R/E/P Community

Please login or register.

Login with username, password and session length
Advanced search  

News:


Author Topic: The Myth of the Accurate Microphone  (Read 12960 times)

Offline Timtape

  • Full Member
  • ***
  • Posts: 101
  • Real Full Name: Tim Gillett
Re: The Myth of the Accurate Microphone
« Reply #135 on: January 08, 2018, 05:55:53 pm »

 ...I have found that the quality of a mic can best be judged with our ears: how that mic processes complex waveforms arriving at the same time: a timpani's spike, a violin's scratchy bow, a trumpet's overtones, etc. all at the same time. Is that multitude processed without smear? Does the soundstage collapse? Can I still pick out individual nuances of each instrument from the whole?...

A wise rule is never to choose a complex explanation when a simple one can explain it just as well.

Many amateur music recordings are fine when only one instrument is playing but when all instruments play  together the sound blurs terribly. Is this because the mic struggles with the "complexity" of many instruments all playing at once?
Often it has little or nothing to do with that.

Quite rightly, a lot has been made here of listening, so here's a listening example, actually from a professional label of a major vocal performer. Admittedly this is from a live TV performance so we need to make allowance for that.

There are multiple audio problems with this presentation but notice what happens as the song progresses. Is it because of the added complexity of the multiple instruments, or is there a simpler explanation? What do you think?

https://youtu.be/uF5E0w2h6wM



 
« Last Edit: January 08, 2018, 09:48:50 pm by klaus »

Offline soapfoot

  • Full Member
  • ***
  • Posts: 223
  • brad allen williams
Re: The Myth of the Accurate Microphone
« Reply #136 on: January 08, 2018, 05:58:42 pm »
A wise rule is never to choose a complex explanation when a simple one can explain it just as well.

On this we can agree.

For instance-- when someone asks "why is this microphone better in this application than that other one?" I'm much more likely to dispense with any objective or pseudo-objective justifications, and just say "it sounds better to me."

Offline Timtape

  • Full Member
  • ***
  • Posts: 101
  • Real Full Name: Tim Gillett
Re: The Myth of the Accurate Microphone
« Reply #137 on: January 08, 2018, 06:59:28 pm »
On this we can agree.

For instance-- when someone asks "why is this microphone better in this application than that other one?" I'm much more likely to dispense with any objective or pseudo-objective justifications, and just say "it sounds better to me."

I take it the person already knows or trusts it sounds better. They want to understand  WHY it sounds better.  How does effectively answering "because it does" help them?

Offline klaus

  • Moderator
  • Hero Member
  • *****
  • Posts: 1501
Re: The Myth of the Accurate Microphone
« Reply #138 on: January 08, 2018, 09:57:20 pm »
Hey Tim,
RE: your reply #135.
When you quote something I said in post #118, i.e. 17 posts earlier, and don't fill in the context, it does not make for good reading or follow up.

WHY something sounds better than something else can be a big, and still largely unanswerable question. But it often has to do with an incomplete rendering by the mic of what we heard in the original. What's missing can be lack of proper processing of many impulses in time at certain frequencies (congestion), or a deviation in the dynamic processing of the original's behavior, or a lack of synergistic interplay of sound-shaping components within a mic's architecture...

Here is an example of the latter, and how misleading it sometimes can be to approach audio improvements by pure logic:
Take tube biasing. Nobody would make the argument that fixed biasing of the impedance converting tube in a condenser mic is superior for the noise floor and frequency response than self-biasing. So I experimented changed the fixed bias of a VF14 tube in the U47 mic to cathode biasing. The effect? Much quieter! Cleaner! Better highs! Yet, it sounded horrible on the scale of listener satisfaction. The dynamic behavior of the mic had turned anemic, its famous characteristics -heft, rich texture in the lower mids, a hyper presence in all the frequencies that matter for a singer, were all gone. It sounded like an audio engineer envisions a linear processor by means of a theory, without feedback from the ears. 

This is just one component in a mic of immense consequences that, once altered, no longer worked in sync with the rest (capsule, transformer, supply voltage, etc.)

We know from the mic's history, that at least some of its building blocks were haphazard, due to post-war shortages, and not what would ideally have made sense, yet the building blocks available fell into place and complemented each other in a most beautiful way.

That again tells me it's fallacy to assume an ideal path towards an ideal mic. The variables are simply to many and their effect on the whole too complex to leave mic design to measurable quantification alone.
 
« Last Edit: January 08, 2018, 10:24:41 pm by klaus »
Klaus Heyne
German Masterworks®
www.GermanMasterworks.com

Offline klaus

  • Moderator
  • Hero Member
  • *****
  • Posts: 1501
Re: The Myth of the Accurate Microphone
« Reply #139 on: January 08, 2018, 10:35:19 pm »
RE:
Quote
There are multiple audio problems with this presentation but notice what happens as the song progresses. Is it because of the added complexity of the multiple instruments, or is there a simpler explanation? What do you think?

https://youtu.be/uF5E0w2h6wM
The mic channel is horribly overloaded on loud passages. So I am not sure how your audio sample relates to the point I brought up.

We don't play games here, where the pupil as homework has to find the answer for the teacher's question; so why don't you tell me how an overloaded vocal mic track relates to my point how poorly executed audio components collapse when too many complex waveforms arrive, and how that is hard to quantify and isolate in technical terms?
Klaus Heyne
German Masterworks®
www.GermanMasterworks.com

Offline Timtape

  • Full Member
  • ***
  • Posts: 101
  • Real Full Name: Tim Gillett
Re: The Myth of the Accurate Microphone
« Reply #140 on: January 09, 2018, 01:28:05 am »
The mic channel is horribly overloaded on loud passages.

Yes the vocal is badly overloaded on loud passages. 

... why don't you tell me how an overloaded vocal mic track relates to my point how poorly executed audio components collapse when too many complex waveforms arrive, and how that is hard to quantify and isolate in technical terms?

Admittedly the Whitney track isnt the best example of the total effect, but generally in a recording which starts quietly with only a few instruments, and then more instruments come in later,  it would be easy to conclude that the cause of the later distortion is the inability of the recording equipment to handle the complexity of multiple instruments themselves. But correlation is not causation. Far more likely the equipment is overloading at some point in the chain due to the combined extra volume of all the instruments playing together, as in a symphony orchestra - or this Whitney remaster, although it's mainly only apparent on the vocal here.

Re the point about  complexity. As I understand it, the many instruments are already "mixed" in the air before they reach the recording microphone, or even our ear. It's now a composite waveform of all those instruments. Like A + B + C + D = X. The mic or ear only deals with the composite X. So long as the mic's fidelity is accurate enough for that waveform that's all that matters. If there's distortion due to the mixing of the various instruments I guess we should  blame the air...

Yes of course  poorly designed, or poorly chosen, or poorly positioned mics - or any other stage in the signal chain- can cause havoc in a recording. It only needs one thing to be wrong. In my experience it's usually less the equipment than  the operator, the "nut behind the wheel"  that causes most problems- as with this Whitney track. I believe today's audio gear, including mics, is fantastic, and  even music recordings made with care many decades ago can be very satisfying to listen to, so long as everything is done well and somebody hasnt messed it up somewhere along the way.

And this track only seems to underline your very important point about listening to the result. I have been a keen listener all my life. Thanks for listening to the track Klaus. I suspect we both have always been ardent listeners and always will be.

Tim 

 





« Last Edit: January 09, 2018, 04:04:22 pm by Timtape »

Offline Timtape

  • Full Member
  • ***
  • Posts: 101
  • Real Full Name: Tim Gillett
Re: The Myth of the Accurate Microphone
« Reply #141 on: January 11, 2018, 05:00:57 am »
Hey Tim,
RE: your reply #135.
When you quote something I said in post #118, i.e. 17 posts earlier, and don't fill in the context, it does not make for good reading or follow up.

WHY something sounds better than something else can be a big, and still largely unanswerable question. But it often has to do with an incomplete rendering by the mic of what we heard in the original. What's missing can be lack of proper processing of many impulses in time at certain frequencies (congestion), or a deviation in the dynamic processing of the original's behavior, or a lack of synergistic interplay of sound-shaping components within a mic's architecture...

Here is an example of the latter, and how misleading it sometimes can be to approach audio improvements by pure logic:
Take tube biasing. Nobody would make the argument that fixed biasing of the impedance converting tube in a condenser mic is superior for the noise floor and frequency response than self-biasing. So I experimented changed the fixed bias of a VF14 tube in the U47 mic to cathode biasing. The effect? Much quieter! Cleaner! Better highs! Yet, it sounded horrible on the scale of listener satisfaction. The dynamic behavior of the mic had turned anemic, its famous characteristics -heft, rich texture in the lower mids, a hyper presence in all the frequencies that matter for a singer, were all gone. It sounded like an audio engineer envisions a linear processor by means of a theory, without feedback from the ears. 

This is just one component in a mic of immense consequences that, once altered, no longer worked in sync with the rest (capsule, transformer, supply voltage, etc.)

We know from the mic's history, that at least some of its building blocks were haphazard, due to post-war shortages, and not what would ideally have made sense, yet the building blocks available fell into place and complemented each other in a most beautiful way.

That again tells me it's fallacy to assume an ideal path towards an ideal mic. The variables are simply to many and their effect on the whole too complex to leave mic design to measurable quantification alone.

That may be true with the vintage U47 but as I understand it, it's not nearly as much a problem with "non vintage" mics where tubes as impedance converters, and transformers for balanced outputs, have been superceded by FETs and balanced active output stages, unless one is wanting the particular distortions those  components  and circuits can produce.
« Last Edit: January 11, 2018, 09:48:51 am by Timtape »

Offline soapfoot

  • Full Member
  • ***
  • Posts: 223
  • brad allen williams
Re: The Myth of the Accurate Microphone
« Reply #142 on: January 11, 2018, 08:13:18 am »
William Bruce Cameron summed it up nicely, for me:

"Not everything that can be counted counts, and not everything that counts can be counted"

Offline leswatts

  • Newbie
  • *
  • Posts: 8
  • Real Full Name: Leslie Watts
Re: The Myth of the Accurate Microphone
« Reply #143 on: January 11, 2018, 03:09:09 pm »
I've reread this thread, and have a few comments in the perspective of a microphone designer.
1) Microphones are not ears. While seemingly obvious, let's give a couple glaring examples

Room walls, nearby things like other microphones, and the like create resonances with nodes and antinodes. We're all familiar with things like bass suckout in the middle of a room when a speaker is along a far wall or corner.
The nodes/antinodes are usually represented as pressure data, but particle velocity (integrated pressure gradient) has nodes/antinodes too, but spatially opposite...That is a pressure node is a velocity antinode.

So let's say a series of pressure antinodes is picked up by an omni ( pressure) microphone or ear.
If we replace that mic with a velocity type (figure 8) it will pick up nodes at those same locations. Peaks and dips are transposed. This sounds completely different!
What about a cardioid? It's equally sensitive to pressure a velocity. So the pressure part picks up a peak, and the velocity part picks up a dip. The sum is 6 dB lower. Again, the sound is completely different from what an ear hears.
Does this imply we should all be using omni mics for best accuracy? No...the ear brain/headmotion localizing element is missing (even with one ear as Jim mentioned) so it also sounds different from what we hear.
These are not subtle effects, they are major.

The other point...about mics being measured with pure sine waves vs the effect with complex music signals.
Any complex signal can be represented by a sum of pure sine waves within an arbitrary interval. A simple
example is the dirac delta function. Think of it a a single infinitely narrow spike...kind of a perfect impulse response. If you add the sum of ALL simple sine waves in phase at that point the sum will be the dirac spike. Everywhere else there is cancellation. The sine waves have to have been always going and always will be going though. If they are started or stopped the sum will not be the single pulse.
Quantum mechanics was mentioned a few pages back, and this happens to to be an example of the QM
Heisenburg uncertainty principle.

So if any instrument does well with a sine sweep (with little nonlinearity) it will do well with a sum. You must have nonlinearities to have intermodulation.

Les
http://lmwattstechnology.com/

Offline klaus

  • Moderator
  • Hero Member
  • *****
  • Posts: 1501
Re: The Myth of the Accurate Microphone
« Reply #144 on: January 11, 2018, 03:29:50 pm »
Microphones are not ears. While seemingly obvious, let's give a couple glaring examples(...)

I am not aware of any post in this thread that made that claim. Microphones, in my opinion, are an approximation of what we hear with our ears. In some cases that approximation- through clever engineering or random happenstance- gives us the emotional connection to the music, much like our ears can.


Quote
Any complex signal can be represented by a sum of pure sine waves within an arbitrary interval (...)
So if any instrument does well with a sine sweep (with little nonlinearity) it will do well with a sum.

If it's that straight forward of a measurement issue, why doesn't any of the mics which promise to be "accurate", or any of the interconnect cables that claim to be artifact-free and "without congestion in the mids" deliver?

You mention you are a microphone designer (if I got this right). Have you succeeded in engineering a mic that, while not as sophisticated in design as ears, at least delivers distortion-free (under any definition) audio?
« Last Edit: January 11, 2018, 03:31:31 pm by klaus »
Klaus Heyne
German Masterworks®
www.GermanMasterworks.com

Offline leswatts

  • Newbie
  • *
  • Posts: 8
  • Real Full Name: Leslie Watts
Re: The Myth of the Accurate Microphone
« Reply #145 on: January 11, 2018, 05:47:15 pm »
Quote
You mention you are a microphone designer (if I got this right). Have you succeeded in engineering a mic that, while not as sophisticated in design as ears, at least delivers distortion-free (under any definition) audio?

Yes, I was a development engineer for Shure and Electrovoice. I now design my own line as well as mics for other companies. That includes Condensers, ribbons, dynamics, and MEMs.

None of our microphones are distortion free. By design.

The best spec mics we have are probably our B&K measurement standards. They generally sound horrible in the studio.

Condenser microphones with flexible diaphragms are inherently distorting due to radial charge movement.
Our piston motion MEMS don't have that problem.

There's science and then there's marketing. Most audio marketing is nonsense of course.

What we try to do is use our scientific knowledge to manipulate parameters to make products that sound good. None of them work like ears...that's not our goal. It's all about expectation bias.

We have a new modeling technology to assist us in that goal...there will be some AES papers about that.

Do we achieve our goal? Guess I need to loan you a mic so you can be the judge.
Right now i'm on winter vacation on a sailboat so all I can do is show you a clip.
I owe one to Sound on Sound for a review too.

I write papers and books (and drink rum) on the boat in winter then return to the acoustics lab at springtime.
Call it semi retirement.

http://lmwattstechnology.com/microphone/polyribbon/soundsamples/soundsamples.html

Les


http://lmwattstechnology.com/

Online Kai

  • Sr. Member
  • ****
  • Posts: 283
Re: The Myth of the Accurate Microphone
« Reply #146 on: January 11, 2018, 07:23:14 pm »
Microphones, in my opinion, are an approximation of what we hear with our ears. In some cases that approximation- through clever engineering or random happenstance- gives us the emotional connection to the music, much like our ears can.
To me a microphone is quite similar to a photo camera, and a recording similar to a photograph.
The photograph will never be the same as the beautiful woman that has been pictured (or would you like to have a date with a photograph 😀?!), but in some cases she might even look better on the photograph than in reality.
This relativates the question if there is a perfectly accurate microphone. What would that accurate microphone deliver?
Is that microphone that makes a strong connection to the music really accurate or just a way to make a very nice looking photograph?
« Last Edit: January 11, 2018, 07:28:05 pm by Kai »

Offline Timtape

  • Full Member
  • ***
  • Posts: 101
  • Real Full Name: Tim Gillett
Re: The Myth of the Accurate Microphone
« Reply #147 on: January 12, 2018, 06:13:40 am »

This reactivates the question if there is a perfectly accurate microphone. What would that accurate microphone deliver?
Is that microphone that makes a strong connection to the music really accurate or just a way to make a very nice looking photograph?

I think it's helpful to distinguish between "accuracy" and "what sounds good on a released recording", but these days there is no need to "choose" between one and the other.

So why use a recording mic as EQ  when we have far more sophisticated dedicated EQ at hand later on? Similar with adding  euphonic distortion such as from tubes or transformers. These things are far better applied and controlled using those tools as  separate steps after the initial recording is made where we can take our time and finesse the sound. The beauty of multitrack recording (from the 50's) and even more so, digital recording was that it allowed each step to be done without compromise, rather than having to get everything finalised at the initial recording take.

In other words, in recording, we can have the best of both worlds. With the mic we can  capture a vocalist cleanly, optimally, even (dare we say it) fairly accurately but it doesnt have to be "the finished product" straight off the mic. It's a very important step and must be done well, but still it's only the first step.
 
« Last Edit: January 12, 2018, 08:04:00 am by Timtape »

Offline leswatts

  • Newbie
  • *
  • Posts: 8
  • Real Full Name: Leslie Watts
Re: The Myth of the Accurate Microphone
« Reply #148 on: January 12, 2018, 07:44:11 am »
Quote
To me a microphone is quite similar to a photo camera, and a recording similar to a photograph.

The photography analogy is a good one. When I watch an old movie I want to see the garish color and contrast distortion of technicolor. I don't want to feel i'm on the movie set.

Quote
This relativates the question if there is a perfectly accurate microphone. What would that accurate microphone deliver?
Is that microphone that makes a strong connection to the music really accurate or just a way to make a very nice looking photograph?

This is the 21st century, not 1935 or 1980. Electroacoustics is not a mystery to us. We can capture
a sound field at a point accurately enough that any anomalies would be inaudible except for noise.

To me approaching that state results in disappointment with most. My customers want something "bigger than life"...not reality. If they wanted reality we'd give it to them.
At Shure we did many many listening tests...some scientific some not. As we approached results with very good numbers we all learned of the characteristic dull sound it produced. We tried to get that high performance in our products, and were often criticized for it. Personally I prefer accurate sound, but i'm not a customer.
 
The mystery is quantifying "bigger than life" so we can engineer it into a product. We have to turn emotion into math.

Part of that is not electroacoustics. People will hear good things because the product is old, rare,  fragile,expensive, has platinum wires, or is red instead of blue. We accept that.

It doesn't have to be that way though...Moore's law works with microphones too. Most microphones made now are silicon chips made by robots.Billions of them.

Soon the robots will make everything (including the best microphones) and all will be free. No one will have to work. All our needs will be met.

The plutocrats will try to prevent this of course.

Les
http://lmwattstechnology.com/index.html

Online Kai

  • Sr. Member
  • ****
  • Posts: 283
Re: The Myth of the Accurate Microphone
« Reply #149 on: January 12, 2018, 10:00:52 am »
We can capture a sound field at a point accurately enough that any anomalies would be inaudible except for noise.
But that single point of a larger soundfield does not even come close to represent the real event that caused it, so why chasing after the idea to capture it accurate.
And even if you did the impossible and captured the real event inside a recording, playback on loudspeakers doesn't reproduce it.
Whatever you do you end up with only a picture of what happened, so why not concentrate on making this picture look as good as possible without taking the constraints in trying to make it accurate, only to end up with a bad recording?

So many times I had the practical experience where bands came into my studio that had done recordings in the so called audiophile field, with just a stereo microphone in front of the whole band because someone had the idea that this is "accurate".
These bands where astonished and pleased when they heard, for the first time, how good they could sound on a serious recording.

This does not mean I cannot make very good recordings with just a single stereo microphone, but the circumstances must fit.
E.g. I'm going to do such a recording tomorrow, and I will use a single pair of Neumann KM84s, because, on that special event, they have proven to give me the best sound of all that I have in my collection.
I would not consider the KM84 an accurate microphone, but a good sounding one on many occasions.
« Last Edit: January 12, 2018, 10:27:03 am by Kai »