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

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
In case you hadn't noticed: AKG only exists as a brand name. They've been closed down by their current owner, Harman.
Which is also another brand name. the actual owner is Samsung.
I'm not privvy to Samsung's work organization, but I would think all mics are now subcontracted to different jobbers. The 414 capsules are probably an OEM derivative of an existing line.
No factory, no lab, no nothing in Austria remains.
From the ashes, the Phoenix is reborn under the Austrian Audio name.
 
Interesting video but not super conclusive due to the way the tests were done. Still, a fun watch.
To me, I find that vocals are what is going to separate these mics into two camps. The ones that become edgy and harsh and the ones that remain sweet even when pushed. Outside of vocals, it gets easier (to me) to make quicker decisions. So many mics can do the job but a few have something magical and that always comes through on vocals.
 
reducing mics to their frequency response, without mentioning transient responce

Aren't mic capsules minimum phase, so that transient response can be derived from frequency response? Or does the delayed acoustic path used to create the polar pattern make all except omni capsules non-minimum phase?
 
Aren't mic capsules minimum phase, so that transient response can be derived from frequency response? Or does the delayed acoustic path used to create the polar pattern make all except omni capsules non-minimum phase?
Transient response can be derived from FR, as they are inseparable. However, since he didn't use a flat reference mic, and used probably low grade car speaker you can't really deduct transient response from his measurements.

Basically if a mic extends to 20K even with some reasonable attenuation of say -6db at 20K you can deduct it has good transient response. You can go into full audiophile mode and ask how much does the FR extend beyond 20K.
 
To me, I find that vocals are what is going to separate these mics into two camps. The ones that become edgy and harsh and the ones that remain sweet even when pushed.
At the time I worked for radio, the general opinion was that a microphone for spoken word was not very critical.
I've always fought that. The human voice produces such a complex audio spectrum, that (anyway for me) it is very easy to hear differences in the sound of microphones. For me a human voice gives me much more information than a microphone comparison on musical instruments. (Guitar, piano etc.) Musical instruments can sound (almost) the same when recorded with different microphones, but on a voice the difference is immediately very clear.
 
I quite liked the sound of the big fat boy with the tube in it, now im getting tempted to make a ribbon mic of my own. first thing is to find a suitable body. i did see a video on youtube where some bloke used a coke can for the body. I would hope to find something a little stronger and sexier. but, this build must be cheap as I dont even need a mic. also as this would be a fun project i would need to wind my own transformer.

would a steel ei core be better than a ferrite core ? I have no idea as i am a mechanical engneer not an electrnics bloke.

once i find a suitable body ill start a fresh thread and welcome any input
 
Aren't mic capsules minimum phase, so that transient response can be derived from frequency response? Or does the delayed acoustic path used to create the polar pattern make all except omni capsules non-minimum phase?
As you surmise, the introduction of delay between the front and rear paths makes pressure-gradient mics non MP. Correlation between frequency response and transient is not verified above about 5kHz.
 
I looked at the video only for a few minutes... I think there is a fundamental flaw here (perhaps later on it takes those things into account, but I comment on the part I saw). The source the guy uses is a coaxial speaker and the microphone used in very close proximity. As such, all the music information comes only on axis and also, all the microphones are compared in the region of the maximum proximity effect buildup, so the low end is wrong.

Indeed, in this situation the main difference between the microphones will be frequency response of mid to the top end response. However, in the real life situation the 'microphone tone' rather comes from off axis response and (especially) from phase response integrity, which can be all over the place for a conventional cardioid microphone. In this case that 'on axis frequency response' becomes almost useless. This is one of the reasons one should always take the published frequency response charts with a (large) grain of salt and always audition a microphone in a real situation with complex sound sources...

Best, M
 
I do appreciate that he is not doing this for some AES paper but rather he is personally invested in a particular sound he wants to recreate and is standing in front of a lot of myths erected by historical narratives in the industry...


The test basically comes down to "I Like the sound of "X"-performer..he used "Y" microphone and it costs "Z" dollars something I cannot afford because my car door handle is broken...


Oh wait! the country music folks just gave me a buttload of $ lets spend it...(and then with ample integrity he leaves his door handle broken to remind us how pure he is which may or may not be part of the projected narrative)...

And then he does what about 90% of the folks here would do he DIY-it...

And he does discover some of the central issues in mic tone...or whatever it is he is chasing...

It's a fun watch...well made and interesting...

Its also Youtube compressed and mp3 grade audio so there's that...and his methodology is less than White Lab Coat by a fair margin...


An SM57 isn't a baseline of anything as far as I can tell but if thats what you choose then the results are whatever you decide they will be.

Small speaker with tweeter is gonna miss a huge amount of frequency material, and transient response...not to mention the proximity effect we have already noticed as no longer part of the test...


He discovers "The capsule is the thing"


Something everyone here is well aware of because...it is.

At the end of the day he is not releasing peer reviewed material he's promoting the country music thing...fair enough...

But it's always fun to see those places in interesting perspectives...

Its fun. Good clean fun...and maybe a little bit of accidental science...which is probably a good thing.
 
As you surmise, the introduction of delay between the front and rear paths makes pressure-gradient mics non MP. Correlation between frequency response and transient is not verified above about 5kHz.
I'm not sure that is the case Abbey. At least 2 different paths is a 'necessary but not sufficient' requirement for Minimum Phase. Certainly, I've not seen non-minimum phase below 20kHz in the mikes I've measured.
 
At the time I worked for radio, the general opinion was that a microphone for spoken word was not very critical.
I've always fought that. The human voice produces such a complex audio spectrum, that (anyway for me) it is very easy to hear differences in the sound of microphones. For me a human voice gives me much more information than a microphone comparison on musical instruments. (Guitar, piano etc.) Musical instruments can sound (almost) the same when recorded with different microphones, but on a voice the difference is immediately very clear.
I absolutely agree with you on this. However, isn't it soul-searching that most voices can be rapidly recognized through all kinds of microphones and speakers, from a carbon mic in an old-style telephone to a Neumann/AKG/AT/whatever?
It shows that voice identification goes way beyond simple frequency and transient response, it includes other things like punctuation and flow, which, of course, are not microphone-dependant.
Voice identification is a big part of what takes part in the assessment of a mic.
It may be why the NOS gurus surmised "that a microphone for spoken word was not very critical."
Here in France we a have, as many other countries, a lot of news shows.
Most use a traditional set-up of Schoeps/Neumann SDC but some have made really different choices; in particular there is one where they use "skullhead" microphones, not even the renown Shure, but cheap asian copies. I guess the looks were determinant in this choice.
 
I looked at the video only for a few minutes... I think there is a fundamental flaw here (perhaps later on it takes those things into account, but I comment on the part I saw). The source the guy uses is a coaxial speaker and the microphone used in very close proximity.
I fully agree with the fact that the experiment is far from absolute scientific rigor. However, had you watched to whole vid, you may have concluded a bit differently.
As such, all the music information comes only on axis and also, all the microphones are compared in the region of the maximum proximity effect buildup, so the low end is wrong.
But isn't it the case of many recorded tracks? Close micing is dominant in many styles of music. Classical music excluded...or not.
Of course the "measurements" have zero absolute value, but as a comparative, they are indicative of the differences that may be experienced in many situations.
This guy is a studio musician, for him it's how mics are used.
Indeed, in this situation the main difference between the microphones will be frequency response of mid to the top end response. However, in the real life situation the 'microphone tone' rather comes from off axis response and (especially) from phase response integrity, which can be all over the place for a conventional cardioid microphone. In this case that 'on axis frequency response' becomes almost useless.
I agree that off-axis response is a very important factor, particularly when using mics in distant or semi-distant micing and regarding feedback.
This is one of the reasons one should always take the published frequency response charts with a (large) grain of salt
I believe we all know that microphone frequency response charts are drawn by the mktg dept. ;)
 
Last edited:
I'm not sure that is the case Abbey. At least 2 different paths is a 'necessary but not sufficient' requirement for Minimum Phase. Certainly, I've not seen non-minimum phase below 20kHz in the mikes I've measured.
I know we already differed on this subject in the past. I lack data to support my claim, but the few microphone transient response graphs I've seen suggest they are not representative of a simple bandpass. The succession of positive and negative pulse of similar amplitude looks to me very reminiscent of the impulse response of the typical 1st-order APF.
I know the LF phase response clearly folows the slope of amplitude response, as has been discussed in another thread, but I believe this should not be extended to HF response, particulaly for large diaphragm mics.
 
I like this video :) Brings a lot of "fresh air" compared to so many dumb mic test videos..
I was impressed of the MD421 comparisons. So much FR deviation i would have never expected for the same type of mic. I fully agree to his findings about mic electronics. If the electronic part is well made the sound differences are negligible.
 
Aren't mic capsules minimum phase, so that transient response can be derived from frequency response? Or does the delayed acoustic path used to create the polar pattern make all except omni capsules non-minimum phase?
If you compare mics to any SDC mic, you can find LDC mics that are brighter but slower, or Ribbon mics, that are darker, but faster.
I don't know, if that is a contradiction, but this is, how I hear some of the mics in practice. I never did any measurements.
 
.. the few microphone transient response graphs I've seen suggest they are not representative of a simple bandpass. The succession of positive and negative pulse of similar amplitude looks to me very reminiscent of the impulse response of the typical 1st-order APF.
I used the Hilbert Transform to check for Minimum Phase (Log amplitude & Phase in a MP system are Hilbert Transform pairs)
Alas I've just lost, in a couple of HD crashes, all my novel work on the subject this and the previous Millenium.

The last time I did serious work on this was an exchange with Prof Angelo Farina on Room Response. An anechoic room is sorta 'MP'. A reverberant room is not. Somewhere between the two conditions, the Room Response becomes non-MP. We were investigating the effect of various 'arrivals' like you describe.

There's a possibility that the back response of a cardioid LDC is non MP but I haven't checked this nor is this beach bum likely to do so. :oops:
 
Last edited:
An anechoic room is sorta 'MP'. A reverberant room is not. Somewhere between the two conditions, the Room Response becomes non-MP.
It's part of the problem. when is the limit between MP and non-MP? It seems commonly admitted that the property can be valid in one frequency domain and not in another.
What I've seen is that at LF mics behave like MP; the controversial debate about dynamics being 90° out-of-phase with condensers clearly demonstrates that.
At HF IDK. I only have hints.
 
I think it's necessary to remember that mic manufacturers - or the sales companies which sell mics - have to sell mics to stay in business, so mics (..like everything else ..like Leica cameras, for example..) are hyped up to be 'the best there possibly is!' ..whether it's true or not.

All those mic reviews in (the UK mag) 'Sound On Sound' always said how great each mic was ..but like any mag, it needs to sell to buyers AND, of course, sell space to advertisers. And (almost) all mics are described as '20Hz - 20KHz' ..but I think only recordinghacks regularly showed, or shows, freq graphs ..which often gave the game away: how many mics really picks up any decent signal at 20Hz? ..When I saw the graphs for the M-Audio Sputnik mic I immediately hired one to find out if it really sounded as good as its - flat - graph showed ..and it sounded better, so I gradually bought 4 of them!

(They've got, er, not a 'notch' in the high mids, but a quirk which I've never read about, whereby it ..how best to describe it? It's like being in a crowded room full of the burble of conversation, and suddenly someone over there mentions a word which means something to you; your name, or an item you've been thinking of buying, or a guitar, or the name of someone you know, and somehow you're 'all ears', and your ears & brain latch onto that particular voice and conversation as your brain pops it to the front of your consciousness, and your ears and brain 'focus' onto it by selecting that delay between the sound coming to both ears, so that you can pinpoint that particular conversation ..your brain and ears become 'active' rather than just 'passive'. That's what the Sputnik does ..so you're not just passively hearing the sound which it delivers, but you're actively engaged with the music ..or at least I am. That'll be the circuitry, probably, not just the capsule.)

'Classic re-issues' are, primarily, a way of capitalising on the reputation of an old mic, and offering a NEW IMPROVED version - as it says on laundry powder. Or NEW IMPROVED ORIGINAL ..or whatever words bring in buyers. But, as a case in point, the NEW IMPROVED RE-ISSUEs of the original AKG C3000 sound NOTHING like the original ..they're just cheaper-made almost-lookalikes without the super-cardioid element which gives the INTIMACY of the original.

But hype of old original mics - like old Leicas - is specific to the period when they first appeared. They were, perhaps, the best then. The Leica M3 was a big step forward in 1954 ..but a current Leica M camera is way behind the spec of other camera makers, though people get suckered in by the reputation of the old original.

I can never hear what's so special about a 414 - or, specifically, a 414 XLII. It sounds 'featureless' to me, but maybe that's the point, or maybe that's just me. It seems to have no 'character', but perhaps those are used to give unaffected, unaltered, true, 'actual' sound of whatever they're being used with: a cello, violins, brass, snare, cymbals, voice, whatever. To me, though, their 'featureless' sound is completely unappealing - same with a TLM103 - and I thing the 414 XLII is roundly beaten by the 'upstart' sE 4400. It has a 'richer', fuller, though similar, sound, but with a little more intrinsic bass.

I want that 'character', not just true-to-life 'verity', from a mic. I want a mic which gives me what I hear in my head, and not necessarily what I hear coming directly from a voice or instrument ..that's why people use companders, and EQ, and the 'density' of tape, isn't it?

Van Gogh didn't paint realistically. He painted what he saw in his mind's eye. If you want 'realistic' try Holbein or Canaletto. If you want an artist's vision, try Turner. If you want what you want, find a mic which gives the sound that you want, no matter what others say.
 
The tone of a mic comes from several things, the diaphragm, and the material that it is made from, which effects where it resonates, how it's damped, the type of mic, if it's a charged plate (condenser) or dynamic (like a speaker with a voice coil), the amplifier if it has one (condensers have at least a FET, maybe more, or a tube or two) the output transformer, if it has one, and what the core is made from, steel, nickel or a blend, the number of windings, etc, then of course how it marries with the input stage of the mic pre.

So the answer is "how long is a piece of string?"

The point of his videos though is not about everything one can think *could* effect tone in perfect listening conditions with super-trained and experienced ears, but normally about subjective differences by regular folks (him) in home-like onditions.. and normally relating to guitars (hence the use of the word 'tone' for mics and his testing method).

So .. it's not really an open question, it's very much limited to his scenarios and basically what makes a *big* difference that everyone can hear easily.

Placement and loudness of the source and how it interacts with the mic polar pattern (which can depend on all the things youy mention) is probably going to make the biggest differences here.
 
Regardless of questions about some of the more academic aspects of his test methods, he does appear to blow a couple of 'myths' right out of the water...
Notably the relatively insignificant differences between silly priced mics and their more sensibly priced counterparts, and also the real world differences - or rather lack of - between tube and FET mics.
None of which is actually very surprising of course...
 
Back
Top