Large Diaphragm Condenser Microphone for Room Acoustic Measurements

[JTee]

Hi everyone,

Been a long time lurker on this forum and found the advice not only helpful, detailed and accurate, but also considerate. Hoping for some of that here! If I make any obvious errors, for instance if I'm asking this in the wrong sub-forum, please let me know.

I'm starting to treat my room, which has been long overdue, and need some advice when it comes to room measurements. From what I've read, small diaphragm condensers with both an omni polar pattern and calibration file are considered the best for measuring room acoustics, with the UMiK-1 being preferred by many. However, a lot of the discussion seems to be around the response in the high range, as this is where most measurement microphones differ from recording mics. I was wondering if, therefore, a large diaphragm condenser like my Aston Spirit (which has an omni polar pattern) would be sufficient in identifying modal problems. This would be going into my Focusrite Clarett 4Pre USB.

Trying to make my money stretch as much as possible here. I would like to avoid spending £100 on a microphone that I won't use for recording after the room is set up, but at the same time I want to get something that at least shows me where some of the problems lie. £100 saved here would be £100 spent on further treatment, most likely!

The final thing I'd point out is this room is always going to be a mess acoustically: The room measures 4.33m x 4.32m, but along one wall there is a 1.32m x 0.98m cutout for a door. So, basically a square with a chunk out of one corner. There's also another door in the opposite corner, an off centred window on the other side of the room, and the ceiling can't support a cloud (at least, it's bowed and it's come down in another room in the house. Really don't fancy testing the weight tolerance with my head and computer as the measuring stick!). The walls are either 6 inches of brick (external) or sandstone (internal), the only stud wall being the cutout, and I'm on the second story, which I've heard isn't the best for acoustics.

In short, all I want is a rough guess as to where I need to target treatment to make this room as good as it can be, so I feel that spending money on a dedicated microphone would almost be a waste. Agreed, the argument could be made that treating it at all is also a waste, but I felt that if I move into a new room I could at least take the treatment with me and buy a dedicated microphone then when I have the funds to be able to do this properly. Right now I feel that perfection (or what is considered perfection in room acoustics) is so far away, that more treatment would probably go further than a device to show me just how hopeless my mission is!

Thank you for your time,
Jake.

 

[Curvature]

The LF drops early and the HF is a little uneven.

The main reason UMIKs or Earthworks microphones are used is that they come with calibration files that ensure flat response in the audible band and are consistently omni, rather than omni-ish. Small diaphragms have generally better polar response at the cost of higher self-noise.

 

[kemmler3D]

Hey, welcome to ASR!

Just for treating low frequency modes your studio mic might be adequate, but calibration is more important than mic quality... It's the limiting factor on how good the room correction can be. You probably need the response to be accurate to +/- 1dB if possible. From that point of view, it may even be preferable to use a iPhone if the LF accuracy of your mic isn't known to that degree.

e: @Curvature s post makes me think the Aston mic isn't going to work for this. -20dB/Oct roll off (before any unit specific variations) means you are almost flying blind below 100hz which is fairly critical for this job.

 

[JTee]

Thank you both for your responses!

I think I'll just bite the bullet and buy the UMiK-1. Wish I had the money for an Earthworks mic! While the graph that Curvature found is the stated frequency response from Aston Microphones, I've seen other people quoting a 20Hz-20KHz (+/-3dB) response. This also seemed to me to be more likely, as I have had audio (albeit rolled off) down to 20Hz recorded by the mic. Could this be interference? There's always some information down at that range which I remove with RX. Just curious. Regardless, I think I just needed the push.

Anyway, thank you both again for your prompt responses.

 

[DVDdoug]

Low frequencies are where the big problems usually are, and they are harder to correct by-ear. High frequency EQ can generally be more broad (less precise).

Ethan Winer says:

Rooms vary by 30dB, and even modest microphones vary much less... In my experience, (measurement) microphones don't need to be calibrated unless you're a professional acoustician who's being paid to provide accurate readings.

He doesn't exactly say you don't need a measurement mic, he just says it doesn't have to be calibrated. And he is a "paid professional".

 

[sergeauckland]

Low frequencies are where the big problems usually are, and they are harder to correct by-ear. High frequency EQ can generally be more broad (less precise).

Ethan Winer says:

He doesn't exactly say you don't need a measurement mic, he just says it doesn't have to be calibrated. And he is a "paid professional".

If you're trying to equalise a loudspeaker to be flat pseudo anechoically, then you'll need a calibrated mic, as the microphone errors are likely to be greater, or at least equal to the errors on the loudspeaker you're trying to equalise. On the other hand, for in-room measurements, it's the other way round, in that the errors you're correcting for are likely to be much greater than any error in the microphone, even without calibration. I bought a cheap Behringer 'measuring' microphone and calibrated it against a borrowed calibrated Earthworks microphone when I did my loudspeakers, but for everyday in-room use, the Behringer is fine as it is. Calibrating a mic at home isn't trivial, and if I had costed my time in calibrating the cheap mic, I could have bought the Earthworks twice over...

If you can go for the Umik, that will give you the most accurate results, and it's something you only need to pay for once, as the mic will be with you forever.

S.

 

[Curvature]

Ethan Winer

He wasn't right about everything.

https://www.audiosciencereview.com/...nes-what-is-the-consensus.33589/#post-1228991

I've seen other people quoting a 20Hz-20KHz (+/-3dB) response. This also seemed to me to be more likely, as I have had audio (albeit rolled off) down to 20Hz recorded by the mic. Could this be interference?

+/-3dB is a fairly significant wobble, IMO.

I'm not surprised the mic picked up LF. LF is usually louder, relatively, than higher frequencies and boundary coupling comes into play. Question is: was that 20Hz a meaningful signal or unrelated noise?

Either way, for measurement purposes it's good to have at least one clear, reliable reference point so that you dont have to reason yourself in circles figuring out what's causing a particular reading.

Pro circles are just as likely to say unverified half-supported things as the retail or audiophile crowd.

Edit: Removed autoembed.

 

[JTee]

Again, not sure if this is following forum etiquette or not, but I have some follow up questions that don't relate directly to the topic. Hope this is okay!

I'm not surprised the mic picked up LF. LF is usually louder, relatively, than higher frequencies and boundary coupling comes into play. Question is: was that 20Hz a meaningful signal or unrelated noise?

To answer your question, it was unrelated noise.

But isn't this also a question in general? According to Amroc, and testing with my own ears, there's a room mode at around 40Hz. A question I had was is that really something I need to dedicate time and space to fixing? I'm leaning towards yes, because I feel that's still a fairly audible frequency, and kick + bass are likely to have at least some energy in that area that'll need some attenuation. This is the reason for wanting a measurement mic, though, as all I'm going on is a very basic (not to mention inaccurate from my input) modal calculator and "Does this sound louder?". Not the most scientific of experiments. I'm also nervous of over-thinking this and just never getting it done because I go down yet another rabbit hole of information that makes me question whether my current plans are actually worth it. My question is this: My monitors can produce 40Hz, but I'm aware the vast majority of speakers can't. Should I push for serious treatment down to this frequency, or focus on issues closer to 50-70Hz range? My guess was the usage of NS-10's was based on this philosophy: If they can't hear it, does it matter if you can't?

If you're trying to equalise a loudspeaker to be flat pseudo anechoically, then you'll need a calibrated mic

Sorry for my ignorance, but in this situation is there a difference? I'd understand that a company wanting to produce a monitor would want a flat response, and that a professional designing a studio space for somebody would want a room with minimal comb-filtering or modal issues. However, in this case, isn't it a combination of a known set of speakers within a room? Fundamentally I agree with your point of view, might as well buy the mic now and have the best setup I can in the moment than wait for the "future, perfect studio", it was just something that I was constantly tripping over. When you have three variables (mic accuracy, monitor accuracy and the room itself) how can you possibly tell which part is creating the problems that are displayed? Finally came to the conclusion that, for the room and monitors, as they both form part of the same system the cause is irrelevant, I just need to know how to fix it.

 

[sergeauckland]

Sorry for my ignorance, but in this situation is there a difference? I'd understand that a company wanting to produce a monitor would want a flat response, and that a professional designing a studio space for somebody would want a room with minimal comb-filtering or modal issues. However, in this case, isn't it a combination of a known set of speakers within a room? Fundamentally I agree with your point of view, might as well buy the mic now and have the best setup I can in the moment than wait for the "future, perfect studio", it was just something that I was constantly tripping over. When you have three variables (mic accuracy, monitor accuracy and the room itself) how can you possibly tell which part is creating the problems that are displayed? Finally came to the conclusion that, for the room and monitors, as they both form part of the same system the cause is irrelevant, I just need to know how to fix it.

The difference I was trying to get at is that if one is measuring a loudspeaker, pseudo-anechoically, i.e. excluding the room effects, and therefore excluding the bass below around 200Hz, the errors in an uncalibrated microphone could well exceed those of the loudspeaker. It should be possible to measure and therefore EQ a loudspeaker to better than 1dB provided of course the microphone is calibrated better than that.

When measuring a loudspeaker in a room, the errors from flat are many times greater, (especially in the bass) so an uncalibrated microphone with a few dBs of its own errors will still produce a useful result as it's difficult to achieve an in-room result better than that, i.e. a few dBs, especially if measured at a few listening positions and some sort of 'average' taken. Clearly, using a calibrated microphone for in-room measurements is better than an uncalibrated mic, but even with an uncalibrated mic, the improvements in-room can be significant.

S.

 

[JTee]

The difference I was trying to get at is that if one is measuring a loudspeaker, pseudo-anechoically, i.e. excluding the room effects, and therefore excluding the bass below around 200Hz, the errors in an uncalibrated microphone could well exceed those of the loudspeaker. It should be possible to measure and therefore EQ a loudspeaker to better than 1dB provided of course the microphone is calibrated better than that.

When measuring a loudspeaker in a room, the errors from flat are many times greater, (especially in the bass) so an uncalibrated microphone with a few dBs of its own errors will still produce a useful result as it's difficult to achieve an in-room result better than that, i.e. a few dBs, especially if measured at a few listening positions and some sort of 'average' taken. Clearly, using a calibrated microphone for in-room measurements is better than an uncalibrated mic, but even with an uncalibrated mic, the improvements in-room can be significant.

S.

Thanks! You've cleared that up for me.

 

[Curvature]

But isn't this also a question in general? According to Amroc, and testing with my own ears, there's a room mode at around 40Hz. A question I had was is that really something I need to dedicate time and space to fixing? I'm leaning towards yes, because I feel that's still a fairly audible frequency, and kick + bass are likely to have at least some energy in that area that'll need some attenuation. This is the reason for wanting a measurement mic, though, as all I'm going on is a very basic (not to mention inaccurate from my input) modal calculator and "Does this sound louder?". Not the most scientific of experiments. I'm also nervous of over-thinking this and just never getting it done because I go down yet another rabbit hole of information that makes me question whether my current plans are actually worth it. My question is this: My monitors can produce 40Hz, but I'm aware the vast majority of speakers can't. Should I push for serious treatment down to this frequency, or focus on issues closer to 50-70Hz range? My guess was the usage of NS-10's was based on this philosophy: If they can't hear it, does it matter if you can't?

Leave in the name of the poster you are quoting, and they will get an alert and link to the exact post

Bass modes are really hard to fix. A microphone may help diagnose issues, and EQ may help address them. For the most part, the goal is simply to reduce issues, not perfection. You can make a tremendous amount of difference just by pulling down bass peaks until they sound more or less even in level. It's usually not worth messing around with simulations unless you know the underlying science decently well. There are a lot of assumptions and variables that may lead you astray.

Say you start (1) with a known (calibrated) microphone response and (2) you have a speaker with highly accurate anechoic measurements like the kind published here. Once you (3) set the speaker in your room and measure, you will have a good idea of how the room has influenced the outcome. The next step (4) is to make the speakers actually usable for studio work. So (5) you take the measured response and determine the relevant perceptual problems and if your tools can to address them. (5) takes a lot of learning and is the most complex. The best guide there is a good book, not random uncontrolled experimentation or advice. In that sense, 5 is also the cheapest, costing way less than a pair of speakers or a computer.

This is the book that is the most helpful: https://www.routledge.com/Sound-Rep...phones/Toole-Olive-Welti/p/book/9781032761930 It combines knowledge of acoustics, psychoacoustics and electrical engineering along with general industry knowledge. Most other books, including well-known books by mixing engineers or acousticians tend to go very light on psychoacoustics, which is by far the biggest determining factor, and readers, even professionals, come away with poor advice by reading only, for example, Everest's and Pohlmann's Master Handbook of Acoustics. The link I posted is to the yet-unreleased 4th edition, but the current 3rd edition will contain almost everything. You can bet the differences and details will be discussed here.

I know this advice is annoying and the opposite of what will help you get started right now. Well, the more you learn, the less money you'll waste, the better time spent. Ask as many questions as you like along the way.

 

[Curvature]

I think the most complicated and maybe even most valuable aspect of starting on this line of learning is avoiding being swayed by persuasive-sounding BS on forums and Youtube and elsewhere, a lot of which is commercially motivated. Knowing enough to ask the right question counts for a lot.

 

[Curvature]

but even with an uncalibrated mic, the improvements in-room can be significant.

Microphones in the LF have broad, predictable slopes (low Q). Acoustical errors will have narrow peaks or dips (high Q). The uncalibrated result can be used as a guide, even if you can't be certain about absolute level.

 

[JTee]

Leave in the name of the poster you are quoting, and they will get an alert and link to the exact post

Duly noted. I was using the "quote" button rather than "reply", and it didn't come up with the name.

This is the book that is the most helpful: https://www.routledge.com/Sound-Rep...phones/Toole-Olive-Welti/p/book/9781032761930...

I know this advice is annoying and the opposite of what will help you get started right now. Well, the more you learn, the less money you'll waste, the better time spent. Ask as many questions as you like along the way.

Honestly, from everything I've read, trawling through decades of discussions on the topic, I think just having a point of reference will be extremely helpful. There's a part of me that just wanted a book, or a web article, or something that I thought "If I follow this, I won't go too far wrong" especially as I've read some books, like Mike Senior's "Mixing Secrets for the Small Studio" which was... Brief, on the subject. The BBC reports were an interesting read, although designed above what I think would be reasonable for what I can actually do, and as far as books go I found a lot of mixed information. For instance, I've seen some people talk about Everest's as if it were the bible, and others say it's a load of crap. I assumed it was somewhere in the middle, probably full of useful information although maybe from a more biased point of view towards certain theories and practises.

Speaking of questions... I'm in a position where I can either have my desk faced in a way that I have a window as an early reflection location, or have mismatched monitor positions for SBIR (One will be a close wall, the other the window about half a foot farther away). Is there a lesser of two evils? Would it be worth experimenting to see if I can get some constructive interference with SBIR from the window / wall or will the mismatch be really destructive?

Thanks for all your help so far.

 

[Curvature]

For instance, I've seen some people talk about Everest's as if it were the bible, and others say it's a load of crap. I assumed it was somewhere in the middle, probably full of useful information although maybe from a more biased point of view towards certain theories and practises.

I don't know why anyone would call it a bad book. That doesn't make any sense. It's an excellent book focused on acoustics. The only missing side to it is something that's outside the discipline, at least right now: psychoacoustics. That goes for all published academic work--you need the right context to understand the perspective embedded in the writing.

Speaking of questions... I'm in a position where I can either have my desk faced in a way that I have a window as an early reflection location, or have mismatched monitor positions for SBIR (One will be a close wall, the other the window about half a foot farther away). Is there a lesser of two evils? Would it be worth experimenting to see if I can get some constructive interference with SBIR from the window / wall or will the mismatch be really destructive?

I would prefer a symmetrical speaker position. SBIR can be dealt with using a combination of panel-based (not porous) absorbers and EQ. You can also try facing the corner, which is often an effective option, if it's comfortable.

 

[JTee]

I don't know why anyone would call it a bad book. That doesn't make any sense. It's an excellent book focused on acoustics. The only missing side to it is something that's outside the discipline, at least right now: psychoacoustics. That goes for all published academic work--you need the right context to understand the perspective embedded in the writing.

I think it may have been a case of "Everyone else says it's great, so I'm going to say otherwise to be cool and edgy!"

I would prefer a symmetrical speaker position. SBIR can be dealt with using a combination of panel-based (not porous) absorbers and EQ. You can also try facing the corner, which is often an effective option, if it's comfortable.

I'll have a think about possible options for placing it towards a corner. Not sure if it would work, but it's an interesting concept. You say non-porous based absorption, so would a diaphragmatic panel be better suited to that job? I posted in another thread by The_Ogre about a Limp Mass Membrane absorber made with MLV. Seemed a fairly cost effective solution, especially with a mass of loft insulation inside to lower the Q value.

 

[Curvature]

I think it may have been a case of "Everyone else says it's great, so I'm going to say otherwise to be cool and edgy!"

I'll have a think about possible options for placing it towards a corner. Not sure if it would work, but it's an interesting concept. You say non-porous based absorption, so would a diaphragmatic panel be better suited to that job? I posted in another thread by The_Ogre about a Limp Mass Membrane absorber made with MLV. Seemed a fairly cost effective solution, especially with a mass of loft insulation inside to lower the Q value.

Yes, a diaphragm/panel absorber. An air gap plus insulation inside the panel is essential not only for tuning but also for absorption to be effective.

 

[kemmler3D]

My monitors can produce 40Hz, but I'm aware the vast majority of speakers can't. Should I push for serious treatment down to this frequency, or focus on issues closer to 50-70Hz range? My guess was the usage of NS-10's was based on this philosophy: If they can't hear it, does it matter if you can't?

Treatment at 40hz in the sense of absorbing materials is probably beyond your level of commitment (think covering half of the entire room in traps) but correcting with EQ is no more difficult at 40hz than 200...

Lots of people have headphones and they almost all go below 40hz...

 

[JTee]

Lots of people have headphones and they almost all go below 40hz...

...Good point...