www.har-bal.com) Harbal will do the FFT for you and show you the histogram or EQ curve. It's the SHAPE of this curve that you want to be interested in. By recording a pure sine wave in your recording environment, played thru your system speakers, then looking at the EQ curve in HARBAL, you can get a real good idea of the acoustics in your soundroom, from end to end. The original pure sine wave is NEVER a pure sinewave by the time it gets to your ears. The mic, speakers, and room acoustics all add their own signature to the mix. hehehehe, no pun intended.