Topic: Mic mod for higher sensitivity?

[szegedin]

Hello
When I hear about mic mods, it's generally about replacing caps to make the mic sound 'better.'
But I was wondering whether [a typical, transformer based moddable] mic can be specifically modded in such a way as to increase sensitivity?
For what I do, I would gladly trade headroom for sensitivity, so wondering if that's done.

[DSatz]

The self-noise of a condenser microphone comes mainly from its capsule capacitance (at low to low-mid frequencies) and its first stage (at mid and high frequencies). If you could modify the circuit for increased sensitivity (i.e. the gain of the amplifier circuitry), you'd inevitably boost that noise right along with the signal, AND you'd lose headroom.

Given that, you might as well skip the troublesome modification process and use an in-line booster preamp (Sanken, Cloudlifter, Triton, etc.) instead--though you'll need to arrange phantom powering for the microphone itself, since most of those in-line preamps take it for themselves and don't pass it through. Or you can simply increase the gain on your preamp--it'll do the same thing.

That said, some microphones are designed with built-in sensitivity options. For example, all Schoeps CMC-series amplifiers have a circuit bridge that allows their output levels to be increased by 5 dB above the standard model. The noise floor rises by about half a dB (relative to any given sound pressure level) when you do that, though, and again the headroom is lowered by the amount of the gain increase. So I don't know whether that would be exactly what you want, or exactly what you don't want ...

--best regards

[szegedin]

The self-noise of a condenser microphone comes mainly from its capsule capacitance (at low to low-mid frequencies) and its first stage (at mid and high frequencies). If you could modify the circuit for increased sensitivity (i.e. the gain of the amplifier circuitry), you'd inevitably boost that noise right along with the signal, AND you'd lose headroom.

Given that, you might as well skip the troublesome modification process and use an in-line booster preamp (Sanken, Cloudlifter, Triton, etc.) instead--though you'll need to arrange phantom powering for the microphone itself, since most of those in-line preamps take it for themselves and don't pass it through. Or you can simply increase the gain on your preamp--it'll do the same thing.

That said, some microphones are designed with built-in sensitivity options. For example, all Schoeps CMC-series amplifiers have a circuit bridge that allows their output levels to be increased by 5 dB above the standard model. The noise floor rises by about half a dB (relative to any given sound pressure level) when you do that, though, and again the headroom is lowered by the amount of the gain increase. So I don't know whether that would be exactly what you want, or exactly what you don't want ...

--best regards

Got it, thanks. So you're basically saying that for a given amount of gain, a preamp section is doing the best it will be able to do for self-noise, and there won't be a mod that lowers self-noise at a given output or increases output at a given self-noise. But is that always the case?
Am I actually off-track with the question, because sensitivity is a function of the capsule, not the preamp section?

That Schoeps option sounds good. For my purposes, I can stand to have a maximum spl way below most people's needs, but I have more need for extremely low self-noise/high sensitivity.

[DSatz]

Again--microphone "sensitivity" by common sense would mean the ability to pick up sounds at very low levels without their being obscured by noise. But the word has a different meaning in audio engineering: the efficiency relationship between the sound pressure level that a microphone picks up and the voltage it puts out in response. So from a technical standpoint, this thread isn't about sensitivity; it's about self-noise. Those two aspects of a microphone are related but not identical.

The only way you can get audibly lower noise from an existing microphone would be if its manufacturer had been cutting costs by using a mediocre FET for the first circuit stage. You won't find that with top-tier professional microphones; they're generally within a fraction of a dB of the lowest noise possible for their given capsule and circuit design. But with some Chinese clones, sure--a good technician could probably reduce the noise a few dB by dropping in a better, selected FET and bias resistors.

The thing is, without trying it, you couldn't know for sure whether you would notice any difference in real-world recording situations or not. Even if a difference was measurable, it might not be audible except during the silences in a direct, 1:1, before-vs.-after comparison, under conditions where the ambient noise was carefully controlled to be the same. That's not most real-world recording, for sure.

While we're on this topic--the very quietest microphones are large-diaphragm models (because a diaphragm with larger active surface area will have greater capacitance, all other things being equal), but the reverse isn't necessarily true; a very high-quality, small-diaphragm condenser microphone can (in practice) be audibly as quiet as the majority of large-diaphragm condensers. It's a mistake to assume that all LDCs are quieter than all SDCs; the two categories overlap quite substantially.

Noise, and our perception of it while we are listening to an intended signal, is a very complex topic. Reducing the perceived effect of noise to a single number is a rather "iffy" thing at best. The frequency spectrum (distribution) of noise is extremely important, plus we don't perceive all frequencies equally at different sound levels; in particular, low frequency noise tends to lose most of its importance at low SPLs (see https://en.wikipedia.org/wiki/Equal-loudness_contour if you want to know more about this). Also, impulse noise is more noticeable than steady noise (though how to characterize and match the levels of such very different sounds is a basic problem), so the time dimension of noise is a big factor.

In general one should set aside all A-weighted, rms noise specifications for microphones; they're bullshit even when they're the result of actual measurement (as opposed to being "aspirational" specifications forwarded from the marketing department). A much more relevant specification is the CCIR-weighted, quasi-peak (or "qpk") noise specification, which is usually some 9 to 12 dB higher than the A-weighted, rms number. (Guess why most manufacturers don't publish that information ...)

That said, even with the best of intentions, different manufacturers can sometimes get remarkably different results when they measure microphones according to the same standard. A decade or so ago, an Audio Engineering Society committee that I'm a member of organized a "round robin" test, in which the characteristics of one microphone were measured by about 10 different manufacturers in turn and the results compared, including the best-known names (Neumann, Schoeps, AKG, DPA, Sennheiser, Microtech Gefell, Audio-Technica, etc.). The frequency response and polar response measurements generally agreed within fairly small tolerances, but the noise measurements were far more divergent, with differences as great as 6 dB in some cases. To my knowledge, no one so far has been able to explain this.

--best regards

edited later to add: I shouldn't have said "The only way ..." above. If a microphone's membrane is DC-polarized, you could try increasing the polarization voltage. That would increase the sensitivity of the microphone without raising its self-noise. However, it would take a substantial increase (e.g. doubling the voltage) to yield enough improvement to be audible, while there are definite risks: The maximum SPL would decrease by the same amount as the sensitivity increases; the distortion of the microphone would increase (though maybe only slightly); and the risk of the diaphragm temporarily getting stuck to the backplate (muting the mike's output until the mike can be disconnected from powering) would increase.

[relefunt]

Thank you very much, DSatz. I can’t remember a time you have posted that I haven’t learned something. You have an ability to explain things in a very clear and simple way without seeming to sacrifice any of the details. That is a rare talent and I hope you consider someday writing a book geared toward tapers, filled with your knowledge and pearls of wisdom.

[Gutbucket]

Great info here, thanks.


relefunt, if your microphone is an omni you can increase it's effective sensitivity and signal to noise ratio +6dB by boundary-mounting it to a large, hard, flat surface, without any other modification of the microphone itself.  This also increases direct-to-reverberant sound ratio +3dB (difference between the gain in coherent direct sound +6 dB and incoherent reverberant sound +3 dB).

Not always doable, but effectively does what you are asking for by modifying the acoustic condition in which the microphone is placed, rather than the microphone itself.