[Top] [All Lists]

Re: testing mic self-noise

Subject: Re: testing mic self-noise
From: Rob Danielson <>
Date: Tue, 3 Jun 2003 14:47:23 -0500
>At 06:55 AM 6/3/2003, Rob Danielson wrote:
>>Hi Caesar et al--
>>I thought it might be useful to follow-up on your original question
>>about ways to test mic self noise. I ran my test using  a 35dB bed of
>>urban night presence (with neighbor's fountain) at 2 a.m. as a
>>source. I compared a Sennheiser mkh-40 with a factory claimed 12dB(A)
>>of self noise and a Rode NT-3 which is spec'd at 17dB(A) by running
>>both thru a MP-2 mic preamp and making 24bit/48K mono sound files
>>through the converters in a RME Hammerfall mutliface.   The NT3 was
>>powered by the mp2, not by the 9 volt battery option. The mp2 was
>>cranked to full gain for both mics.
>>I studied a section of the sound files where the bed was the
>>steadiest. The sound file created by the MKH-40 has a peak of -14.7dB
>>whereas the Rode's sound file over the same time period peaks at
>  >-22.9dB. The published sensitivity of the MKH-40 is 25 mV/Pa while
>  >the Rode's is stated as 12 mV/Pa.
>>I normalized both sound files to -15dB and truncated bit depth to 16
>>in the same step. I used a file header change in Soundhack to change
>>the sample rate from 48K to 44.1K so there's identical processing and
>>no sample rate processing involved. I linked them into a single
>>stereo .wav file with the MKH-40 on the left (top) ($1100) and the
>>Rode NT-3 ($160) on the right (bottom).
>>(1.9 mb)
>>Remember,  both mics are running through a $650 mic pre and through a
>>high quality A->D converter with no compression. From a rough test I
>>ran on my Sharp MT-90 MD recorder, the quality is significantly less
>>(more hiss) when taking an NT-3 directly into the Sharp's mic input
>>at FULL GAIN and through its A-D convertors and compression. I
>>couldn't do this later test which would give us more info about the
>>popular,  "should I get a better mic or a good pre" question.  I
>>don't have digital out for my MD discs and my two DAT decks recently
>  >departed.
>I'm still uncertain just what you demonstrated here, Rob. I looked at the
>sound way up at 14,450Hz and find the Rode did a much better job of
>capturing that faint sound, as well as the sound of the fountain. Is this
>because of the 12 vs 24 mV sensitivity?
>Doug Von Gausig
>Clarkdale, Arizona, USA
>Nature Recordists e-mail group

Hi Doug-

Thanks for listening to such an unsavory recording and sharing your reactions.

The wav file plays loud so its easy to forget that the test pushes 
the limits of the mics and pres like nature recordists. You are 
listening to a highly amplified signal; 35dB is quiet for an urban 
environment. The tiny fountain is about 80' away. Yes, there's more 
noise with the Rode, but not a lot as the combined noise and 
sensitivity specs would suggest. As you point out, the Rode's sound 
file is brighter. That's really a surprise--not only because one 
expects more high end with higher $  but the Rode is unidirectional 
and the mkh-4o is cardioid and both were pointed at the same 
spot--away from the fountain!  Of course, the Rode's brightness comes 
along with the extra noise hiss. It also occurred to me that the 
Sennheiser @25 mV/Pa sensitivity should have produced an original 
file (pre-normalized) that was lot greater than 8dB more saturated. 
The Rode's sensitivity at  12 mV/Pa is 15 dB less. Does this mean 
that a $160 Rode is as good as a state of the art field mic? Or,.. 
that the specs are suspect?

I guess its another confirmation for me that responses of a 
recording, when the gain goes way up, are very complicated.  Could 
the MP2 mic preamp be leveling the playing field that much? I've 
heard the Rode used a lot with MD recorders because that what my 
students use and  believe me, the nt3 is a lot noisier in that 
application. I've heard from experienced recordists that preamp 
quality is very important when gain goes up; maybe there's more to 
that with high gain than we currently appreciate. A way to further 
test the effects of the pre would be to run the Rode through a lesser 
quality pre like that of an MD and and then an MP2 back to back. If I 
had a digital out, I'd try it. Should the noise with the MD pre jump 
way out, then that would make a  pretty good case for buying a better 
pre before dumping a lot into a big bucks mic if you're not planning 
on buying a high-end recorder and you plan to use high gain often. 
BUT, If you only record hot subjects with an MD pre, maybe the Rode 
would produce 90% of what you need all alone. The Nt3 specs are very 
similar to the stereo version  people have been posting about. Rob


<Prev in Thread] Current Thread [Next in Thread>

The University of NSW School of Computer and Engineering takes no responsibility for the contents of this archive. It is purely a compilation of material sent by many people to the naturerecordists mailing list. It has not been checked for accuracy nor its content verified in any way. If you wish to get material removed from the archive or have other queries about the archive e-mail Andrew Taylor at this address: andrewt@cse.unsw.EDU.AU