Help interpreting some measurements with speakers and a microphone of unknown quality

[SSSheridan]

Hey all, I'd appreciate some pointers in interpreting these early results and where I should go from here.

I have some secondhand Bose computer speakers (active) and a Yamaha subwoofer (active). The one piece that I trust is my Focusrite Scarlett Solo. And for now, I've picked up a couple of cheap XLR mics on steep discounts from Amazon Warehouse: a Behringer XM8500 dynamic mic for 15€ and a G-Mark G-77 (seems to be a new/unknown Chinese brand) condenser mic for 9€.

I'm looking to characterize my system so I can optimize what I have and prioritize my upgrades on a limited budget. It's okay that the accuracy won't be great; I'd just like to get a basic testing setup so I can try things and see what happens. For example, my subwoofer has a really weak response around 60-70Hz. If I try to fix that with EQ, how badly will it distort? That sort of thing.

Here are a few ARTA graphs. These are not careful measurements; the room was noisy. The point is just to learn how to measure and how to interpret what I'm seeing.

Prominent harmonic distortion. I checked the signal being sent to the (active) speakers and it was a clean 1kHz signal, with no harmonics. Can analog (XLR) microphones produce harmonic distortion, or is that definitely the speakers?


Next, comparing the two microphones' readings of white noise:

Most of the peaks and troughs are shared between the two graphs, which makes me think that the peaks and troughs are mostly characteristic of the speakers.

How good/bad are cheap XLR mics plugged into a decent ADC like the Focusrite? Given a limited budget, I'm not sure whether to return both of these fellas and invest in a better mic, so I can have more trustworthy measurements, or whether I should just save the money for a speaker upgrade.

Appreciate any thoughts and pointers!

 

[HarmonicTHD]

Hey all, I'd appreciate some pointers in interpreting these early results and where I should go from here.

I have some secondhand Bose computer speakers (active) and a Yamaha subwoofer (active). The one piece that I trust is my Focusrite Scarlett Solo. And for now, I've picked up a couple of cheap XLR mics on steep discounts from Amazon Warehouse: a Behringer XM8500 dynamic mic for 15€ and a G-Mark G-77 (seems to be a new/unknown Chinese brand) condenser mic for 9€.

I'm looking to characterize my system so I can optimize what I have and prioritize my upgrades on a limited budget. It's okay that the accuracy won't be great; I'd just like to get a basic testing setup so I can try things and see what happens. For example, my subwoofer has a really weak response around 60-70Hz. If I try to fix that with EQ, how badly will it distort? That sort of thing.

Here are a few ARTA graphs. These are not careful measurements; the room was noisy. The point is just to learn how to measure and how to interpret what I'm seeing.

View attachment 228924


Prominent harmonic distortion. I checked the signal being sent to the (active) speakers and it was a clean 1kHz signal, with no harmonics. Can analog (XLR) microphones produce harmonic distortion, or is that definitely the speakers?


Next, comparing the two microphones' readings of white noise:
View attachment 228925
View attachment 228926

Most of the peaks and troughs are shared between the two graphs, which makes me think that the peaks and troughs are mostly characteristic of the speakers.

How good/bad are cheap XLR mics plugged into a decent ADC like the Focusrite? Given a limited budget, I'm not sure whether to return both of these fellas and invest in a better mic, so I can have more trustworthy measurements, or whether I should just save the money for a speaker upgrade.

Appreciate any thoughts and pointers!

The problem is that there is no easy way to calibrate your microphones neither in their frequency response nor in their SPL level.

Currently you are measuring the microphones characteristics overlayed with your speaker and your room characteristics. Hopeless to know what effect you observe is caused by what. Therefore people use calibrated measuring mics of which the cheapest option is the UMIK1 for ca 100 bucks.

As for SPL you might get away with by using a handheld SPL meter to calibrate against, but not very accurately. If you don’t have one already save the money and get the UMIK instead. Or just ignore it and only be content with relative SPL values.

 

[Blumlein 88]

Both of your mics appear to be dynamic cardioid microphones. Cardioids are directional and generally have dropping response below 200 hz. Except if you are about 1 meter or closer where proximity effect raise the low end. They may be fine for voice, but not a good choice for measuring speakers. Here is the response graph of the 8500. It is described as remarkably flat from 50hz-15khz. Hmmmmm, this is what could be called marketing or lying.

Omni's are typically used for measurements. Like these:

Home

If you had an omni recording microphone, you could probably make good use of it. There are some ways to get somewhat close to calibrate the SPl. These dynamics however have poor response, plus the proximity effect interferes below 200 hz. It is like using a crooked ruler. You listed one measurement at 10 cm from the speaker. The low end will be heavily boosted with a cardioid at that distance, but we have no way to know how much.

A Umik 1 solves a number of issues. Software will pick it up and your SPL readouts will be correct. You could roughly calibrate non_USB microphones, but you'll need another SPL meter or some good specs to base it upon from the maker.

 

[AnalogSteph]

The XM8500 is a remarkably good-sounding cheapie dynamic mic, but obviously the low end needs plenty of help from proximity effect. At one or two inches off, the response should be quite a different story.

Prominent harmonic distortion. I checked the signal being sent to the (active) speakers and it was a clean 1kHz signal, with no harmonics. Can analog (XLR) microphones produce harmonic distortion, or is that definitely the speakers?

Everything produces harmonic distortion to some degree, from speaker over microphone to analog input stage and ADC. The magnitude, order and modest analog input level would seem to suggest hysteresis-related distortion in the speaker driver though - I wouldn't think the Scarlett input stage would be that bad, and dynamic mics can generally take insane levels (as much as >150 dB SPL). You can easily determine who the culprit is by varying recording distance or input gain and observing the effect on the ratio of fundamental and 3rd harmonic. If there is little to no change, it has to be the speaker. If dialing back the gain has a positive effect, it's the input stage. Same for distance could be either input stage or microphone.

If you need a measurement mic on the very cheap, your best option may be DIY. Even a Behringer ECM8000 or Superlux ECM999 (much the same animal, apparently) will set you back 34-39€, and then you still don't have an individual calibration curve.

 

[SSSheridan]

Thanks for the replies! I've got a better idea what's what now.

So a UMIK-1 is on my wishlist now. But, I'm also looking at the Behringer ECM8000 (+ my Focusrite). I understand that the UMIK-1 would save me a bit of effort and uncertainty, but it would be a bit painful for this student's budget.

From what I'm seeing, the Behringer ECM8000 is the standard for ultra-budget measurement mics. Its disadvantages compared to the UMIK-1 are:
1) Being an XLR mic, it can't give absolute SPL values.
2) It doesn't come with an individual calibration profile.
3) Anything else?

For my use case, (1) relative SPL values are fine, I think. (2) frequency response does matter; however, looking at this sample of 130 Behringer ECM8000s:

...it looks like for basic system tuning, they should be trustworthy enough without calibration from perhaps 40Hz-3kHz, with tuning-by-ear above and below that range.

Is there anything that I'm not thinking of?

 

[Blumlein 88]

Home

Dayton Audio EMM-6 Electret Measurement MicrophoneThe Dayton Audio EMM-6 is a precision electret condenser microphone designed for measurement and critical recording applications. The extremely flat frequency response and true omni-directional polar pattern make it perfect for use with room...

www.parts-express.com

I think the Dayton mic is a better deal than the ECM8000 because it comes with individualized calibration curves for the mic you get. With the ECM8000 you are just depending upon the inherent response of the microphone. Also Cross Spectrum which sales even better calibrated versions of measurement microphones stopped selling the ECM8000 saying the quality had deteriorated too far. The EMM 6 from Dayton Audio would work with the Scarlett you already have.

 

[SSSheridan]

If you need a measurement mic on the very cheap, your best option may be DIY.

Do you mean DIY as in breadboarding/constructing from scratch? I've dabbled a bit, but boy oh boy is it a rabbit hole.

Home

Dayton Audio EMM-6 Electret Measurement MicrophoneThe Dayton Audio EMM-6 is a precision electret condenser microphone designed for measurement and critical recording applications. The extremely flat frequency response and true omni-directional polar pattern make it perfect for use with room...

www.parts-express.com

I think the Dayton mic is a better deal than the ECM8000 because it comes with individualized calibration curves for the mic you get. With the ECM8000 you are just depending upon the inherent response of the microphone. Also Cross Spectrum which sales even better calibrated versions of measurement microphones stopped selling the ECM8000 saying the quality had deteriorated too far. The EMM 6 from Dayton Audio would work with the Scarlett you already have.

Okay, if the quality has deteriorated, that is a concern. With the Dayton being half the price of the UMIK already, I understand why y'all tend to just recommend going for the real deal. And I respect the conventional wisdom 'round these parts.

So, Dayton EMM-6 sounds like a good idea for my use case. (And I like the feeling of using the perfectly good ADC I've already got.) As I understand it, the UMIK-1 holds the following advantages over the Dayton:
- absolute SPL levels
- plug-n-play (which does sound dang handy if it works on smartphones as well)

Are there other advantages to consider?

Edit: Sonarworks is good, right? I'm in Germany, and the Sonarworks XREF20 seems slightly easier to get my hands on. However, given that the price is 60-70% of the UMIK-1, I might end up just biting the bullet.

 

[Roland68]

Do you mean DIY as in breadboarding/constructing from scratch? I've dabbled a bit, but boy oh boy is it a rabbit hole.


Okay, if the quality has deteriorated, that is a concern. With the Dayton being half the price of the UMIK already, I understand why y'all tend to just recommend going for the real deal. And I respect the conventional wisdom 'round these parts.

So, Dayton EMM-6 sounds like a good deal for my use case. (And I like the feeling of using the perfectly good ADC I've already got.) As I understand it, the UMIK-1 holds the following advantages over the Dayton:
- absolute SPL levels
- plug-n-play (which does sound dang handy if it works on smartphones as well)

Are there other advantages to consider?

Edit: Sonarworks is good, right? I'm in Germany, and the Sonarworks XREF20R5 seems slightly easier to get my hands on.

This is one of the best ways to get an affordable calibrated measurement microphone.
https://shop.hifi-selbstbau.de/produkt/omnitronic-mm-2usb/
Alternatively, you can also send in your measurement microphone for calibration.
https://shop.hifi-selbstbau.de/produkt/mikrofonkalibrierung/