Speakers distortion

[sergeauckland]

You should always have the mic steady, ideally on a stand with a long boom to minimise reflections. If you're doing a listening window measurement,then the microphone is moved between measurements, but kept stationary for each individual sweep.

You should also always use a calibrated microphone as the errors can be several dBs at any frequency. If you're only setting up subwoofers, an uncalibrated mic is OK, but for setting up equalisers or measuring loudspeakers, calibration is necessary.


For distortion, one on-axis sweep is fine.


S

 

[RayDunzl]

My microphone is on a small stand on the top of the couch (at ear position), and placed at the acoustic center of the stereo speakers.

Taking two sweeps, using acoustic reference option, left and right measures, the impulse response can indicate how far off-center the mic is, down to one sample time (or maybe less in newer REW versions).

Sometimes the middle is not really where you think it is.

With both speakers active, if the mic is not well centered, the high frequeny measurement will reveal comb filtering (wavy response line) as some frequencies will cancel out (timing difference).

 

[Blumlein 88]

If you have an uncalibrated mic, you can grab a screenshot of the response from the manufacturer of the mic. Use Web Plot Digitizer to create a csv file which REW can use for a cal file and improve your results. Even cheap recording condenser mics are very consistently made. Works best with omnis but even cardioids can be usefully improved for REW use.

 

[Blumlein 88]

https://www.audiosciencereview.com/...if-you-dont-have-calibrated-microphones.1459/

Here's a thread on the mic calibration.

 

[andreasmaaan]

Should I hold the mic steady at one position while doing RTA measurement or should I move it like when I'm doing MMM RTA frequency response measurement?

Hold it in one position, preferably on a stand I’d say.

I'm using microphone with calibration file.
It would probably be the best to calibrate SPL in the SPL Meter window but I don't know how to measure reference SPL.

Which mic is it? I believe most USB measurement mics have SPL calibration.

If that’s not the case, you could take a shortcut and use a smartphone SPL meter, which won’t be very accurate but will get you in the ballpark (+/-5dB) for the better ones IIRC. Or you could get an SPL meter.

When doing sweep distortion measurement is it necessary to make several measurements from different positions and average it like when doing frequency response measurement or it is enough do do it from single position?

It depends what you’re trying to achieve. Really, the best way to do it (other than in an anechoic chamber) is nearfield. I’m not sure I see the value in in-room distortion measurements at the listening position personally, but that’s just my 2c...

 

[Blumlein 88]

snip........


It depends what you’re trying to achieve. Really, the best way to do it (other than in an anechoic chamber) is nearfield. I’m not sure I see the value in in-room distortion measurements at the listening position personally, but that’s just my 2c...

The value is to see if you can beat Ray's results at the LP.

 

[Krunok]

You should always have the mic steady, ideally on a stand with a long boom to minimise reflections. If you're doing a listening window measurement,then the microphone is moved between measurements, but kept stationary for each individual sweep.

That is exectly how I'm doing it whem making frequency response measruement.
When I do 7-9 sweeps and make average of them it correlates really well with the results I'm getting from RTA moving microphone method.

You should also always use a calibrated microphone as the errors can be several dBs at any frequency. If you're only setting up subwoofers, an uncalibrated mic is OK, but for setting up equalisers or measuring loudspeakers, calibration is necessary.

My microphone is calibrated and cal file is loaded into REW, but I don't know how to calibrate SPL level in REW.
When I'm making log sine sweeps and RTA MMM there's a difference of app 9dB between their graphs in REW although I'm doing them at the same volume level.

For distortion, one on-axis sweep is fine.

That was the answer I was looking for, thank you!

 

[Krunok]

Even cheap recording condenser mics are very consistently made.

Mine is of that type, it's calibrated and I'm satisfied with the consistency of the measurements.

 

[Krunok]

Hold it in one position, preferably on a stand I’d say.

Which mic is it? I believe most USB measurement mics have SPL calibration.

If that’s not the case, you could take a shortcut and use a smartphone SPL meter, which won’t be very accurate but will get you in the ballpark (+/-5dB) for the better ones IIRC. Or you could get an SPL meter.

It depends what you’re trying to achieve. Really, the best way to do it (other than in an anechoic chamber) is nearfield. I’m not sure I see the value in in-room distortion measurements at the listening position personally, but that’s just my 2c...

This is my mic. I believe the problem is i don't know how to do calibrate SPL in REW. I set recording slider to the max in Windows sound properties but I don't know how to set SPL level in REW.

It depends what you’re trying to achieve. Really, the best way to do it (other than in an anechoic chamber) is nearfield. I’m not sure I see the value in in-room distortion measurements at the listening position personally, but that’s just my 2c...

How far from speakers should the mic be in nearfield distortion measurement? And at what height?

 

[mitchco]

I have found only one measurement mic that publishes it's distortion spec: http://www.isemcon.com/datasheets/EMX7150-US-r04.pdf

 

[Krunok]

The value is to see if you can beat Ray's results at the LP.

Oh, I wasn't even hoping to do that, my Castles and his MLs are not even in the same league!

Is there some other distortion measurements on this forum so I can check how my Castles compare to the other non-esoteric speakers?
@RayDunzl, do you maybe have measurements of some other speakers?

 

[Krunok]

I have found only one measurement mic that publishes it's distortion spec: http://www.isemcon.com/datasheets/EMX7150-US-r04.pdf

Pity they didn't specify distortion at lower SPL. Can in it be assumed that THD will lower for the same ammount of dB when SPL is lowered?
3% is app -30dB. If SPL drops for 60 dB to 85dB will aTHD also drop to -90dB (0,003%)?

 

[SIY]

Pity they didn't specify distortion at lower SPL. Can in it be assumed that THD will lower for the same ammount of dB when SPL is lowered?
3% is app -30dB. If SPL drops for 60 dB to 85dB will aTHD also drop to -90dB (0,003%)?

The limitation in that measurement may be the self-noise. It's been a while since I measured the EMX7150's noise, but if memory serves, the one I have was something like 27-28dBSPL over a 20kHz bandwidth, A weighted.

 

[sergeauckland]

Pity they didn't specify distortion at lower SPL. Can in it be assumed that THD will lower for the same ammount of dB when SPL is lowered?
3% is app -30dB. If SPL drops for 60 dB to 85dB will aTHD also drop to -90dB (0,003%)?

The graph on page 2 shows that distortion stays low, around 0.1%, until it rises at around 125dB SPL, so I would expect for that mic that distortion doesn't get much lower. Measurement mics all have small capsules, so their self-noise is higher than large capsule mics, so ultimately, the THD+N figure will be N limited.

If your loudspeakers are below 0.1% distortion, you have nothing to worry about.


S

 

[SIY]

If your loudspeakers are below 0.1% distortion, you have nothing to worry about.

GedLee metric might be more useful than THD in this particular instance. I've also been experimenting with intelligibility indexes, specifically ABC-MRT, and so far, I seem to get a nice correlation between that and my subjective impressions of clarity (limited sample size so far, so I don't want to draw any sweeping conclusions!).

 

[Krunok]

The graph on page 2 shows that distortion stays low, around 0.1%, until it rises at around 125dB SPL, so I would expect for that mic that distortion doesn't get much lower. Measurement mics all have small capsules, so their self-noise is higher than large capsule mics, so ultimately, the THD+N figure will be N limited.

If your loudspeakers are below 0.1% distortion, you have nothing to worry about.
S

Which graph on page 2 are you referring to?

Even ML and JBL M2 don't go below 0.1% on the graph @RayDunzl posted.

Here is how noise vs THD looks for my left speaker:

 

[andreasmaaan]

How far from speakers should the mic be in nearfield distortion measurement? And at what height?

20-25cm is about right, with the speakers as far away from reflective surfaces as possible. If the nearest reflective surface is say 2m away and the mic is 25cm away, your direct wave will be about 20-25dB louder than the first reflection. But of course the near(er) field measurement will skew the amplitude response. So it’s all about balancing those two factors. Which is why I think 20-25cm is probably about right in most cases.

 

[Krunok]

20-25cm is about right, with the speakers as far away from reflective surfaces as possible. If the nearest reflective surface is say 2m away and the mic is 25cm away, your direct wave will be about 20-25dB louder than the first reflection. But of course the near(er) field measurement will skew the amplitude response. So it’s all about balancing those two factors. Which is why I think 20-25cm is probably about right in most cases.

So mic should be at the tweeter height aimed directly at it, 20-25cm from the tweeter?

 

[sergeauckland]

My understanding is that for a near-field measurement, the microphone has to be within 1-2mm of the middle of the driver, as they're mostly used to make measurements on the bass driver below the critical frequency in-room, which for most UK rooms will be around 200Hz.

Far field measurements are done pseudo anechoically at 1m, perhaps 2m if the room is large (or outside) and the 'speaker is raised off the floor by several metres.

S

 

[sergeauckland]

Which graph on page 2 are you referring to?

Even ML and JBL M2 don't go below 0.1% on the graph @RayDunzl posted.

Here is how noise vs THD looks for my left speaker:

I was referring to Page 2 of the spec for the EMX7150 measuring mic.

S