Speaker/Room Measurement Witchcraftery

[camverton]

Good points above re stereo or single speaker measuring. In practice, once I’ve setup the sub taking individual readings, I take three measurements, the first with left, right and sub, the second left and sub together, and the third right and sub together. For checking mid to high frequencies I use the left and right channels alone, but for bass frequencies and up to about 250 Hz I use the first measurement with both channels measured together. IIRC this was suggested by the designer of the software particularly when using the spectrogram and waterfall plots. Needless to say it is also instructive to look at each channel individually when considering lower frequencies.

 

[tuga]

I don’t really understand this. Surely most of us are using measurements in order to double-check what we are hearing from the listening seat in our rooms? As such that will always be the sound of two speakers playing in that space. To my ears the sound of one speaker alone is way less than half the sound of a well setup stereo system, it just doesn’t drive the room in anything like the same way. Surely we want to be measuring what we are actually hearing, not some theoretical abstraction. FWIW I’ve similarly never understood the point of measuring a speaker at 1m unless you are trying to track a very specific fault in pair-matching or whatever. Without an anechoic chamber any such measurement won’t tell you much about the speaker design, so really all that interests me is the sound of my full system measured at the listening seat in the way I listen to it. I’d never cite any of my measurements as being representative of a Tannoy Monitor Gold, a JR149 or whatever as they are just an REW measurement from the listening seat in my room and as such valueless to anyone other than me (assuming they actually have any value to me!).

I guess that there could well be phase/time errors that a single mic placement would suffer from that a pair of ears wouldn’t that could give spurious results, but in that case surely a crossed pair and simultaneous stereo measurements should be done. The whole point is to measure what we are experiencing, and we are not hearing two entirely isolated mono signals. That is not happening as both speakers are driving the room and any reflections, time and phase delays etc are happening in real time between both speakers and that listening space. That is what needs to be measured.

You are not measuring what you are listening to but the response that each speaker produces in the room in relation to the listening position.

You don't want to be measuring both speakers simultaneously.

 

[ToTo Man]

I believe the issue here is the single mic, and I don't think a simple pair of crossed mics would resolve this. I suspect the most accurate way to capture sound from a stereo pair of speakers as we hear it would be to use a pair of mics flush-mounted on a dummy head and take a stereo measurement at each mic, and then average the two.

I would flush-mount the mics at the opening of the ear canals instead of recessing them into them, in order to avoid the canal resonances that can be seen on many uncorrected headphone measurements*. I might even do away with the pinnae too.

Like a human head, a dummy head provides a physical barrier that partially shields/deflects the direct output of one speaker from reaching the opposite ear, and thus may help to mitigate the interference issues described in my last post.

* An increasing number of headphone enthusiasts are now using DIY "flat-plate couplers" instead of dummy heads, as they are much cheaper and produce results more comparable to anechoic speaker measurements with less correction compensation required.

 

[tuga]

I believe the issue here is the single mic issue, and I don't think a simple pair of crossed mics would resolve this. I suspect the most accurate way to capture sound from a stereo pair of speakers as we hear it would be to use a pair of mics flush-mounted on a dummy head and take a stereo measurement at each mic, and then average the two.

There is no issue with the mics.

If one wishes to record the sound of their speakers then one can use a stereo mic or a mic pair and should obviously play both speakers simultaneously.

For technical measurements (room response, speaker and listener positioning) use one mic and measure one speaker at a time.

 

[Tony L]

For technical measurements (room response, speaker and listener positioning) use one mic and measure one speaker at a time.

I think you mean ‘for convention’s sake’ rather than ‘for technical measurements’ as to be blunt you stand no chance of measuring the effect of two speakers interacting with an acoustic space by measuring one of them at a time! Basically you are measuring and forming opinions on something that never exists in the real world as no one ever listens to their stereo one channel at a time! Averaging two independent mono measurements is not the same as measuring a room being driven by two speakers. It is simply not the same event. Beyond pair matching, setting crossover level controls etc I can’t see much use at all for mono measurements, I want to know what is happening when the room is loaded in the normal way with two speakers.

 

[ToTo Man]

I suspect the most accurate way to capture sound from a stereo pair of speakers as we hear it would be to use a pair of mics flush-mounted on a dummy head and take a stereo measurement at each mic, and then average the two.

I would flush-mount the mics at the opening of the ear canals instead of recessing them into them, in order to avoid the canal resonances that can be seen on many uncorrected headphone measurements*. I might even do away with the pinnae too.

Like a human head, a dummy head provides a physical barrier that partially shields/deflects the direct output of one speaker from reaching the opposite ear, and thus may help to mitigate the interference issues described in my last post.

Another potential solution would to swap the dummy head for the real thing by using a pair of in-ear mics like "Oluv's Gadgets" uses in his YouTube videos to audition headphones. Need to make sure you keep your head nice and still during the measurements though!

 

[tuga]

I think you mean ‘for convention’s sake’ rather than ‘for technical measurements’ as to be blunt you stand no chance of measuring the effect of two speakers interacting with an acoustic space by measuring one of them at a time! Basically you are measuring and forming opinions on something that never exists in the real world as no one ever listens to their stereo one channel at a time! Averaging two independent mono measurements is not the same as measuring a room being driven by two speakers. It is simply not the same event. Beyond pair matching, setting crossover level controls etc I can’t see much use at all for mono measurements, I want to know what is happening when the room is loaded in the normal way with two speakers.

Yes, by convention. I've never seen anyone measuring two speakers simultaneously except newbies, by mistake.

What do you expect to quantify/illustrate with such measurement?
If you listened exclusively to mono then perhaps it might make sense to measure both speakers.
And let's not forget that no two speakers produce the exact same frequency response.

Yes there is some interference but try measuring using the moving mic method and the troughs will even out nicely, which is what happens when you are listening because your head is not fixed onto a tripod and the difference in response between speakers.

An in-room response of a single speaker will give you an idea of how that speaker's response below 300Hz in a particular spot is affected by the room boundaries when measured/listened from a particular seating position. Move the speaker and the response in that range will change, move the seat in any horizontal or vertical direction and the response will also change.
Above 300Hz you can get an idea of the tonal balance reaching the listener's ears. Here toe-in and listener to axis height will make the most difference, followed by distance to side wall.

An example:

 

[tuga]

For those who can't cope with the interference there's always the possibility of cancelling crosstalk using DSP or a screen:

Alternatively, listen with headphones.

 

[JemHayward]

I've spent a couple of days measuring my speakers using REW and a Umik USB calibrated microphone. My method was to measure each speaker from 20Hz to 15kHz with a microphone pointing upwards at the listening position (well, between the two of us on the sofa). I was only applying corrections below 300Hz as the DSP unit is only active in that range. Measure, then apply corrections, the measure again. I applied the calculated biquad provided by REW, and then manually added a room curve to boost the bass (as we like a lot of bass...) and finally added a notch at 60Hz to stop the buzz in the door to the cellar. It sounds superb. We then did a sweep with both speakers to make sure no anomalies cropped up - thankfully none, and finally did it again with the two humans in place to see if we changed things - not much, thankfully as I don't think my wife would have tolerated any more frequency sweeps (though I have to listen to her Motown which is much worse).

 

[igor_xxxx]

This individual or stereo measurements was something that I was also wondering in the last days.Due to some isolation/positioning changes in my system, I was redoing my measurements and I used mostly the individual measurements to adjust the subwoofers, but there is a difference between the individual measurements and the stereo measurement.... just I was not sure if the differences are anomalies of the measuring procedure and to which extent is their influence an the actual sound that I am hearing.

According to my stereo measurement, I have some peaks and dips which I am not able to notice to that extent when listening (per the graph), which are reduced on the individual measurements.

In the end I setup everything according to the individual measurements, listened, everything sounded great and I packed the measuring equipment, but I am bit wondering what is the "correct" way.

 

[JohnPM]

Best is to measure each speaker separately with a timing reference. To see the behaviour with both speakers active the measurements can then be summed. The main issue with multiple speakers running is comb filtering at high frequencies due to path length differences to the mic. At low frequencies resonances which are excited by an individual speaker may be suppressed when a pair reproduce content that is essentially mono, if they drive the resonances in opposite phase. That suppression will also be evident in the summation of individual measurements that have a common timing reference.

 

[darrenyeats]

The only mid/high frequency measurement I'm interested in is reverberation time. I'm looking for a fairly even reverb time across frequency. That's it. In room FR per se for mid/high frequencies isn't of interest to me.

I'd never EQ mid or high frequencies - it's important the direct sound is flat. At bass frequencies the rules are a bit different, so I wouldn't have to kill you if you're doing that (grin).

 

[ToTo Man]

The only mid/high frequency measurement I'm interested in is reverberation time. I'm looking for a fairly even reverb time across frequency. That's it. In room FR per se for mid/high frequencies isn't of interest to me.

I'd never EQ mid or high frequencies - it's important the direct sound is flat. At bass frequencies the rules are a bit different, so I wouldn't have to kill you if you're doing that (grin).

You must be lucky to own speakers that produce your exact desired in-room frequency response without the need for any EQ!

On a serious note, I agree that EQing out narrow peaks and dips in mid and high frequencies is misguided and very likely detrimental, especially if you're only using a single mic location as a reference. However, IME applying EQ with wide Q to ameliorate broad elevations or depressions that are consistently present at multiple listening locations has very little, if any, downside that I can hear. YMMV.