Measure Stereo Setup simultaneously or Speakers individually

[5omeone]

Hello everyone,

I’d like to measure my stereo system, which consists of two active speakers (B&O Balance). I’ve connected a Wiim Pro Plus with a 10-band parametric EQ beforehand. For the measurement, I’m using the classic combination of REW and UMIK-1. I’ve set up a test configuration, but during my research I’ve come across different approaches:

I’m now wondering whether it’s best to measure the room response of both speakers simultaneously, or to measure each speaker individually. On the EQ side, I can separate left and right; otherwise, that wouldn’t even be an option. One speaker is placed against a straight wall, the other directly in a corner, so I can definitely see how measuring them individually makes sense. BUT: In the end, both speakers always affect the room response...

What would you recommend? Is measuring them individually overkill, or is it worth the effort?

Thanks, everyone, and best regards!

 

[john61ct]

Depends what you are measuring.

Assuming both speakers are identical, "per-speaker as anechoic as possible" is how to EQ their response independent of room resonances / reflections.

If doing crossovers / phase-timing issues then you do separate AND together, follow the methodology for the tool creating the filters.

Overall "room compensation" is separate from the former two goals but some tools combine them.

See the GSonic Reference thread for examples of combining both single speaker and both channels.

The Acourate thread has a lot of good info as well, beyond just about that tool

 

[levimax]

Hello everyone,

I’d like to measure my stereo system, which consists of two active speakers (B&O Balance). I’ve connected a Wiim Pro Plus with a 10-band parametric EQ beforehand. For the measurement, I’m using the classic combination of REW and UMIK-1. I’ve set up a test configuration, but during my research I’ve come across different approaches:

I’m now wondering whether it’s best to measure the room response of both speakers simultaneously, or to measure each speaker individually. On the EQ side, I can separate left and right; otherwise, that wouldn’t even be an option. One speaker is placed against a straight wall, the other directly in a corner, so I can definitely see how measuring them individually makes sense. BUT: In the end, both speakers always affect the room response...

What would you recommend? Is measuring them individually overkill, or is it worth the effort?

Thanks, everyone, and best regards!

You will get a lot of different answers and reasons on this and I have done it both ways. My answer is to do it both ways i.e. take right, left, and combined measurements and look at them to see what is going on. The individual measurements/ corrections are technically more correct but in some cases it sounds better correcting the combined response (there can be many reasons for this). Try it both ways and see which you prefer. You didn't mention it but I would suggest MMM measurements if you aren't doing it that way, much faster and easier and just as accurate if not more accurate than sweeps for EQ.

 

[janbth]

Hello everyone,

I’d like to measure my stereo system, which consists of two active speakers (B&O Balance). I’ve connected a Wiim Pro Plus with a 10-band parametric EQ beforehand. For the measurement, I’m using the classic combination of REW and UMIK-1. I’ve set up a test configuration, but during my research I’ve come across different approaches:

I’m now wondering whether it’s best to measure the room response of both speakers simultaneously, or to measure each speaker individually. On the EQ side, I can separate left and right; otherwise, that wouldn’t even be an option. One speaker is placed against a straight wall, the other directly in a corner, so I can definitely see how measuring them individually makes sense. BUT: In the end, both speakers always affect the room response...

What would you recommend? Is measuring them individually overkill, or is it worth the effort?

Thanks, everyone, and best regards!

Always measure L+R and EQ both speakers together in the bass region. The reason for that is that bass content tend to be mono, and if L and R is EQd separately you risk problems with room modes and possibly destructive interference at the listening position.

EDIT: I added room modes. These are probably much more important than interference in the bass region.

 

[levimax]

Always measure L+R and EQ both speakers together in the bass region. The reason for that is that bass content tend to be mono, and if L and R is EQd separately you risk destructive interference at the listening position.

Lots of people swear by Stereo Bass these day and it has been proven using Deltaware that "most music is summed to mono" is a myth, even for vinyl. One thing for sure if you "sum to mono" electrically you are guaranteed destructive interference of any out of phase content and guaranteed to lose information unless the bass is already mono. How important that is is open to debate, you can search and read about it on these forums, there is no consensus on this but lots of arguing and strong opinions.

 

[janbth]

Lots of people swear by Stereo Bass these day and it has been proven using Deltaware that "most music is summed to mono" is a myth, even for vinyl. One thing for sure if you "sum to mono" electrically you are guaranteed destructive interference of any out of phase content and guaranteed to lose information unless the bass is already mono. How important that is is open to debate, you can search and read about it on these forums, there is no consensus on this but lots of arguing and strong opinions.

I'm well aware of the discussions about "stereo bass", directional bass, and AEV in other threads. I consider that an advanced topic. I don't deny the existence of non-mono content in music, but I have yet to be convinced about how relevant it is. Many people - including me - can't have directional bass due to a non-optimal setup (room modes, bad location of the speakers and subs, bad listening position). I think the best advice is to learn how to get proper mono bass first, then explore directional bass later.

 

[Pareto Pragmatic]

Personally I use stereo not L/R, but that's because my very old house has a floor that is pretty much a passive radiator, and I need to load that up as it would be energized for stereo.

That said, for other spaces I found L/R to be more accurate for a single listening position, but more "touchy" for any adjustments I wish to make to what REW suggests. Stereo measurements are more robust for EQ for me, in that I can touch things up multiple places without everything going wacky. L/R not so much.

I think asymmetrical spaces will benefit more from a L/R EQ approach than symmetrical ones.

You didn't mention it but I would suggest MMM measurements if you aren't doing it that way, much faster and easier

Even faster is if you have a single position that mimics MMM results for the most part. The one I use has a positional null at 40hz, but other than that looks pretty close to MMM.

 

[john61ct]

Personally I use stereo not L/R

What does "L/R" mean besides Left and Right channels in stereo?

Summing them to a mono signal, same content sent to both sides?

 

[Pareto Pragmatic]

What does "L/R" mean besides Left and Right channels in stereo?

L/R is separate measure and eq, stereo is both firing at once. That was my intended distinction.

 

[john61ct]

Aha thanks.

So I thought measuring separately AND together was the norm.

Actual filtering, each speaker as needed "as anechoic as possible" EQ

separately from crossovers, which one would think usually mirrored.

But checking timing delays / phase alignment tuning MUST be done together?

All of which my use case requires keeping separate from room compensation EQ

 

[Pareto Pragmatic]

So I thought measuring separately AND together was the norm.

I think the norm in society is not to measure. For all people playing music. For those who do, I would guess both speakers at the same time is most common. But ASR is WAY outside the norm as a sub population, that's for sure. Out in the world people think I am an expert, here I am barely competent at the basics.

I do think measuring separately AND together is best practice. At least once, to see how it goes and what works in a give room for a given listener. I would certainly do both if it made my results better. But it doesn't, not for me, not in my room.

 

[john61ct]

Yes, I meant out of those using REW and other filter design tools that have measuring tools built in.

I would certainly do both if it made my results better. But it doesn't, not for me, not in my room.

Measuring has no "results". It's just one information input to help inform creation of crossovers (magnitude FR vs )

or "compensation EQ"

whether of the speaker / driver-level response ("as anechoic as possible")

or the room response.

There are dozens of DSP filter creation tools oriented toward one or more of these goals

varying in compatibility with many different convolver tools.

Each of which has its own learning curve, and different approaches to how and what and why to measure.

Some people just use REW to help with their physical room treatments.

All this just say, whatever your purpose in measuring, you've only scratched the surface of that aspect, keep an open mind.

 

[marin.weigel]

... 10-band parametric EQ beforehand ... whether it’s best to measure the room response of both speakers simultaneously, or to measure each speaker individually. On the EQ side, I can separate left and right ...

Obviously, it's about room acoustics correction here.
And since the correction will be done using Wiim's 10 band PEQ only, any single driver or phase correction is out of the question, so moving mic measurements are most practical to not try and compensate for relatively confined nulls and peaks at a single measurement position.
It's definitively very informational to compare separate l/r graphs due to the differing speaker positioning, and advisable to use separate corrections if dissimilar by more than, I'd arbitrarily say, 3dB in any given band while applying psychoacoustical smoothing.
Checking the net result by combined l/r mmm is mandatory.
As a general recommendation, try to avoid boosts, as well as too high (narrow band) q-values, and concentrate more on the frequency range of lower mids and bass.
Above all, try to ameliorate acoustical problems as much as feasible by means of listener/speaker positioning and physical room treatments before EQ'ing.