The No.1 Website for Pro Audio
Microphone distortion measurements
Old 9th November 2008
  #1
Lives for gear
 

Microphone distortion measurements

Hi!

This thread will be updated later on but in the meantime I'd like to share some of our first findings on mic distortion.

The following mics were measured:

Earthworks QTC1
Sennheiser MKH8020
Sennheiser MKH8040
Line Audio CM3
Behringer ECM8000

We measured intermodulation distortion using 9+10kHz and 19+20kHz at aprox. 80dB SPL and for this test only focused on the difference tone/distortion component at 1kHz.

The ECM800 had about 15-20dB higher levels of distortion than the others.

The rest of the gang had very similar performance which came as something of a surprise.

I don't have the notes here (will add more info after a follow up session of mesurements on several different SPL's and also lower in frequency), that said the 1kHz product was about 80dB down at 9+10kHz and 75dB down at 19+20kHz.

It's interesting to see the CM3 having similar performance as mics costing +10 times more.

Also surprising that the claims from Sennheiser about lower IMD than other mics due to the symmetrical capsule at least don't hold at these levels. We will see if they keep this good behaviour to a higher degree than the other mics when pressing them with higher SPL's.


To be continued..


/Peter
Old 9th November 2008
  #2
DAH
Lives for gear
 
DAH's Avatar
Im justcurious abot the procedure- could u share some light on it?
Old 10th November 2008
  #3
Lives for gear
 

Hi!

The test tones were fed into two tweeters, one sine into each. This method avoids any intermodulation from DAC, power amp or tweeters.

The signal was captured by the mic using Audiofire4 as preamp and AD. The excellent analysing software ARTA was used in spectrum analyse mode to see what was going on.

There's a possibility that a preamp and AD adds intermodulationdistortion that interfer with the measurement but AF4 is a low distortion device and is much more linear than the mics at these levels. We also checked by turning up and down the gain of the preamp with no difference in distortion reading which is a clear indication of the readings being ok with only mic distortion.

The set up is okay and gives relatively correct results but we will arrange with a suitable test baffle with two small-flanged-tweeters so we can exclude combfiltering from various reflections. When performing the measurements we had a piece of fiberglass on the floor taming the floor reflection.

Since we used steady state stimuli (sine waves) reflections does not mock up the readings as long as you can get the actual SPL at the mic to be the same but that prooved a little hard at 19+20kHz due to the short wavelength.


/Peter
Attached Thumbnails
Microphone distortion measurements-img_0651.jpg  
Old 10th November 2008
  #4
Lives for gear
 

Here is a spectrum view performing IMD measurment on the Echo Audiofire4 using the software RMAA.

A 19kHz and 20kHz sinewave handled by a non-linear system will result in sum and difference tones at various frequencies and can be seen at 1kHz, 18kHz and 19kHz for example. The peaks at 38kHz and 40kHz is the 2nd order harmonic distortion and also sum and difference tones (IMD) surrounding the pure harmonic tones at 38kHz and 40kHz.

When we calculated the distortion during this test we started with the 19+20kHz rms levels and subtracted the 1kHz level from that. We did not include other orders of IMD than the 1kHz one.

The meaning with this test was just to get started and to get a relative grip on the distortion and most of all to see how the distortion compared between different microphone designs.


/Peter
Attached Thumbnails
Microphone distortion measurements-imd.png  
Old 10th November 2008
  #5
Lives for gear
 
nosebleedaudio's Avatar
 

Peter,
What kind of numbers are you getting?
Is that red material absorption?
Put a 2" piece of 703 between them to reduce reflection off the cabinets.
Old 10th November 2008
  #6
Lives for gear
 

Mike,

we used fiberglass at the floor when doing the measurements. The red you see is my friends ugly carpet in his livingroom.. heh


The speakers were angled to avoid reflections but at the same time get the tweeters close to on axis towards the capsules.

Small movements of the mics made the levels of the respective sines move up and down so we used the average between the two sine tones and counted from there down to the difference tone at 1kHz.

The 1kHz product was between -72dB and -84dB down on all mics except for the Behringer. In general (but with some exceptions) the distortion was higher at 19+20kHz than at 9+10kHz.

For the next test round we will use a common baffle for two tweeters and mount them close so that both tweeters are on axis towards the microphones. The baffle and mics will be well away from any objects or surfaces so reflectionfree "free field" results can be had. Again, this is not criticall for this kind of measurement but better precision will be the result I guess, especially when comparing the mics.


This measurement round was not very precise but gave what we felt was decent reults within a couple of dB.


One thing we learned was that you can not use just any mic to measure loudspeaker distortion since what you see may be mic distortion + speaker distortion. A QTC1 or similar seems to be safe though since few loudspeakers has distortion that low.


/Peter
Old 1st December 2008
  #7
Gear Guru
 
John Willett's Avatar
 

Exclamation

Are you just measuring around the 1kHz mark?

Most mics are pretty good here - the major discrepancies are in the high frequencies.

This picture shows the IM distortion of various microphones measured by Sennheiser in the 1980s (when the MKH 40 was released - the blue line is the MKH 40. This diagram was published in various brochures and publications over the years.

Click on the picture to show the high quality version.



It would not be fair of me to name the other mics, but they are all top quality ones. The test was two frequencies (I think 70Hz apart) and these were swept from 20Hz to 20kHz keeping the same spacing between the two frequencies. The test was done in an anechoic chamber with the two different frequencies coming from two generators via two separate amplifiers into two separate loudspeakers to remove any possibility of any IM distortion being introduced by the amplifiers or loudspeakers.

But you will see that all are good around the 1kHz mark.

The old Sennheiser MKH CD (1985 vintage) has results you can listen to - I still have a copy I can burn copies of if you want to listen.

But these tests are very useful as IM distortion, although very low, does muddy and mask the music even though you may not particularly be aware of it.

Good luck with your new tests.
Old 1st December 2008
  #8
Lives for gear
 

Great, thanks for input John!


The numbers I mentioned above is representative for HF distortion.

How shall I put it.. hmm, we did not meaure 1kHz distortion with a 1kHz sine wave or anything thereabout but we used a similar method as the one you describe.

The difference being that we only tested at a few fixed high frequencies instead of sweeping. Also instead of 70Hz apart we used two sines 1kHz apart.

A flaw one could argue in our preliminary test is that we did not calculate total distortion for the two tone stimuli but restricted us to looking at the 1kHz product.

But we did not state that the total distortion was 80dB or so down, but that the 1kHz difference tone was around that level.

So, my/our results would be the equivalent of the 10kHz and 20kHz readings in the curves of your picture, not the 1kHz readings.

Hope I did not caused confusion.. :-)

Oh, the relatively high distortion from all mics in your graphs indicates that the Sennheiser test was done at much higher spl than our test and as I mentioned earlier, it's going to be interesting to see if the MKH mics performs better than the others at these higher levels.

Also we suspected that the distortion would be higher towards the upper end of the audible spectrum and therefore we started out with only HF measurements for this first round.


/Peter
Old 1st December 2008
  #9
Gear Guru
 
John Willett's Avatar
 

Smile

Quote:
Originally Posted by Audiop View Post
Oh, the relatively high distortion from all mics in your graphs indicates that the Sennheiser test was done at much higher spl than our test and as I mentioned earlier, it's going to be interesting to see if the MKH mics performs better than the others at these higher levels.
Ahhh - just found the figures - The Sennheiser test was done at 104dBSPL and it *was* 70Hz apart (I mentioned it in my 1998 AES paper).
Old 10th December 2008
  #10
Gear Maniac
 

If you really want something useful from these tests, you need a more complex test signal, such as the Jon Risch test set - see Jon_Risch's Web Site (he's the first guy to be able to measure the difference between speaker single and bi-wiring).
Old 10th December 2008
  #11
Lives for gear
 

Quote:
Originally Posted by TimmyP1955 View Post
If you really want something useful from these tests, you need a more complex test signal,
Not at all. Two tone stimuli tells us a lot about the nonlinearity without letting playback equipment interfer with the readings. Make sure you read carefully and understand this tests, it seems that most people miss some basic things and misunderstand this.

Quote:
such as the Jon Risch test set - see Jon_Risch's Web Site (he's the first guy to be able to measure the difference between speaker single and bi-wiring).
When I was discussing with Jon back at madisound board when it was most active he never mentioned anything about such a test but people I know looked into it and nailed down effects that make it out into the room decades ago.

I don't mind critique but I prefer some substance and not "that is wrong" kind of statements.


/Peter
Old 10th December 2008
  #12
Lives for gear
 

Timmy,

I did some reading on Jon's site. The signal you propose can not be used in this test and would render the results/readings totally useless.


/Peter
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Forum Jump
Forum Jump