"Where does the tone come from in a microphone?"

[Co.]

Got some examples where you see this?

For example: Let's take a mic, that has a resonance on a specific frequency. Let's say, it is the head basket, that is resonating. This is going to ruin the transient response in that area of frequencies, because any impulse is going to be elongated by that resonance. With an EQ, i just can make this resonance quieter, but I can't make it shorter in time.
How ist an EQ going to help here, be it MP or not?
Is this a problem of terminology, or am I missing something?

 

[Tim Campbell]

Years ago before his death I worked a bit for and had many conversations with David Blackmer concerning his development of Earthworks mics. He felt that the time domain was where you could solve many of the problems of current microphone design.

Examples of improved off axis response for cardioid can easily be found in Neumann KM84s and Sennheiser MKH40s

 

[Co.]

My problem with judging stuff that by ear is that to my ears many of these "smeary" mics sound just as fast as another mic, but just darker.

So you never heard a mic, that is both bright and smeary?

 

[MicUlli]

I am very reserved using sound attributes like "smeary" or "fast" or whatever is used in the community and rely on measurements. What i see in measurements may be audible..
Here are some measurements of a super cardioid SDC in order to get more facts whether this mic is "minimum phase". The 0° and 90° phase responses are nearly minimum phase, but the 180° phase response shows clearly the behaviour of a 2nd order allpass and is therefore NOT minimum phase.
That leads me to the conclusion that the impulse responses for sound coming from the front up to an angle of 90° (the preferred use case of this type of mic) are completely linked to the frequency response via f->t or t->f transformation principles.
The excess phase behaviour for sound coming from behind is not that critical because the group delay is smaller than approx. 100 us and therefore not audible at all...

 

[Co.]

I am very reserved using sound attributes like "smeary" or "fast" or whatever is used in the community and rely on measurements. What i see in measurements may be audible..
Here are some measurements of a super cardioid SDC in order to get more facts whether this mic is "minimum phase". The 0° and 90° phase responses are nearly minimum phase, but the 180° phase response shows clearly the behaviour of a 2nd order allpass and is therefore NOT minimum phase.
That leads me to the conclusion that the impulse responses for sound coming from the front up to an angle of 90° (the preferred use case of this type of mic) are completely linked to the frequency response via f->t or t->f transformation principles.
The excess phase behaviour for sound coming from behind is not that critical because the group delay is smaller than approx. 100 us and therefore not audible at all...

So we can conclude, that whatever other phenomena we perceive on transient behavior, have to be related to some kind of nonlinear behavior?
There is no doubt that there are some nonlinearities in microphones, right?

 

[MicUlli]

Of course the transient response is dictated by additional properties.
For low and medium frequencies saturation and clipping have to be mentioned.
For higher frequencies slew rate plays a role.
Never believe in "max SPL" ratings without any information regarding acoustical limits and frequency ranges..

 

[ricardo]

For example: Let's take a mic, that has a resonance on a specific frequency. Let's say, it is the head basket, that is resonating. This is going to ruin the transient response in that area of frequencies, because any impulse is going to be elongated by that resonance. With an EQ, i just can make this resonance quieter, but I can't make it shorter in time.
How ist an EQ going to help here, be it MP or not?

This is exactly the situation where modern DSP voodoo can help. eg Simple Arbitrary IIRs
Provided the resonance is MP (various caveats about when this applies. see my answer to MicUlli) I can dream up a MP IIR filter that will flatten the response AND get rid of the wonky artifacts in the IR.

Of course I would want to alleviate as much of the yucky resonances first cos the EQ would only apply exactly to the 'measured' direction. It would help in other directions but not eliminate it.

 

[ricardo]

... but the 180° phase response shows clearly the behaviour of a 2nd order allpass and is therefore NOT minimum phase.
That leads me to the conclusion that the impulse responses for sound coming from the front up to an angle of 90° (the preferred use case of this type of mic) are completely linked to the frequency response via f->t or t->f transformation principles.

Thanks for this MicUlli. You have demonstrated exactly the situation I hinted at earlier about 'back response' possibly being non-MP

I suspected this from working with Prof Angelo Farina, on whether rooms, with their reflections, were MP.

What we found was that when later arrivals were larger than the first, the transfer function could be non-MP. It was unclear from our small investigation whether this is sufficient or just necessary. Definitely worth a PhD

As the 'back response' is usually a small part of the total sound, it isn't important .. unless you record important stuff on the null of your cardioid

The excess phase behaviour for sound coming from behind is not that critical because the group delay is smaller than approx. 100 us and therefore not audible at all...

Thanks also for showing the scale of the problem. 100us might be at the accuracy limits of modern test gear.

I forgot to mention that some modern test gear will do the Hilbert Transform for you (Clio?) ... including removing the delays etc. The 'Excess Phase' is a measure of how non-MP something is.

MicUlli, what gear are you using for these supa dupa measurements?

An important point which IIRC, I mention in Simple Arbitrary IIRs, is that if a Transfer Function is mostly MP but has non-MP bits (as MicUlli's example). you should only EQ the MP bits. That way you ALWAYS improve the Transient (Impulse) Response.

 

[soliloqueen]

there's also the that systems can perform hugely differently depending on the period of time over which energy is inserted. i have seen electronics perform very differently when injected with a dirac pulse vs swept with a sine wave w the same energy (especially transformers), even though ideally there would be no difference between a sweep->IR deconvolution and a true dirac IR. systems can genuinely have different response to transients on a purely time/amplitude scale outside of the whole time/FR domain.

 

[Co.]

This is exactly the situation where modern DSP voodoo can help. eg Simple Arbitrary IIRs
Provided the resonance is MP (various caveats about when this applies. see my answer to MicUlli) I can dream up a MP IIR filter that will flatten the response AND get rid of the wonky artifacts in the IR.

Of course I would want to alleviate as much of the yucky resonances first cos the EQ would only apply exactly to the 'measured' direction. It would help in other directions but not eliminate it.

Then the video isn't as bad as I thought.
Then the most reasonable approach should be, to get some low noise, low distortion small diaphragm mics with interchangeable capsules or something like an OC818 and apply filtering and/or saturation as needed.
Doesn't sound very sexy, actually.

 

[MicUlli]

@ricardo: MicUlli, what gear are you using for these supa dupa measurements?
You can derive all information from my document mic measurements part 2, find it in the thread

A poor (wo)man's microphone measurement equipment​

BR MicUlli

 

[Icantthinkofaname]

So you never heard a mic, that is both bright and smeary?

I haven't heard many mics I'd actually think smeared transients together, except maybe super cheap Chinese stuff.

The Oktava MK-012 is a mic that I've seen be called smeary, yet I've seen a few shoutouts where it sounds like it's more accurate than a KM184. You can obviously do better than both of those (Schoeps and the higher end Sennheisers come to mind), but I don't think the "smeary" mic of the two is smeary at all. In fact the only time I've thought that was in a shootout where they used separate takes for each mic so both sounded like they smearing stuff like body resonance and strong noise during certain takes.

This is exactly why I have issues with using terms like "smeary". Too many variables to accurately judge by ear unless we're comparing drastically different mics. In fact I think a certain insanely popular Shure mic that gets used on guitar cabs a lot sounds smeary compared to a lot of $200~ condenser mics, it actually sounds like it smears stuff together when I hear it on guitar or drums (especially acoustic guitar or used for hi hats). They'll work, and sound fine, but you sacrifice a lot of detail and nuance IMO.

 

[ccaudle]

perform very differently when injected with a dirac pulse vs swept with a sine wave w the same energy

If the pulse was the same energy as a sine wave, because of the short time duration of the pulse wouldn't that imply it had a much higher amplitude?
I think that a more fair comparison would be to use reasonable amplitude for each, and average if you needed for better noise performance.

 

[soliloqueen]

If the pulse was the same energy as a sine wave, because of the short time duration of the pulse wouldn't that imply it had a much higher amplitude?
I think that a more fair comparison would be to use reasonable amplitude for each, and average if you needed for better noise performance.

did it both ways (within reason) while troubleshooting, same behavior.

one of the weirdest behaviors I observed in a circuit wasn't even nonlinear with amplitude, it was purely length of stimulus-dependent. it rang louder with shorter pulses, and the intensity of the ringing was proportional to amplitude w no nonlinearity. really weird stuff. it tested flat with white noise and a sweep, and when deconvolved, but rang like a bell with a dirac pulse at the same level, and the effect was continuous. that is to say, the closer to a pulse the input was, the more it rang. So it rang with drums, but not with saxophone. I've never seen anything like that before or since. then again, the circumstances were very extreme.

 

[GeorgeToledo]

So it rang with drums, but not with saxophone.

One reason I am always suggesting people test a mic as a drum overhead after their assessment on vocals. It tells so much.

 

[soliloqueen]

One reason I am always suggesting people test a mic as a drum overhead after their assessment on vocals. It tells so much.

the circumstances were very extreme (there was a transformer running massively underloaded by accident) but it's an important reminder that in bizarre situations, things can act very, very weird. so acting slightly weird in normal situations? not out of the question.

 

[Co.]

This is exactly why I have issues with using terms like "smeary". Too many variables to accurately judge by ear unless we're comparing drastically different mics

I think, the more variables there are, the more you have to know exactly, what you are looking for. Easy to miss something. Easy to measure something and forget, what other factors might contribute to the sound.
If there is only one variable, it is pretty easy to identify, measure and evaluate. If there are many variables, I prefer to use my ears and intuition. Not as a designer, but as a user, of course.

And regarding the mentioned Shure mic: I guess it is smeary, because the transient response can't be on par with your average condenser or ribbon mic, because the mass of diaphragm and coil is much higher, so the response is relatively slow. The harsh and phasey upper midrage won't make up for it, but many people seem to like it or at least got used to it. I don't. Always prefered other mics.

 

[abbey road d enfer]

And regarding the mentioned Shure mic: I guess it is smeary, because the transient response can't be on par with your average condenser or ribbon mic, because the mass of diaphragm and coil is much higher, so the response is relatively slow.

Smearing is not necessarily related to "slowness". Slowness that accompanies normal roll-off is well-known and so natural it is not perceived by the audition as an anomaly.
What could probably be the cause for the perceived smearing is the fact that a dynamic mic has many individual resonances and break-up modes.
These resonances show up as "discontinuities"* in the graph, that are usually smoothed out either by choice or by the limits of the test equipment.
An unsmoothed graph is generally unreadable and unusable. Unreadable because they show a blurred/hairy line with many very narrow peaks and troughs, and unusable because the cause of these is partially intrinsic to the mic and partially due to reflections in the test space.
These "discontinuities" are undoubtedly cause for smearing; the audibility of which being an unknown, since it's in the beholders ears.

* I use incorrectly the word "discontinuity" because I have never found the correct english equivalent to "accident" as in mathematics.

The harsh and phasey upper midrage won't make up for it, but many people seem to like it or at least got used to it. I don't. Always prefered other mics.

I must say I've never been a fan of SM57, particularly on acoustic guitar or snare, where the difference with an SDC is immediately conclusive IMO.
However I was shocked by George Massenburg's extensive use of SM57's in conjunction with "The Brick" preamps during one of his master classes.

 

[Marc HENRY]

i have several proposals for this guy : make another video comparing vintage cameras, from Hasselbald to Instamatic, while looking the comparison pictures throught the bluring glass of his bathroom.
Next video : great and low cost wines comparison, just adding 15% of Coca light in every glass before tasting...

 

[MicMan]

i have several proposals for this guy : make another video comparing vintage cameras, from Hasselbald to Instamatic, while looking the comparison pictures throught the bluring glass of his bathroom.
Next video : great and low cost wines comparison, just adding 15% of Coca light in every glass before tasting...

Dear Marc,

The man who made the video - linked at the top of this thread - starts by saying "I don't know anything about microphones.." (..he's a musician, not an engineer..) but he goes on to examine different BASIC ideas and characteristics about various mics, and gives simple comparisons which may be useful, to a reasonable extent, to people who, as he says, "..don't know anything about microphones".

He gives a broad-brush overview, and some useful comparisons for a beginner. I personally can't see what all the fuss is about wines ..I'm not stupid, but wines simply don't hold any interest for me. I do have, though, a (weirdly?) fine discrimination between different kinds of sweetness and can tell immediately the differences between Coke and Pepsi ..and other brands.. and between drinks with - ugh! - artificial sweeteners (aspartame, saccharine, etc) and between the hideous, for me, (natural) stevia, and normal sugar ..from either beet or cane.

We each have different capabilities and talents (..and for many years I wrote professionally assessing 'Blads, Minoltas, Nikons, Canons, Leicas, Olympus and all the rest of them ..but in a way which was understandable to photography newbies besides cognoscenti..) but why ridicule a man who's trying to show - and has put plenty of effort into it - a basic, and entertaining, grounding in mics? He's an 'enthusiast' and he doesn't profess to be an expert.