Other than frequency response, what makes a mic sound good? Say there is a cheap microphone, and an expensive microphone with same frequency response/selfnoise/sensitivity/polar pattern.
Now we compare the recordings, they probably will sound different..and the winner will probably be expensive one.
I think your main issue with understanding this is that you're looking at it from too simplified a view.
There are other issues involved as well: such as response time (which factors in to sensitivity, though it is rarely mentioned), or subtle phase differences that contribute to the mic's overall tone/color (independent of frequency response).
As has been mentioned it's largely a case of design. Mics are designed for specific purposes, and trade offs are allowed to control cost versus specified use. It can also depend on the source you're recording. I've been editing some fabric sound I recorded, and two of the mics I used are nearly identical sounding in some of those files (a Neumann TLM-170R in hyper-cardioid and an AKG SE 300B with with CK93 capsule). They would sound very different on a voice recording though.
-- edit --
this question piqued my curiousity enough that I ran a test for the tonebenders podcast. Check out the results here:
-- edi t--
I honestly think this is a good question that's worthy of a little thoughtfulness.
IMO it is possible to eq one mic's frequency response to closely resemble another's. As an example, I can (and have) take a simultaneous recording made with an NT5 and a Schoeps CMC6.Mk4 and use EQ (mellowing of the high end on the NT5) to make the two mics play well together and even feel very simlar frequency-response wise.
Every capsule size has inherent resonant frequencies that interact with the overall recording in complex terms though - so simple slopes and rises won't compensate for that part of a mic's character effectively.
With that said, I actually believe you can go quite a long way with EQ - especially in the upper mids and high end.
The bigger differences are in low end extension (you can't add what isn't there) and in transient response.
I think transient response specifically, is what many people refer to as "detail" and is what can truly separate the sound of one mic from another. As an example, my Line Audio CM3s have a very similar tonal characteristic to the Schoeps CMC6.Mk4 mic, but they track transients much more slowly. This means that they yield similar recordings on things like traffic ambiance, but very different sounding recordings on applause - with the Schoeps yeilding much more detail in the individual claps and the CM3 sounding very different with regads to dynamics.
The other thing about transient response is that it isn't uniform across the frequency spectrum. Its entirely possible for mic A to track transients faster at 10k than mic B (again possibly due to capsule-based resonances), but slower at 2k.
Another factor is electrical components. Mic circuit boards are filled with components that further tune the frequency output, adjust impedance, create pads and filters, and change polar patterns.
The quality of these components can do things like add harmonic distortion, boost or cut frequencies artificially, and further shape transient response at different frequencies. We recently did a shootout with a U87 versus an AT4050 on VO, and the 4050 sounded much cleaner than the U87 (though it lacked that U87 "character", probably mostly due to the differences in the level of harmonic distortion due to the different designs of the internal electronics.
The membrane is an analog device. It's operation varies continuously. Thus, you cannot change the overall recorded signal to sound like what another microphone would have picked up using static post-processing. Regarding noise, different microphones also have different noise prints. Different microphones also have different housings, which affects how the signal is directed to the membrane (reflections and phase cancellation occurs). There's just so many variables.
I don't think it's a question of cheap vs expensive nowadays, at least in studio microphones. It's just good design vs bad design. You can find many microphones with good design for cheap nowadays, although the used inventions are likely ripped from expensive microphone designs (e.g. Neumann). It's just the manufacturing that's cheap.
In field-capable microphones there's much less competition and less cheap copycats.
I wouldn't always say the winner is the expensive one, like a bad mic used well is always better than a good mic used poorly. Weeelllll sometimes : P
But my understanding (and I possibly think this just to sleep better at night) was that the reason you can't truly just process sound from a cheap microphone is that it may have not captured those frequencies in the first place. Or at least not in the same way. So by trying to bring those characteristics out, your enhancing certain aspects, likely the bad along with it.
That and of course the physical components may differ greatly, meaning colouring in the more expensive mics will be closer to what we hear on a daily basis and be what we class as more professional. Plus we need to keep thinking this was, or else who would make money making heart breaking & expensive mics.
expensive/good mics basically pick up more detail and clarity as well as having lower self noise. It's impossible to EQ in detail.
Think of it this way. Do you think you could ever EQ Jimmy Durante singing to sound like David Clovedale of Whitesnake, Whitney Houston, Robert Plant, etc.? Good sound is usually made by having GOOD equipment at each stage of the process; not by having trashy sound that some plugin is trying to make up for. If you don't put it in in the first place, good EQ is unlikely to make up for it never having been there to begin with.
Thought of in terms of bandwidth, can you compress a large text file down to three bytes, and expect to get the original back? Can you take an eight bit recording, upsample it to 16, and expect the lower eight bits to be of similar quality as if you had recorded it in 16 in the first place? You are starting with less detail; i.e. less information. How do you propose to add the missing information? With static EQ settings? I have a friend who has been in music for 30+ years; and still cannot see why upsampling a low bit depth signal doen't yield a high quality one. Save yourself a lot of bad mixes and listen to me: do every part of the process right, if you want high-quality results. You may be able to leave out some steps and get your result sometimes, but you may never let one step degrade the signal, thereby losing information, and expect to get similar quality to if it had been done right in the first place. By "polishing a turd", you will never get a silk purse or a juicy steak.
Listen, I put an MXL V67 ($100 mic) up against a U87 and ran it them through a Millennia STT1, and the cheap V67 won. You can't negate the fact that if you throw a cheap mic onto a cheap pre, you get blah blah. Somewhere in the signal flow, pre's do matter...and I ditto the guy who said, get the organic sound to minimize EQ in the first place. The best engineers (believe it or not) rarely use EQ. And btw, the a/b test on the mics I shared above, was tested at the Spirit Ranch when I was working with Bud Snyder (Grammy nom. Engineer for the Allman Bros.) - he was the one who hipped me to comparison. We laughed.
basically, get a U87 and a cheap condenser, make a 'hissing' S sound into each and record the results. you'll have your answer.
Steven Slate seems to have figured out a way to do this with his Virtual Mic System. http://www.slatedigital.com/vmspreview/ it's worth a look.
There are some complex well thought out answers here, but to me the straight up question "can you make a cheap mic sound like an expensive one?" The answer is still unequivocally NO! You cannot skirt around it with long winded explanations. Making two different mics "work well" together with use of eq is a totally different topic.
A microphone recording is one-dimensional, the audio landscape at a location quite more complex once we leave pure pressure recording (omnidirectional microphones). You cannot EQ away any recorded noise: good microphones have a much lower noise floor. And much of the difference between expensive and outrageously expensive microphones lies in the directional characteristic over their frequency range: the diagrams for the really expensive ones don't change all that much over frequency, so color coming from off-axis does not get colored all that much.
You also cannot really EQ good impulse response back in: that's about as hopeful as making a good quality picture from one out-of-focus: when attempting to do so, you get lots and lots of noise in. That may be tolerable for investigative purposes (trying to understand words in an audio recording), but it will not make for good quality recordings.
Any non-trivial post processing comes with associated costs in noise and clarity. You can improve recordings according to several objectives, but it turns out that the objective "enjoyable" is rather hard to establish after the fact. Usually the best results will be achieved when the microphone already delivers results rather close to what you want.
Which may actually be something different than "most expensive", particularly for close-captioning microphones that are part of a live setup rather than of a recording intended to be sort-of authentic: singers' microphones for example.
