mic capsule ideas

[Mr. PC]

Charge both the diaphragm and backplate, for extra signal!

Curve the diaphragm (and backplate to match)... maybe it's make the front signal stronger?

2+ diaphragms, making rings around eachother. Maybe the outer ring would be end-terminated, and the center ring center-terminated? (For some reason I can't get these images to embed)







https://ibb.co/2h1FN3S
https://ibb.co/nmjf4mY

[DVDdoug]

Condenser mics have been around 100 years so I assume they've "tried everything".


And, I doubt cost is a problem since some microphones are outrageously expensive!


There are some rectangular diaphragms:

Audio-Technica has developed a 4-part rectangular element for the AT5047. These four matched ultra-thin diaphragms function as a single unit (with outputs proprietarily summed) and provide a surface area that is twice the size of a standard 1" diaphragm. By using four diaphragms in a single capsule, you get all the benefits of an extra-large element with an extremely low noise floor.

[Philbo King]

I once started a design for a mic that used laser interferometry to convert the audio to electrical signal; then stopped when I realized my employer would own the idea due to the nature of my job.

In the long run, like speakers, mics end up being affected the most by the acoustic environment they're in far more than by their basic design. And unlike ears, mics don't have an auditory cortex to unravel the reflections and phasing.

So I've decided if I can't make something sound good with a $60, $150, $800 mic, it probably just doesn't sound good to start with, and is akin to putting lipstick on a boarhog. That philosophy has served me well, while saving me a pile of cash on mics I didn't truly require and time wasted on crappy sounding tracking.

[Mr. PC]

Originally Posted by Philbo King

I once started a design for a mic that used laser interferometry to convert the audio to electrical signal; then stopped when I realized my employer would own the idea due to the nature of my job.

In the long run, like speakers, mics end up being affected the most by the acoustic environment they're in far more than by their basic design. And unlike ears, mics don't have an auditory cortex to unravel the reflections and phasing.

So I've decided if I can't make something sound good with a $60, $150, $800 mic, it probably just doesn't sound good to start with, and is akin to putting lipstick on a boarhog. That philosophy has served me well, while saving me a pile of cash on mics I didn't truly require and time wasted on crappy sounding tracking.

Woah, so you make mics? Can you say who's your employer? That's pretty interesting.

So basically lasers would measure the movements of the diaphragm?

Ya, I have an issue with being incapable of action when things don't make sense to me... I should just record everything with my m201 and ND967... I'm just afraid of noise build-up from 'thermal noise' and preamp noise (hoping an sE Dynamite could help that). But now I'm getting back to my specific IRL problems, and this thread is more just about having fun with "What if"

[Philbo King]

I'm retired now. I worked for an aerospace defense contractor (I prefer not to name them) in their reasearch and development center, building all kinds of wild & crazy stuff.

Yeah, the idea was to stretch an extremely thin diaphragm, like gold-sputtered mylar, bounce a laser beam off it in an optical cavity, and detect the variations in the optical standing wave with a fast photodiode. It was just a concept. My reasearch showed they use a similar techniqque for some seismic mics (earth tremor detectors), but at that time I didn't find it applied to audio mics. Ideas are easy; ones that can be developed are harder, and ones that work can be very difficult indeed.

[clepsydrae]

My non-expert take:

Originally Posted by Mr. PC

Charge both the diaphragm and backplate, for extra signal!

AFAIK the polarization voltage is already tuned to a "proper" level; too high and the membrane collapses into the backplate (sometimes it works for a while and then collapses when wind hits the capsule and gets it close enough to attract to the plate...) There is also the issue of the resonance of the membrane, which IIRC the voltage affects, along with other characteristics. I would imagine "charging" the backplate wouldn't change any of these considerations, since AFAIK it's the voltage difference between the two components that matters, and that's already arbitrarily adjustable and optimized in mics.

Not sure about the motivation for the other ideas you mentioned, or what exactly you meant, but yeah, the dark wizards behind mic capsules have certainly tried just about everything and more, or have solid theoretical grounding why it's not a good idea to try something. The mic building I've done has been a series of lessons in how deep that rabbit hole goes (I just poke my nose in.)

That said, it's always great to shake things up a little for the sake of discovering new things.

Originally Posted by Philbo King

Yeah, the idea was to stretch an extremely thin diaphragm, like gold-sputtered mylar, bounce a laser beam off it in an optical cavity, and detect the variations in the optical standing wave with a fast photodiode.

You're probably aware of the laser mics used in surveillance. Apparently they are pretty hard to make work even in good circumstances, and the SNR is poor. But the concept is sound, AFAIK.

There was also some research done that translated the imperceptible shift in pixels of objects recorded on video in to sound. E.g. point your video camera at a thin/light/rigid thing, make some noise in the room, and recover the sound by doing computational analysis of the slight variations in the resulting pixels. Very terrible SNR, but you can hear something coming through.

Originally Posted by DVDdoug

There are some rectangular diaphragm

Interestingly (to me anyway), the capsules in the AT5047 are electret elements. It's a good mic to point to when people are describing membrane condeners as "true condensers" and turning their noses up at electrets. :-)

[rogs]

It's easy enough to find cheap 34mm large diameter condenser mic capsules from China (or at least it has been up until recently! ) and incorporate them into low cost DIY microphone projects ...just for fun...
I did this project last year: www.amx.jp137.com/ and that used a large condenser mic capsule - with no DC polarisation voltage at all! ... works pretty well!

[Mr. PC]

Originally Posted by rogs

It's easy enough to find cheap 34mm large diameter condenser mic capsules from China (or at least it has been up until recently! ) and incorporate them into low cost DIY microphone projects ...just for fun...
I did this project last year: www.amx.jp137.com/ and that used a large condenser mic capsule - with no DC polarisation voltage at all! ... works pretty well!

What would be the benefit of that? Lower self noise?

[rogs]

Originally Posted by Mr. PC

What would be the benefit of that? Lower self noise?

The main benefit is the ability to use the mic for outdoor recording without any worries regarding noise problems from moisture ..
Conventional DC bias condenser mics are incredibly high impedance (they normally work into a load of 1Gohm - or even higher) - and that can bring serious problems with humidity, when that type of mic is used outdoors.

Only Sennheiser and Rode currently use the RF concept commercially (see here for more details on the idea: https://assets.sennheiser.com/global...tePaper_en.pdf )

Those mics tend to be expensive, so trying to create a cheap DIY alternative seemed like am interesting project to try out..

And yes, the mic does have a pretty low noise floor (self noise is around 4dB - on a par with the new style Rode NT1 )

[Philbo King]

Originally Posted by rogs

It's easy enough to find cheap 34mm large diameter condenser mic capsules from China (or at least it has been up until recently! ) and incorporate them into low cost DIY microphone projects ...just for fun...
I did this project last year: www.amx.jp137.com/ and that used a large condenser mic capsule - with no DC polarisation voltage at all! ... works pretty well!

Looks like a very cool project! Thanks for sharing.

[rogs]

Originally Posted by Philbo King

Looks like a very cool project! Thanks for sharing.

It's been a fun project --- worked out way better than I first imagined, mostly because of the way I was able to play around with optimising the 'Q' and tuning of the inductor assembly, to get some extra 'noise free' gain...

[Mr. PC]

OK, another idea.

Most multi-pattern mics are made from 2 cardioids, right?

Could you make one from an omni+figure 8?

Couldn't that get you a more accurate response with less proximity effect?

[clepsydrae]

Originally Posted by Mr. PC

Could you make one from an omni+figure 8?

Couldn't that get you a more accurate response with less proximity effect?

Hah, funny, I went down a similar path once and it turns out that that's how engineers got cardioid patterns before cardioid capsules were invented. And IIRC the first cardioid mics were just fig8+omni in a shared housing, but not finding a source on that just now.

Figure 8 mics have high proximity effect, cardioids in the middle, and omnis have none, so the fig8+omni isn't going to have less than a regular cardioid capsule, AFAIK, but not sure if there's a detail there I'm missing. Otherwise, maybe the pattern would be more pure, i.e. more even through the frequency spectrum, as opposed to the standard cardioid capsule method, and overall might match more closely the mathematical ideal of the pattern. But overall there seem to be good reasons that the twin cardioid approach has won the day. Not sure there'd really be a huge benefit to a fig8+omni approach in the real world, but it's always worth experimenting.

[Mr. PC]

Originally Posted by clepsydrae

Hah, funny, I went down a similar path once and it turns out that that's how engineers got cardioid patterns before cardioid capsules were invented. And IIRC the first cardioid mics were just fig8+omni in a shared housing, but not finding a source on that just now.

Figure 8 mics have high proximity effect, cardioids in the middle, and omnis have none, so the fig8+omni isn't going to have less than a regular cardioid capsule, AFAIK, but not sure if there's a detail there I'm missing. Otherwise, maybe the pattern would be more pure, i.e. more even through the frequency spectrum, as opposed to the standard cardioid capsule method, and overall might match more closely the mathematical ideal of the pattern. But overall there seem to be good reasons that the twin cardioid approach has won the day. Not sure there'd really be a huge benefit to a fig8+omni approach in the real world, but it's always worth experimenting.

So it turns out there are still some mics doing this, or even better, 2 diaphragms with a single back plate, so you can mix the pattern more accurately.

https://microphone-parts.com/product...ophone-capsule

All the mics I'm finding that use that capsule are $2,000+, but the capsule is cheap; maybe I should learn to build mics and put it into a cool bottle-case.

I don't really care about the frequency response sooo much, as long as I can EQ it flat; I just want something that is smooth, fast, EQable (meaning not narrow peaks which are hard to EQ out)... so I wonder if this capsule would be better in general. You'd need the right circuitry to mix the two diaphrams into a supercardioid though, right?

Another interesting mic with a mix of capsules

http://www.josephson.com/pdf/srs7.pdf

To expensive for me, and I don't really care about side info, unless there were a way to subtract the sides from the 'center' even further and get a shotgun-like pattern; but I doubt that's possible without harming the signal.

Mostly I'm just thinking about how folks say only omnis can get a pristine/accurate/lifelike sound, so would something like a rk47 capsule be more like an omni in that regard?

[clepsydrae]

Originally Posted by Mr. PC

All the mics I'm finding that use that capsule are $2,000+, but the capsule is cheap; maybe I should learn to build mics and put it into a cool bottle-case.

I'm in full support. If I have a cognitive bias in me, it's that I'm an anti-snob when it comes to mics. I will only grudgingly admit that expensive mics are worth it at all. I do admit it in some cases, but rarely. I admit this bias, but I'm also suspicious that I'm correct. Awaiting further data before I get arrogant. :-)

I made a mic with the RK47 capsule. It used the electronics from a CAD E-350 (I got a few of the boards off eBay years ago, but I don't think they make much difference to the sound). I love the mic. I'd like to make one from an RK-12 or similar next.

I would say: even if you're working from an existing schematic, don't underestimate the time and effort involved in securing the parts and assembling the circuit for a multipattern condenser circuit. Most of the high-yield/minimal-effort/low-cost low-hanging fruit with DIY mics comes in the small/simple stuff, like electret capsules that cost $3 and in conjunction with a couple parts make 80dB SNR super-flat mics that hold their own against multi-thousand dollar mics.

(Ribbon mics are also totally DIY-able because the circuit consists of a single part: the transformer. You can buy the "motors" online and save a ton of time that way. I made my own motor, but that part is much more involved.)

You'd need the right circuitry to mix the two diaphrams into a supercardioid though, right?

Yes.

http://www.josephson.com/pdf/srs7.pdf

Neat. I've lusted after one of these, myself.

Mostly I'm just thinking about how folks say only omnis can get a pristine/accurate/lifelike sound, so would something like a rk47 capsule be more like an omni in that regard?

I wouldn't necessarily agree with the premise; omnis are of course by default the flattest, and thus used for measurement mics, etc., don't have proximity effect, and so on, but in the real world I'd contend that the sense of "realness" has more to do with the room and the mic technique than the mic type. But to your question, if I wanted the most "omni-like" cardioid mic, I'd just look for the flattest response... the RK-47 is not a super flat capsule. See e.g. the RK-12.

So what exactly is the goal? A smooth, flat, "real" cardioid for not too much $?

In terms of cardioid mics, lots of folks are excited about the "Alice" electret mic circuit and related variants, in conjunction with Transsound capsules (TSB2555 or TSB165). (Here's an example.) That's what the Ear Trumpet Labs mics are (+some tweaking and stringent quality control, plus build quality.) Or get a kit (just found this via google). Or buy a cheapo "donor body" mic on Amazon and replace the capsule and circuit with whatever.

If you value your time, it's never worth it (there are so many fantastic and cheap mics now), but if you value the learning and empowerment and the ability to make it how you like it, it totally is. :-)

[Mr. PC]

Originally Posted by clepsydrae

I'm in full support. If I have a cognitive bias in me, it's that I'm an anti-snob when it comes to mics. I will only grudgingly admit that expensive mics are worth it at all. I do admit it in some cases, but rarely. I admit this bias, but I'm also suspicious that I'm correct. Awaiting further data before I get arrogant. :-)

I made a mic with the RK47 capsule. It used the electronics from a CAD E-350 (I got a few of the boards off eBay years ago, but I don't think they make much difference to the sound). I love the mic. I'd like to make one from an RK-12 or similar next.

I would say: even if you're working from an existing schematic, don't underestimate the time and effort involved in securing the parts and assembling the circuit for a multipattern condenser circuit. Most of the high-yield/minimal-effort/low-cost low-hanging fruit with DIY mics comes in the small/simple stuff, like electret capsules that cost $3 and in conjunction with a couple parts make 80dB SNR super-flat mics that hold their own against multi-thousand dollar mics.

(Ribbon mics are also totally DIY-able because the circuit consists of a single part: the transformer. You can buy the "motors" online and save a ton of time that way. I made my own motor, but that part is much more involved.)

Hmmm, how do you make a super flat high SNR mic for cheap? Does it really sound accurate? Like, maybe the diaphragm is slow or muddy in places? I guess it'd be omni? I really am not a DIY guy, I just want a good mic to make music with (I only like to compose really).

Neat. I've lusted after one of these, myself.

So that's 2 figure 8s being summed into a multi-pattern? Also very interesting. It seems you could sum pretty much anything except 2 omnis? Or I guess you could? Maybe that'd be the most accurate?

I wouldn't necessarily agree with the premise; omnis are of course by default the flattest, and thus used for measurement mics, etc., don't have proximity effect, and so on, but in the real world I'd contend that the sense of "realness" has more to do with the room and the mic technique than the mic type. But to your question, if I wanted the most "omni-like" cardioid mic, I'd just look for the flattest response... the RK-47 is not a super flat capsule. See e.g. the RK-12.

So what exactly is the goal? A smooth, flat, "real" cardioid for not too much $?

Actually, I prefer super cardioid. I like very very dry recordings (and I travel always). I've got a laptop and Tascam 2x2, so I could sum 2 inputs. I was thinking of getting a DPA 4099, because it's clip on, dry, flat, and easy to position, and then maybe combining that with my m201 or nd967, which I already have (maybe not even use them, but since I have a second input, might as well for backup. The m201 is supposed to have the same capsule as an MD441, right? That's what I thought when I bought it, but now I can't find that info).

But then I started thinking about the interference tube in the 4099, and thinking how it's likely only flat because of circuitry, and has a low SNR (and in my past recordings, preamp/mic noise has been a small issue. Recordings are fine, I would just prefer no noise so I can freely EQ the highs).

I've been thinking of the Schoeps CCM 41, which is accurate supercardioid, small, and OK SNR. Or a mkh-40, which is similar, though slightly bigger. If getting a good SNR and detail means a big mic, I'd get big, but I prefer small (or alternatively, awesome looking, and shaped in such a way I could position it easily without bending the XLR, like this)... but those are the least important. Still, if I were to build my own mic, would be cool to put it in such a case, and I'm sure the shock mount here is great.

In terms of cardioid mics, lots of folks are excited about the "Alice" electret mic circuit and related variants, in conjunction with Transsound capsules (TSB2555 or TSB165). (Here's an example.) That's what the Ear Trumpet Labs mics are (+some tweaking and stringent quality control, plus build quality.) Or get a kit (just found this via google). Or buy a cheapo "donor body" mic on Amazon and replace the capsule and circuit with whatever.

If you value your time, it's never worth it (there are so many fantastic and cheap mics now), but if you value the learning and empowerment and the ability to make it how you like it, it totally is. :-)

I actually prefer supercardioid to cardioid... I want as little 'room' as possible. I tend to make frankenstein recordings, mixing parts from things I've made over the years, so having as much dry info as possible means I can shape it to fit in whatever I have, and don't have to worry about a bad room or room-noise.

I don't want a mic to 'warm' 'brighten' or do anything nice to the sound; I can always do that with EQ... for me it's just accuracy and detail (actually, even if the mic weren't flat, if the frequency response were smooth enough that I could EQ it, or send it to a company to calibrate flat and give me a ReaEQ file) that would be find. Maybe having a boost in the highs is better, because that's also the area with preamp/mic noise, so turning that down in the DAW would help remove noise, right?


Now I'm starting to think of something like the 4099, but an omni (if omnis are really a 'cleaner' signal). But if that's untrue, then I'm back to square 1.

This is the thread where people are telling me that for accuracy you need omni, btw. They also suggest I buy a variety of different mics for different purposes, but that's not at all what I want, since I've no interest in being a recording studio. I just want to piece together soloists (my next project I'm wanting to record ethnic instruments in Russia, and combine them with some classical and electro).

https://www.gearslutz.com/board/remo...nse-mic-4.html

Maybe I should be looking at some kind of clone?

[clepsydrae]

Originally Posted by Mr. PC

Hmmm, how do you make a super flat high SNR mic for cheap? Does it really sound accurate? Like, maybe the diaphragm is slow or muddy in places? I guess it'd be omni?

I shouldn't have said "super-flat" without qualifying it as "super flat considering the cost and simplicity." The mics I'm describing aren't super-flat compared to $4k "super-flat" mics. It sounds like your standards might be a tad higher/different than what I'm describing (you describe the 80 dB SNR of the CCM 41 as "OK", implying that you'd like better), and the off-axis response of the mics I describe probably isn't at the level of pro mics, and the pattern isn't as narrow, etc, but just for fun, here's a comparison I did a while back:

This is between a Primo EM184 cardioid capsule (they used to go for like $15 but are no longer made but can be had for $40 or so I think) against a DPA d:vote 4099.

mic A: mp3 | wav
mic B: mp3 | wav

My ABX tool (see sig) can be useful for such comparisons.

Those are EQ'd a little bit. The DPA 4099 is not really a flat mic; it rolls off the bass some and has a slight treble boost (entirely appropriate for an instrument mic). The point wasn't to compare raw recordings but to see if the DIY mic could match the DPA with some EQ. I think it succeeded. Which do you prefer? Can you tell which is which in a blind test with the ABX tool?

I really am not a DIY guy, I just want a good mic to make music with (I only like to compose really).

Yeah if you're not pumped on DIY it's not worth the time/effort. After all, you can buy fantastic used mics on eBay for shockingly little money these days.

And if you want a super-flat >80 dB SNR super-cardioid with controlled off-axis response and high SPL tolerance and good sensitivity in an attractive package... DIY is going to be a long, hard road.

That said, it's always worth putting in perspective how important those requirements really truly are, which is why I posted the above demo. :-)

So that's 2 figure 8s being summed into a multi-pattern?

I don't understand it all, but apparently it's a dual-ribbon mic where only one of the ribbons is used at once, depending on the pattern. Details here. There are intricate highly-engineered acoustic pathways being mechanically switched around to generate the patterns, as well as switching between the ribbon elements.

Also very interesting. It seems you could sum pretty much anything except 2 omnis? Or I guess you could? Maybe that'd be the most accurate?

Summing two omnis sounds like little-to-no benefit to me... I'm not an expert but i suspect that phasing issues would far outweigh whatever small benefit might be achieved in SNR and smoother overall response...

Maybe having a boost in the highs is better, because that's also the area with preamp/mic noise, so turning that down in the DAW would help remove noise, right?

Not sure if there's truth to this. Sounds reasonable in principle!

This is the thread where people are telling me that for accuracy you need omni, btw.

If one was to pick a mic at random from the ocean of mics, omni mics are certainly, on average, flatter and truer. But IMO your preferred pickup pattern should easily outweigh the subtle differences in high frequency response. E.g. the MK41 you pointed to is flat as heck, so, problem solved, except for the cost.

In general, the off-axis response is a consideration, SNR is a consideration, etc. But you can always just check the spec sheet to see if it meets your needs (when/if you can trust the spec sheet, of course.)

My conviction is that anything we're ascribing to the subtle nuances of the high frequency response on mics (within that pro class) is much more likely to be a combination of imagination and random variation in performance/technique/etc.

[Mr. PC]

Ya, I can't tell any difference.

I can hear some unpleasant room in the guitar, but that's in both. (I'm listening on a bluetooth pool speaker, since my monitors broke. That'll be another thread). There was maybe some resonance around 4k in track A in the accordion? But I'm really trying there; really I think they sound exactly the same, which is remarkable since you couldn't have placed them exactly the same, and I'd have imagined mounted on the instrument there should be a bigger sound difference just from placement?

Also, one being cardioid, and the other supercardioid with an interference tube, I'd have thought the 4099 would be way way drier, but I can't hear that at all.

Also, you happen to have the mic I was looking at, and play a bunch of the instruments I wanted to record!! Wow.

I wonder if liking a lot of old music makes me less sensitive to low sound-quality? As a kid, I thought old music sounded bad (as in, low-fi) but now I don't even notice it, and find I instantly dislike new recordings, even before I hear the content, just based off sound quality.


So... maybe I should stick with my m201 and ev967, add an SE-Electronig Dynamite for less preamp noise, and just EQ them flat? Or maybe I should just buy any clip-on mic? In that thread, someone said the m201 sounded "woolly", but maybe that's just an EQ issue? I'm amazing how many aspects of sound are simply a matter of EQ.

[clepsydrae]

Originally Posted by Mr. PC

There was maybe some resonance around 4k in track A in the accordion? But I'm really trying there; really I think they sound exactly the same, which is remarkable since you couldn't have placed them exactly the same, and I'd have imagined mounted on the instrument there should be a bigger sound difference just from placement?

It's been a long time since I recorded those, but IIRC the mics weren't mounted on the instruments, but they were positioned in a similar spot. That hides one "problem" with the DIY mic which is that being full-range it picks up low-frequency body sounds more than the DPA mic which rolls the bass off, but it's easy to solve with EQ, even in a live setting. Since both mics are so tiny it wasn't hard to place them in the "same" spot; I would have spaced them enough to allow the interference tube to work, though (maybe an inch apart?, but again, it was a long time ago.)

I find that accordions (that's actually a melodica, but same idea) and banjos are great for drawing out differences in mics, what with all the upper partials.

I can DM you which mic is which when/if you want.

Also, one being cardioid, and the other supercardioid with an interference tube, I'd have thought the 4099 would be way way drier, but I can't hear that at all.

I dunno, personally I wouldn't really expect too a noticeable difference with close-micing. It's akin to how omni mics can be used even on stage sometimes; if they're close to the source, the ratio of instrument signal to ambient signal is really high. I would expect better feedback rejection from the DPA, but I never tested that. (In retrospect, I should have. But the DIY mic is totally usable on stage, at any rate.)

Also, you happen to have the mic I was looking at, and play a bunch of the instruments I wanted to record!! Wow.

Yeah, that worked out, huh? :-) I don't actually own the DPA, but had it on a long loan at that time. It was really handy to compare the instrument mics I was making to a pro offering.

I wonder if liking a lot of old music makes me less sensitive to low sound-quality? As a kid, I thought old music sounded bad (as in, low-fi) but now I don't even notice it, and find I instantly dislike new recordings, even before I hear the content, just based off sound quality.

I've definitely found that friends of mine in the old-time and folk communities like lo-fi cardboard-sounding recordings with zero reverb (real or artificial). E.g. recordings made in smallish dead rooms with dynamic mics. They also seem to be really attuned to mid-range tones, not care about bass, and don't like treble. I think this is why the Edwina from ETL is so popular in those circles [after aesthetics, which is always the real reason with mics :-)] -- it's mid-forward (well, in the sense that it rolls off the bass and low-mids and doesn't boost the highs significantly), and that aligns with all the recordings that they listen to. (Great mic, don't get me wrong, just not my choice personally except as an up-close vocal mic.)

And of course everyone listens to music on 0.5" speakers nowadays so, whatever. :-)

[Mr. PC]

Hmmm, listening to that mic, it does sound nice to me... but probably the guitar and room just sound nice. I find when I mic guitar the 'proper' way, angling toward the 12th fret etc., it sounds really 'plastic' to me. I get the same feeling from the internal pickup. Maybe just the HF information up close are what make me feel 'plastic'. And ya, I remember making some songs where I rolled off above 10k, because it just sounded more "smooth/round" to me... and I got horrible feedback from mixing engineers.

So it seems 90% of the sound profile (not counting room) is the EQ curve, which can be done ITB, and the other 9% is noise. Do you think what people say about large diaphragms being 'slow' and 'muddy' is bogus, and it's just a frequency-balance thing?


I'd love to know which is which (though I still don't hear a difference), either by PM or if you could message me through my site, since my inbox is almost full, even better.

http://albertmckay.com/


Now I'll go write some sad music, because I might never see my student again due to Covid). He was like my brother.

[clepsydrae]

Originally Posted by Mr. PC

So it seems 90% of the sound profile (not counting room) is the EQ curve, which can be done ITB, and the other 9% is noise.

I've heard the claim about "micro peaks" in the frequency response, and it makes sense to me, though I haven't read anything definitive on it. The idea that ribbons are "easier to EQ" as a result of that also bears out in my experience.

Consistency of the polar pattern off-axis seems like a reasonable thing to think about as well. And yeah of course noise, EMF rejection, etc.

Other than that, the only other characteristic I'm aware of (unless I'm forgetting something) that affects the "sound" of the mic could be what you bring up:

Do you think what people say about large diaphragms being 'slow' and 'muddy' is bogus, and it's just a frequency-balance thing?

I've heard educated people claim that the frequency response tells all, and that the idea of transient response being different, given equal frequency response, is a myth. I've heard other educated people claim that SDC vs LDC matters in part because transients are temporally smeared in the LDC due to the large diaphragm. The "frequency response" people then counter with the reasonable-seeming argument that "transient response" is just a muddled way to think about frequency response; that is, if a diaphragm responds quickly to a transient, that's equivalent to having response in the higher frequencies (Fourier, etc.)

There's also the question of diaphragm ringing post-transient; i.e. nonlinearities in the response that can smear transients through time. This basic physics-of-mics stuff is at the frontier of my mic knowledge, and has been next on my to-do list in terms of research for a long time. :-)

My personal experience backs one aspect of the SDC/LDC difference: to my ear, SDCs do indeed seem to capture transients of, say, guitar or mandolin, in a more "crisp" fashion than do LDCs, even if the tested frequency responses of the mics seem to be comparable. Might just be confirmation bias, or my shortcomings as a recordist/mixer when trying to make a LDC sound like a SDC, but it sure seems true.

Now I'll go write some sad music, because I might never see my student again due to Covid). He was like my brother.

Good health to you both!

[clepsydrae]

(sent a PM -- your website contact form didn't work for me -- gave an error: "Form PHP script is missing from web server, or PHP is not configured correctly on your web hosting provider. Check if the form PHP script has been uploaded correctly, then contact your hosting provider about PHP configuration.")

[Mr. PC]

Thanks for the PM!! Ya, I gotta update my site, but lost all the original content, which would mean rebuilding from scratch.

I wonder why there are no mics shaped like a coax speaker.



Also, about the ribbon - that makes sense, since ribbons are the simplest design. I wonder if there could be some kind of low-noise small 'diaphragm' ribbon?

Close-mic with it, and maybe stick a large wedge behind it to stop info coming in the rear?

Or maybe 2 ribbons summed to create a cardioid/super cardioid?

[clepsydrae]

Originally Posted by Mr. PC

Also, about the ribbon - that makes sense, since ribbons are the simplest design. I wonder if there could be some kind of low-noise small 'diaphragm' ribbon?

Yeah as with LDCs I'm not personally clear on the true effect that ribbon size could make. I know that a shorter ribbon = less magnetic flux = less voltage = more noise, but if you were willing to sacrifice some aspects (noise, frequency response) maybe you could get "truer" transients or something.

Or maybe 2 ribbons summed to create a cardioid/super cardioid?

I'm not aware of a way to combine two theoretical figure 8's to produce anything other than another figure 8, an omni, or silence, or points in between.

[Mr. PC]

Originally Posted by clepsydrae

Yeah as with LDCs I'm not personally clear on the true effect that ribbon size could make. I know that a shorter ribbon = less magnetic flux = less voltage = more noise, but if you were willing to sacrifice some aspects (noise, frequency response) maybe you could get "truer" transients or something.

I'm not aware of a way to combine two theoretical figure 8's to produce anything other than another figure 8, an omni, or silence, or points in between.

Couldn't you

Front @ +60V + Rear = -30V?

but I guess a figure 8 and an omni would also work, using the fig. 8 as the main diaphragm for a smoother response?


Anyway, this is all just speculative talk now... I'm back at square 1 figuring out which mic to buy.

[clepsydrae]

Originally Posted by Mr. PC

Couldn't you

Front @ +60V + Rear = -30V?

but I guess a figure 8 and an omni would also work, using the fig. 8 as the main diaphragm for a smoother response?

Hmm... you proposed combining two ribbon mics to form cardioid or super-cardioid patterns, which is what I was responding to... they don't have a "front" or "rear" capsule, and don't use polarization voltages...

But yeah a ribbon+omni to get cardioid would in theory benefit from the smoothness of the ribbon, presuming there's truth to that idea.

I'm back at square 1 figuring out which mic to buy.

I say: mics and their effects are overrated, and don't stress it. :-)

[Mr. PC]

Originally Posted by clepsydrae

Hmm... you proposed combining two ribbon mics to form cardioid or super-cardioid patterns, which is what I was responding to... they don't have a "front" or "rear" capsule, and don't use polarization voltages...

But yeah a ribbon+omni to get cardioid would in theory benefit from the smoothness of the ribbon, presuming there's truth to that idea.



I say: mics and their effects are overrated, and don't stress it. :-)

So I'll just keep using my dynamics? Do you think the realism I heard from the Earthworks was a placebo, or just the placement?

I've had issues with preamp noise (that high frequency slanted noise; clearly not room-noise).

Maybe just any kind of mounting-mic so I can get ultra-close?

[clepsydrae]

Originally Posted by Mr. PC

So I'll just keep using my dynamics? Do you think the realism I heard from the Earthworks was a placebo, or just the placement?

I've had issues with preamp noise (that high frequency slanted noise; clearly not room-noise).

Maybe just any kind of mounting-mic so I can get ultra-close?

Well, lemme walk that back a little bit. I don't mean that mic choice is truly irrelevant. As mentioned, I like the sound of SDC's on transient sounds, for example. The choice matters, but I think that once you've narrowed it down to a certain technology (dynamic, SDC, LDC, ribbon, whatever), a pattern (or patterns), and a set of specs you want to meet (noise, frequency response, EMI resistance, etc), that most mics that fall in to that category and which have good reviews are going to be more or less identical, or close enough not to stress it.

So the difficulty (as you already know) is deciding how to weigh those choices, and one of my main points is not to focus too much on the potential supposed "drawbacks" of the categories. E.g. if you want a tight pattern, don't worry about the fact that omni mics tend to be flatter/truer, just get the flattest response you can find in a tight-pattern mic: it'll be fine. Or, if all you want is a flat/true mic, don't worry that ribbon mics are reputed to be smoother/easier to eq, just get a flat omni that is reputed to be smooth, etc.

If you're struggling to balance all these choices, my personal opinion is that pattern is most important, followed by frequency response, followed by SNR, followed by other stuff (pattern consistency through the frequency range, sensitivity (in the electrical spec sense), EMI, durability, etc.)

Digesting everything you've written here, my guess would be that you'd be happiest with a nice-n-flat SDC pencil mic of some kind, like the ones you listed or any alternative. Some are ruler flat, should (in theory) be easy enough to EQ, are low-profile, durable, easy to travel with, etc.

I wouldn't want a DPA instrument mic to be my main field recording mic. Not having the flexibility in mic position would be a big drawback. E.g. some instruments (fiddle, banjo) don't usually want to be mic'ed close if you can help it. And what do you do if they are playing a flute, or singing, etc. But if you are cool with that drawback, they're certainly fine mics and are super-small and low-profile and have tight patterns.

But yeah, a nice stereo-matched pair of quality SDCs is a classic choice for flexible/honest field recording. If you fear the pattern may be too wide, you can push in a little and thus lower the amount of ambient sounds.

[Mr. PC]

Originally Posted by clepsydrae

Well, lemme walk that back a little bit. I don't mean that mic choice is truly irrelevant. As mentioned, I like the sound of SDC's on transient sounds, for example. The choice matters, but I think that once you've narrowed it down to a certain technology (dynamic, SDC, LDC, ribbon, whatever), a pattern (or patterns), and a set of specs you want to meet (noise, frequency response, EMI resistance, etc), that most mics that fall in to that category and which have good reviews are going to be more or less identical, or close enough not to stress it.

So the difficulty (as you already know) is deciding how to weigh those choices, and one of my main points is not to focus too much on the potential supposed "drawbacks" of the categories. E.g. if you want a tight pattern, don't worry about the fact that omni mics tend to be flatter/truer, just get the flattest response you can find in a tight-pattern mic: it'll be fine. Or, if all you want is a flat/true mic, don't worry that ribbon mics are reputed to be smoother/easier to eq, just get a flat omni that is reputed to be smooth, etc.

If you're struggling to balance all these choices, my personal opinion is that pattern is most important, followed by frequency response, followed by SNR, followed by other stuff (pattern consistency through the frequency range, sensitivity (in the electrical spec sense), EMI, durability, etc.)

Digesting everything you've written here, my guess would be that you'd be happiest with a nice-n-flat SDC pencil mic of some kind, like the ones you listed or any alternative. Some are ruler flat, should (in theory) be easy enough to EQ, are low-profile, durable, easy to travel with, etc.

I wouldn't want a DPA instrument mic to be my main field recording mic. Not having the flexibility in mic position would be a big drawback. E.g. some instruments (fiddle, banjo) don't usually want to be mic'ed close if you can help it. And what do you do if they are playing a flute, or singing, etc. But if you are cool with that drawback, they're certainly fine mics and are super-small and low-profile and have tight patterns.

But yeah, a nice stereo-matched pair of quality SDCs is a classic choice for flexible/honest field recording. If you fear the pattern may be too wide, you can push in a little and thus lower the amount of ambient sounds.

Do you think getting a little closer with a cardioid would be safer than using a supercardioid?

Like, getting an equal amount of 'room sound' from both, the cardioid would in general pick up more of the instrument?

e.g. I'm recording a violin as dry as possible. Using a supercardioid, if the performer moves 1 foot off axis, the sound changes more than using a supercardioid... although I guess if you're putting the cardioid closer to keep the amount of 'room' consistent, you'll have the same issue?


Also, should I just keep using my small diaphragm dynamic? Or a condenser will really be more detailed? (The idea that dynamics pick up less room is a myth, right?) Have you ever used an m201?

[clepsydrae]

Originally Posted by Mr. PC

Do you think getting a little closer with a cardioid would be safer than using a supercardioid?

Like, getting an equal amount of 'room sound' from both, the cardioid would in general pick up more of the instrument?

Hmmm I'd really have to test it on a given instrument, and I haven't done so. I bet it would be roughly a wash... Like, a shotgun four feet back, vs a cardioid close enough to get a similar amount of room noise... they'd obviously sound different (e.g. proximity effect of the cardioid) but in terms of how they hear the instrument... would be a little more position-dependent with the cardioid but I bet you could get the same sounds from both arrangements with some care and EQ.

When you say "supercardioid" are you imagining like a handheld SDC or more like a shotgun mic?

I've used the Rode NTG3B (location dialog recording) and it was great. Very natural sound, no sacrifices for the on-axis frequency response (not totally flat, but close), etc. They often use a shotgun mic as the central mic in the NPR tiny desk concerts. Kind of a dry, focused, sound, IMO, but sounds like that's what you're after.

e.g. I'm recording a violin as dry as possible. Using a supercardioid, if the performer moves 1 foot off axis, the sound changes more than using a supercardioid... although I guess if you're putting the cardioid closer to keep the amount of 'room' consistent, you'll have the same issue?

I'd have to test to be sure of an opinion on that one, but my gut is that, given equal-room-noise distancing, a shotgun would be a little bit touchier, but that's a guess.

Also, should I just keep using my small diaphragm dynamic? Or a condenser will really be more detailed?

Yeah as mentioned before I have personally come to the opinion that SDC condensers are the best for clear transients, which is what everyone says, but I don't have a strong theoretical grounding to explain it (see previous discussion re: Fourier, etc.) I personally avoid dynamics because I've haven't liked how they sound, but I haven't used many nice ones, so I shouldn't weigh in on that. As usual, I would advise that if you like what you hear, keep it and use it. And if you suspect there may be benefits to some other type, borrow or rent or buy/return a mic and try it out -- might be the only way you can really answer any of these questions.

(The idea that dynamics pick up less room is a myth, right?)

Yeah, all things being equal, it's a myth. But there is obviously a historical pedigree of dynamic mics being stage mics, and so their patterns were traditionally tight, their HF response not as strong, etc, as opposed to LDCs/SDCs being more recording mics, so their patterns might not be as tight and they were sensitive and bright. I think that's pretty much all vestigial mythology at this point, though. Modern studio and stage dynamics and condensers can all crossover just fine. If you have a trustworthy spec sheet in hand, no need to worry about that particular historical reputation.

Have you ever used an m201?

Nope.

[Mr. PC]

So I'm not really interested in shotguns, because I feel the interference tube would colour the sound... so ya I'm thinking a handheld SDC. Probably the schoeps CCM41 is my ideal mic (but then again, is there CCM22 more accurate? The response is flatter, but such a small difference, I probably would never hear it... although I guess the CCM41 wouldn't be good for bass? Or maybe it would if I got close enough for the proximity effect to flatten the response out?

So, about placement, being going a bit psycho trying to get a perfect mic placement on the acoustic guitar on my last album worsened my wrist injury (holding the guitar in the perfect position relative to the mic). I can't release the album yet until the art's finished, but this is an example track (and the guitar tone isn't even that great... plus using Soothe on it). I liked the idea of the clip-on because sound wouldn't change if I move around while playing (or ditto for whoever I'm recording).

https://soundcloud.com/phryq/coffee-.../s-3jw3IaHcKQY


Now I've got good monitors, and still those 2 recordings you made (between the 4099 and your self-made mic) sound very very similar. If anything, I'd say your mic sounds better (not saying which is which). Maybe my ears just aren't so good to tell the difference though...

Actually, I really love the look of the Sontronics Saturn... but I guess I shouldn't go for looks.

http://recordinghacks.com/microphones/Sontronics/Saturn


So... would I be right to say that the mic really isn't important, other than SDC, LDC, dynamic, and polar pattern? And that positioning/placebo accounts for what most people consider the difference between mics?

Like, when I've used expensive mics in the past, and thought they sounded great, was it just a combination of slightly different placement and a placebo affect?

[clepsydrae]

so ya I'm thinking a handheld SDC. Probably the schoeps CCM41 is my ideal mic (but then again, is there CCM22 more accurate? The response is flatter, but such a small difference, I probably would never hear it... although I guess the CCM41 wouldn't be good for bass? Or maybe it would if I got close enough for the proximity effect to flatten the response out?

Yeah I think those response differences would be very easy to EQ... it's only a couple dB down in the bass. I think you'd hear the difference pretty easily if recording something bassy, but a little boost and you could match 'em easy, I'd think (and/or use proximity, yeah). Obviously the polar pattern in the CCM41 looks closer to what you've been describing. Personally if I had to choose one I'd probably take the CCM41 (both seem like great mics.)

I liked the idea of the clip-on because sound wouldn't change if I move around while playing (or ditto for whoever I'm recording).

Those Schoeps are so tiny you could probably clip those on pretty easy, too, eh?

So... would I be right to say that the mic really isn't important, other than SDC, LDC, dynamic, and polar pattern? And that positioning/placebo accounts for what most people consider the difference between mics?

Like, when I've used expensive mics in the past, and thought they sounded great, was it just a combination of slightly different placement and a placebo affect?

As mentioned before, I'm an admitted anti-snob when it comes to this, so I tend to pooh-pooh the fetish for expensive mics, and I'm maybe being a little too quick to do so, especially lacking personal experience with the highest-end mics and not having compared every conceivable pair of mics, etc. (E.g. I've never used a Schoeps, so what do I know?) I do think that placebo can easily outweigh many of the differences people are listening for, not to mention performance variation and placement, and that it's very unwise to trust the opinions of people that haven't blind tested themselves extensively when making such evaluations (e.g. almost everyone that has opinions about audio.)

If e.g. the Schoeps CCM 41 is the feature combo you like, but the price tag is off-putting, I would argue that if you find a mic with a reliable* spec sheet that matches it (close enough anyway), and that sells for less $ and has good reviews, you would very likely not be able to tell them apart when blinded. The reliability of the spec sheet is of course tricky, so issues like the "mini spikes that don't show up on the frequency plot" theory, or off-axis coloration or build quality or whatever are worth considering, but my personal style is to be open to cheaper options, when they are informed.

E.G. the KSM137 looks to be extremely flat (with flatter bass than the Schoeps) except for a moderate bump at 9k, which I bet could be EQ'd easily. It goes for $300. A little bit larger. Same noise spec as the Schoeps, and like the Schoeps has a very well-controlled pattern. But it's cardioid, not supercardioid. (The KSM141 is ~the same and adds an omni option.)

If I had both I'd love to test the differences, but I bet they'd be inaudible after EQ, aside from the pattern difference, which itself may not be crucial. To me the question then becomes whether changing to a supercardioid pattern is worth $1500.

[Geoff Waddington]

I have been fortunate to have recorded with some really good pics, not ALL of the "holy grail" mics, but enough to know, and yeah, there really is a difference IF you have extremely good monitors in an extremely good room and an extremely good console.

Can you record fabulous tracks with average mics -- ABSOLUTELY.

Tom Dowd: "The mic is there to capture, not to interfere"

One under reported, very important, factor is the basket design, it's REAL easy to get a lot of nasty HF resonances there, the dimensions are small.

Given your interest in capsules, etc., you really should check out the Townsend Sphere L22.

It is a modelling mic system that allows you to record the front and back capsules separately, and then adjust polar patterns AFTER recording.

It's also a great sounding mic on its own

[cyrano]

If you are serious about mic DIY, you might consider joining the micbuilders list. That's where the DIY crowd and a lot of professional designers hang out. It's low traffic, high-quality content:

https://groups.io/g/MicBuilders

Since it's gone from Yahoo to groups.io, it has gotten a lot more accessible and reliable. You can read the archives online, if that's more to your liking.

You might discover that getting the right parts is harder than anything else, sometimes

[cyrano]

Another great series to watch, if you're serious about mic DIY:

https://www.youtube.com/watch?v=_rTA...260yJBRQngBKiw

[Mr. PC]

That group looks a bit over my head, but I might ask someone to build me a mic for $$.

This post

"Typically semiconductors optimized for switching compromise linearity- so I was grumbling about the "chopper/switching use" parts. They are optimized for speed. Fairchild Process 51 used for J111, J112, J113, PF5102, BF246, BSR58, MMBF5103, U1897 to name a few.

It's very difficult to compare JFETS because the same part number will have different specs/data between manufacturers, and tests have different parameters. I think the majority here are high-Z applications, that RF excitation circuit looks nice though.

The 2SK3666 I'd earlier mentioned is a bit noisy, but 2SK3557 was pretty quiet in the EDN JFET noise test.

TO-92 package is being abandoned by the major semi manufacturers, leaving the boutique like Linear Systems, Calogic, InterFET, or the jump to SMT."

"https://groups.io/g/MicBuilders/message/30248"


No way I'm gonna go down the rabbit hole of transformers, but I guess if you have a bit of lag, you can get a flatter transformer?

I'd be find for that, since I'm not using the mic live, and don't need instant monitoring.

[Geoff Waddington]

Originally Posted by Mr. PC

That group looks a bit over my head, but I might ask someone to build me a mic for $$.

This post

"Typically semiconductors optimized for switching compromise linearity- so I was grumbling about the "chopper/switching use" parts. They are optimized for speed. Fairchild Process 51 used for J111, J112, J113, PF5102, BF246, BSR58, MMBF5103, U1897 to name a few.

It's very difficult to compare JFETS because the same part number will have different specs/data between manufacturers, and tests have different parameters. I think the majority here are high-Z applications, that RF excitation circuit looks nice though.

The 2SK3666 I'd earlier mentioned is a bit noisy, but 2SK3557 was pretty quiet in the EDN JFET noise test.

TO-92 package is being abandoned by the major semi manufacturers, leaving the boutique like Linear Systems, Calogic, InterFET, or the jump to SMT."

"https://groups.io/g/MicBuilders/message/30248"


No way I'm gonna go down the rabbit hole of transformers, but I guess if you have a bit of lag, you can get a flatter transformer?

I'd be find for that, since I'm not using the mic live, and don't need instant monitoring.

If I understand you correctly, don't worry about transformer lag in terms of performance, it's orders of magnitude smaller than DAW buffer latency.

[cyrano]

The micbuilders group welcomes beginners.

That particular discussion about fets isn't typical. It's about the pain involved finding the right fet for optimal specs. These days, some typical audio fets are no longer available. As a beginner, you''l get by with what's available though.

And that list is the home of the Alice mic you mentioned, which is easy to build and affordable.