R/E/P Community

Please login or register.

Login with username, password and session length
Advanced search  

Pages: [1]   Go Down

Author Topic: Km56c performance test  (Read 2977 times)

afterlifestudios

  • Full Member
  • ***
  • Online Online
  • Posts: 102
  • Real Full Name: John Raham
Km56c performance test
« on: September 17, 2016, 04:15:41 am »

I have a KM56c that sounds generally wonderful.  I'd like to know if its self-noise is within spec.  I can use it on quiet sources, but if recording a very quietly finger picked acoustic guitar from a distance of 18 inches, the self noise is noticeable.  Especially compared to one of my KM84s. 

Obviously I realize that the AC701 won't be silent, but I'm just wondering where my specimen sits on the range of noise...

Is there a clever method to test this?


Logged

Jim Williams

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 575
Re: Km56c performance test
« Reply #1 on: September 17, 2016, 11:22:07 am »

The standard method for noise tests are to load the input through a capacitor that matches the capacitance of the capsule. After correct and complete screening a noise sweep can be done with a graphical analyzer like Audio Precision once the 48 volt power is applied. It must be a very clean 48 volts or that will add to the noise results.

Noise vs frequency sweeps not only show the totality of the noise but its sprectral contribution. Spot s/n ratio tests are also available for a summed noise response.
Logged

afterlifestudios

  • Full Member
  • ***
  • Online Online
  • Posts: 102
  • Real Full Name: John Raham
Re: Km56c performance test
« Reply #2 on: September 17, 2016, 12:14:16 pm »

Thanks Jim.  Would I have anything to compare my results to?  Or would I just have some nifty numbers?  (Are there published figures?  Or other users who have done this?)

An Audio Precision analyzer is out of my reach at this point, but I will ask some local technicians...

Also, I'm not sure where the +48v you mention comes into play.  (The KM56 of course has its own PSU and its voltages have been calibrated.)

I suppose A/B comparison to a known excellent specimen would put my mind at ease, but I'm having trouble finding anyone in Vancouver who has one...


Logged

klaus

  • Moderator
  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1664
Re: Km56c performance test
« Reply #3 on: September 17, 2016, 01:34:34 pm »

Yeah, I don't have one of those Audio Precisions either and never missed it in 35 years of testing tube mics for noise. But I have an advantage over the original poster: I usually have access to the same mic that is to be tested.

Here are two simple and quite legitimate ways to test whether your KM56's noise level is ok:

1. If the noise you hear at average recording and monitor levels (no signal input, just a quiet room) is irregular, sputtering, frying, or in any other way sporadic, suspect AC701 tube troubles. Leave the mic on for 24 hours before you make that test.


2. Plug in all of your professional condenser mics, one after another under the above described conditions, and listen for their self noise. You will establish a noise floor range, usually ±4dB or so. Now plug in your KM56 and compare. If its noise floor is comparable, you are OK for now. 

There are a few tweaks that can lower a tube mic's noise to the low end of its design spec, but kind of work is downstream from your initial evaluations described here.  Please report back once you have results.
Logged
Klaus Heyne
German Masterworks®
www.GermanMasterworks.com

afterlifestudios

  • Full Member
  • ***
  • Online Online
  • Posts: 102
  • Real Full Name: John Raham
Re: Km56c performance test
« Reply #4 on: September 17, 2016, 02:10:46 pm »

Thank you, Klaus. 

No problems with #1

Re: #2...  So would all mics be tested with the same preamp gain?   If not, how do I establish a uniform level by which to measure the noise? 

I will be measuring against km84, u67, u87, sm69 etc...  Would the difference in output levels of these mics not interfere with the test method?

Thanks again,
John


Logged

klaus

  • Moderator
  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1664
Re: Km56c performance test
« Reply #5 on: September 17, 2016, 03:33:13 pm »

I should have been more clear about the test method for #2.

Equalize the gain for each mic by either speaking or singing with the same volume (or use a 1kHz or white noise test tone, reproduced through a speaker) positioned about 10" from the mic.

Then set the input gain level on the pre to achieve the same output level out of the pre for each mic under test.

Though the pre's gain setting will vary, its noise when amplifying a bit more for a lower gain mic will still be considerably below any noise emanating from any of the condenser mics.

The comparative mics you mentioned are all in the same relative noise spectrum where a healthy KM56 should be able to compete, ± 3-4dB.

Or, put more simply: properly conducted, this test will reveal whether your KM56 sticks out in its noise performance from the rest, even though some of the other mics are solid state and may be a tiny bit quieter. Yet, the noise floor of a quiet AC701 will be competitive with that of a U67, after proper gain leveling.
Logged
Klaus Heyne
German Masterworks®
www.GermanMasterworks.com

afterlifestudios

  • Full Member
  • ***
  • Online Online
  • Posts: 102
  • Real Full Name: John Raham
Re: Km56c performance test
« Reply #6 on: September 17, 2016, 05:17:04 pm »

Excellent. Thank you, Klaus. I will post my results...
Logged

David Satz

  • Full Member
  • ***
  • Offline Offline
  • Posts: 122
Re: Km56c performance test
« Reply #7 on: September 17, 2016, 07:15:59 pm »

Neumann specified the self-noise of the KM 56c with dummy head attached (a substitute capacitance like that which Jim W. described, in a shielded enclosure that can be attached to the microphone amplifier in place of the capsule head) as -100 dBU RMS unweighted, and about 11 dB lower according to IEC 268-3 with "A" weighting.

This is from Neumann document 11459-918-002.00 "Amplifier Specifications for Microphones" originally dated April 1973, but apparently revised in 1981. It's in the "InfoPool" on Neumann's Web site and can be downloaded via http://www.neumann.com/download.php?download=docu0048.PDF .

--best regards
Logged

afterlifestudios

  • Full Member
  • ***
  • Online Online
  • Posts: 102
  • Real Full Name: John Raham
Re: Km56c performance test
« Reply #8 on: September 17, 2016, 09:10:27 pm »

Excellent. Thank you for the link!
Any suggestions for a suitable capacitor (is that the 2x45pf I see on the chart?) and the ideal place to connect it in place of the kk56?

Logged

David Satz

  • Full Member
  • ***
  • Offline Offline
  • Posts: 122
Re: Km56c performance test
« Reply #9 on: September 18, 2016, 09:53:43 am »

Having posted that link, I want to say that noise measurements are very tricky--problematic even for the best manufacturers, let alone an individual with more limited resources. Shielding around the ultra-high-impedance input of the microphone amplifier is really important, since stray EM fields are everywhere. You can't just plug in a capacitor and expect to get reasonable results, especially if you don't have a spectrum analyzer and filters on hand. Try listening to the output of the preamp with a capacitor attached to the microphone amplifier, and you'll probably hear what I mean--the hum and buzz levels can be enormous.

For the microphones I use the most, I've bought shielded test heads from the manufacturers. Neumann used to sell these as regular catalog items, and Schoeps still does. But even then the EM protection isn't perfect. Even if it were perfect, my noise measurements probably still wouldn't agree with the manufacturers', since the test conditions--e.g. the time averaging of the meters, the frequency weighting networks--are critical but not identical between us.

Even among the best manufacturers, measurements of the same microphone according to the same standards was found to differ by as much as 6 dB (AES SC-04-04 "round robin" test a few years back).

I've put real time and study into this, bought the test gear and learned how to use it, and can now make a few basic measurements of a few types of condenser microphones and get reasonable results (frequency response, THD+N, overload point). But noise is in a whole other class of difficulty. If I really suspected a noise problem in one of my better microphones, I might try to make my own measurements just to see what I happened to get--but I wouldn't trust the results very far, and would send the microphone to the manufacturer for checkout.

--best regards
Logged

Jim Williams

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 575
Re: Km56c performance test
« Reply #10 on: September 18, 2016, 01:19:41 pm »

The reason all the major manufacturers use the Audio Precison is its precision and repeatability. You don't have to strain and decipher small noise differences, it's all plotted out on a 10~1mhz graph. Spectral noise contributions are also easily seen, low end hum/psu isssues are easily seperated from upper frequency hiss. That is informative to the user as to the source of the noise.

A weighted or unweighted measurements are also available so you can pick your poison. Plus you have a plethera of other tests, THD, IMD, FFT, frequency response, phase response, etc.

There are labs that accept outside devices to test as well as rental agencies that will rent an AP for a couple of days. These are not out of the hands of the curious if one is willing to make the effort. It will be a learning experience for all that try.
Logged

Kai

  • Sr. Member
  • ****
  • Online Online
  • Posts: 314
Re: Km56c performance test
« Reply #11 on: September 19, 2016, 12:53:40 pm »

...Audio Precison is its precision and repeatability...
Although the Audio Precition is a nice thing to have, it's like shooting with a canon on sparrows for a simple noise measurement.
It's not even what I would call plug and play, you still have to do know what you do.

What you need is a very quiet environment. This can even be a closed box  inside a closed box in a quiet room.
Then you need some kind of speaker in a low reverberation room, generating a test tone, e.g. 1kHz, as a fixed level reference.
First record the test tone from a defined distance like 10", then put the microphone into the boxed box and record the noise without changing anything else.
Do this with all the microphones you want to compare, it does not matter if you change gain settings between different microphones or not.
Only the test tone and distance needs to be the same all the time.
Then adjust the tracks of the different microphones so that the test tone comes out with the same level for each.
Now you can compare the noise by ear, any type of meters and even analyze it with software analyzers of your choice.
If you want to measure absolute values you needs a calibrated reference microphone for determination of the test tone level, but that's not absolutely necessary.
If you have an iPhone you can use the app AudioTools as Sound Level Meter, it's default calibration is usually quite close (within about 1-2dB).
Logged
Pages: [1]   Go Up