Measurement Microphones - what is the consensus?

[Blumlein 88]

Test mikes are the one place where I'll splurge a bit. At a minimum, I'd want something like an Earthworks or iSemcon. Not terribly expensive but not cheap. It's not just the performance/calibration, it's reliability, stability, and if necessary, traceability. You can get good results with the cheaper mikes for sure, but I have more confidence in the day to day and measurement to measurement performance of the phantom powered lab-grade mikes.

Most of my testing for publication uses PCB Piezotronics phantom power condensers, but they are a significant step up in cost. Superb mikes, though, the peer of things like DPA or GRAS.

I saw your article on those, but it didn't mention price. What is the price? They look like fantastic microphones.

 

[SIY]

I saw your article on those, but it didn't mention price. What is the price? They look like fantastic microphones.

Don't know current, but I'd guess $1500-ish?

 

[Curvature]

A DIY channel I trust recently claimed that all measurement mics need recalibration every 5 years. What do you guys think?

I wouldn't say that. It depends on the diaphragm and construction. As long as they are kept in decent shape, not abused, and in reasonable temperature the drift in response should be low. Certain conditions require studier mics, like for boom operators, roadies or consultants who keep equipment in trunks and have to use them in very hot, cold or humid environments. And then there are precision lab mics which come with guarantees about dB drift for a certain number of years, like those by B&K.

 

[Cbdb2]

Okay you buy the ECm 8000. Now what can you do with it? It outputs an analog signal over XLR. There is no such thing as an XLR to USB adapter. You will have to purchase an audio recording interface which connects to your computer and has input for the ECm 8000 microphone. That will cost at least about $80. There are some I would not recommend for maybe half that. It will also be a little more complex to use versus a Umik or UMM which simply plugs into to the computer ready to use.

Cheaper options are like suggested using a smartphone or ipad. If you have an Iphone that would work though not directly with REW. Android phones usually roll off low end response. There are some USB lavalier microphones and some that plug into the audio in jack on your computer if it has those. They likely would work for your purposes and some are $35 or so. I haven't used one to recommend. It would work probably.

If you have a little measuring microphone for your AVR, it might plug into the audio input of your computer and work. I've not done that either so not sure it would work, but it might. Of course many newer computers no longer have that kind of 3.5 mm audio input jack. Don't know your particular gear.

Dont know about other recievers but my Denon setup mic needs a 5v phantom power.

 

[jhaider]

Test mikes are the one place where I'll splurge a bit. At a minimum, I'd want something like an Earthworks or iSemcon. Not terribly expensive but not cheap. It's not just the performance/calibration, it's reliability, stability, and if necessary, traceability. You can get good results with the cheaper mikes for sure, but I have more confidence in the day to day and measurement to measurement performance of the phantom powered lab-grade mikes.

I don’t disagree with you in word or deed (my current measurement rig for in room measurements* is 4x iSEMcons using REW Pro, though there are CSL calibrated UMIK-1’s and a CSL/Dayton EMM-6 around too, and a Neumann MA-1) but I think there are two issues one should consider:

First, are you measuring loudspeakers or setting up rooms/running room EQ software? IMO one does not benefit much from elite performance for the latter, though for loudspeaker design or characterization it may be more useful. The reason is systemic errors in the microphone don’t affect the biggest things, as they are all relative. Furthermore, the regions where mics are least accurate (ULF, top octave) are really areas that need fine tuning by ear. There’s a danger one regularly sees on the internet of being anal about what in room graphs show without benefit of listening and thinking, and being super picky about room measurement gear plays into reliance on the rig rather than using the measurement rig to inform two ears (+chest/face) and a brain. Also there are mistakes that can lead one astray, such as using an individualized Earthworks calibration for room measurements. (They only offer axial individualized cals.)

Second, are you taking measurements for yourself or for publication? For the latter it’s probably wise to use fancier stuff. Honestly that’s the main reason I personally bought the iSEMcon mics instead of a bunch more Dayton EMM-6’s or similar. In the latter case, something like miniDSP or Dayton makes a lot of sense. If you want an in option in between Dayton and iSEMcon, Neumann’s MA-1 is really nice. It looks fantastic and I trust their calibrations. Here you have the opposite issue as Earthworks; MA-1 is only calibrated for grazing. So not ideal for speaker measurements.

*I’d like to get back into measuring speakers at a high level but have not found suitable space. Our in-town Chicago roof deck is simply too loud/noisy for outdoor measurements (I’ve tried) and I haven’t been able to look seriously for other options.

 

[Bernard23]

No, you are assuming that the error in question will affect my explicitly stated use case.

Exactly, blind work and tweaking will result in very undesirable results which is why a microphone is vital. Improving low frequency performance of a system can be done by reducing low frequency peaks if the center frequency and Q is known, and any microphone will reveal that information. Calibrated or not. Same goes for setting a subwoofer's crossover frequency, you need to know the corner frequency of the filter or the acoustic roll-off, but a microphone doesn't need to be calibrated to show that. And for adding multiple subwoofers too, you can see how they sum together without a calibrated microphone.

I'm talking about looking at a specific attitude that speaker radiation and microphone roll-off have little bearing on. I'm concentrating on the peaks and dips.

I'm pointing out that there's inexpensive measurement tools available. That is all. You are right that person must come to understand what a measurement is telling them for it to be of any use, but you must also know that quality and cost of the microphone has no effect on that understanding. Hands-on experience can, and some might be motivated to get that experience if they know they can use something very affordable or that they already have on hand.

This interests me, because I am not interested in an absolute reference figure, I'm looking for relative changes. My room has some pretty monstrous resonances below 150hz, and a couple of null nodes that are easy enough to hear, so all I need to be able to see is the Fr, relative amplitude and the Q of each peak / trough. I have an android phone, and tried an app called spectroid which of course uses the in-built mic in the phone. I'm wondering if buying a cheap USB mic will help a little, if only showing the response to 20Hz as I'm presuming the background noise suppression is a problem with the built in mic. Are there any USB mics that are suitable, Amazon is full of lapel mics, but they all have DSP built in, which I don't want. I'm not planning to invest in something like an UMIK for the one time I'll use it.

 

[HarmonicTHD]

This interests me, because I am not interested in an absolute reference figure, I'm looking for relative changes. My room has some pretty monstrous resonances below 150hz, and a couple of null nodes that are easy enough to hear, so all I need to be able to see is the Fr, relative amplitude and the Q of each peak / trough. I have an android phone, and tried an app called spectroid which of course uses the in-built mic in the phone. I'm wondering if buying a cheap USB mic will help a little, if only showing the response to 20Hz as I'm presuming the background noise suppression is a problem with the built in mic. Are there any USB mics that are suitable, Amazon is full of lapel mics, but they all have DSP built in, which I don't want. I'm not planning to invest in something like an UMIK for the one time I'll use it.

Yes it will. The UMIK1 for about 100 bucks does all you want. Mainly people use the free REW software to run the FR sweep and even derive the compensation filters.

Plenty of YouTubes out there on how to get started.

(I have one too. Best bang for the Audio buck I ever spent).

 

[Bernard23]

This is the results of a sweep test I did using my Galaxy S23 ultra and spectroid app, after adding eq in the Wiim app. I'm really just experimenting here, so trying to address the null troughs. I added a, slight lift at 54Hz then max 12dB at both 92Hz and 135Hz. Without EQ, the dip at 135Hz is -18db but that's beyond the limits of adjustment in the PEQ.
What I'm not sure if is whether those are errors in the measurement rather than room effects, though they are audible in a sweep test, and they are consistently repeatable. I just bought a iMMC usb mic for £30 so I'll try that out tomorrow and see if I get a similar shaped response.
If I'm honest, this is far more fun than messing about with a turntable as one did back in the days before any form of consumer digital media was around.

 

[No. 5]

I use Spectroid too. On my Pixel 7 Pro using the internal microphone it compares very favorably to my calibrated mics, but I have no idea how a Galaxy S23 Ultra compares. I would, however, change some of the settings for better results as the defaults seem to prioritize time resolution over frequency resolution. FFT size to its max is the big one, "desired transform interval" also to max. The rest of them you can probably leave as is.

This is the results of a sweep test I did using my Galaxy S23 ultra and spectroid app, after adding eq in the Wiim app. I'm really just experimenting here, so trying to address the null troughs. I added a, slight lift at 54Hz then max 12dB at both 92Hz and 135Hz. Without EQ, the dip at 135Hz is -18db but that's beyond the limits of adjustment in the PEQ.

Personally, I'd avoid adding that much boost to a null since it's asking a lot out of your amplifier and speakers. For me, I'd address any audibly objectionable nulls with speaker/listener positioning and/or multiple low frequency sources and signal processing.

What I'm not sure if is whether those are errors in the measurement rather than room effects, though they are audible in a sweep test, and they are consistently repeatable.

If you can hear them and they show up repeatedly in measurements, they are room effects. knowing that, you can experiment with how to deal with them.

 

[Bernard23]

I use Spectroid too. On my Pixel 7 Pro using the internal microphone it compares very favorably to my calibrated mics, but I have no idea how a Galaxy S23 Ultra compares. I would, however, change some of the settings for better results as the defaults seem to prioritize time resolution over frequency resolution. FFT size to its max is the big one, "desired transform interval" also to max. The rest of them you can probably leave as is.

Personally, I'd avoid adding that much boost to a null since it's asking a lot out of your amplifier and speakers. For me, I'd address any audibly objectionable nulls with speaker/listener positioning and/or multiple low frequency sources and signal processing.

If you can hear them and they show up repeatedly in measurements, they are room effects. knowing that, you can experiment with how to deal with them.

That's useful advice thanks. Thought about the impact of adding too much boost, and have set up a 6dB LF shelf from 80Hz for the sub to handle, I'm not too worried about killing it as it wasn't expensive. I can't do much with the room, it's just a useless shape for audio with a bizarre shaped ceiling that was probably a feature back in the 70s when it was built.

 

[No. 5]

happy to help!

Small changes in subwoofer placement, or moving to another acceptable location, or changing how many sound sources are operating over that bandwidth by adjusting the crossovers settings, (if any of those are options) may be things to try too. But yeah, we can just be stuck with what we have sometimes.

 

[Bernard23]

Actually thinking about it, the sub is in the same place it's always been. It's way off to the left of the main speakers so probably phase issues too. This is one reason why I've been thinking about a network receiver with room correction DSP

 

[Jukka]

This is the results of a sweep test I did using my Galaxy S23 ultra and spectroid app, after adding eq in the Wiim app. I'm really just experimenting here, so trying to address the null troughs. I added a, slight lift at 54Hz then max 12dB at both 92Hz and 135Hz. Without EQ, the dip at 135Hz is -18db but that's beyond the limits of adjustment in the PEQ.
What I'm not sure if is whether those are errors in the measurement rather than room effects, though they are audible in a sweep test, and they are consistently repeatable. I just bought a iMMC usb mic for £30 so I'll try that out tomorrow and see if I get a similar shaped response.
If I'm honest, this is far more fun than messing about with a turntable as one did back in the days before any form of consumer digital media was around.


View attachment 345238

You have a thousand dollar phone, but complain about price of UMIK-1????

Some things to consider: the use of a measurement mic is seldom one-time. It will most likely outlive your current phone. The most useful part of a proper mic is that you can use it with proper software on a real computer. The software will do most of the work. If you are reluctant to buy a new mic, but already have another mic, go ahead and try it. Almost any mic, except probably head-set mics, will yield some kind of results. Also if you have friends who own a mic, ask if they could loan it. UMIKs are quite common and not too expensive to be loaned.

PS. Have you tried to connect your phone to a computer and it as a USB mic? I wonder if that's possible...

PPS. Buying second hand is risky for measurement mics, because they can be easily broken if not handled with care.

 

[Bernard23]

Thanks, though comparing a phone which is used for hours a day for multiple tasks to a single use (even if used to run a room check regularly) isn't valid, not in my opinion anyway.

Fully aware that using something like rew and a umik is most reliable and easy to use, but I'm trying to see if it possible using hardware that I already own therefore as little extra outlay as possible.

 

[Jukka]

Yeah, I use a 350 € phone daily and own a UMIK-1.