Unique room EQ for loudspeaker comparisons

[QMuse]

Are you able to try how the speaker would sound with the filters I made based on Amir's measurement with Klippel?

@amirm Do you by any chance still have the speaker with you so you can try it? This idea of EQ-ing the speaker based on your measurements sounds really intriguing to me, but somebody has to try how it works in real life.

 

[amirm]

@amirm Do you by any chance still have the speaker with you so you can try it? This idea of EQ-ing the speaker based on your measurements sounds really intriguing to me, but somebody has to try how it works in real life.

Yes, I have been wanting to do this. Let's plan it on a future speaker review. For now, my efforts in reduction of one or two peaks has worked.

 

[QMuse]

Yes, I have been wanting to do this. Let's plan it on a future speaker review. For now, my efforts in reduction of one or two peaks has worked.

Excellent, I'll be happy to assist. My proposal is to do a 2 step EQ - first EQ-ing the speaker based on your measurements with Klippel from 400Hz up and then room EQ for below 400Hz based on your MMM measurement from the LP where you listen the speaker. The idea is to evaluate by a trained listener how much of an improvement would that bring to SQ. If that turns to be usefull folks would be able to generate speaker's EQ filters based on the Klippel measurement you provide. This would of course be one time affair to verify the concept. Does this makes sense to you?

 

[aarons915]

Well, I think it is safe to say that a quote where Toole is explaining that room response below 300-400Hz should be EQ-ed has been quoted at least hundred of times on this forum. I find it difficult to believe that you are not familiar with it, or very surprised that you, as a member with Technical Expert badge, are not accepting it.

Toole doesn't say EQ is necessary for speaker comparisons, the Harman double blinds don't use room EQ for their evaluations either. I'm sure you're also aware of Dr. Toole's study that showed the same speakers were preferred in the same order in various rooms and is the basis for the phrase that we "hear through the room".

I don't think anyone would argue that EQ below the transition frequency won't make our systems sound better but I think it's just adding unnecessary variables to a speaker comparison and is pointless since it won't make a difference in preference.

 

[amirm]

Excellent, I'll be happy to assist. My proposal is to do a 2 step EQ - first EQ-ing the speaker based on your measurements with Klippel from 400Hz up and then room EQ for below 400Hz based on your MMM measurement from the LP where you listen the speaker. The idea is to evaluate by a trained listener how much of an improvement would that bring to SQ. If that turns to be usefull folks would be able to generate speaker's EQ filters based on the Klippel measurement you provide. This would of course be one time affair to verify the concept. Does this makes sense to you?

Yes. For now, I want to just keep my one PEQ for the room mode and no more for that. For the rest, we just correct based on Klippel data. I will then test each one of the filters individually to see how effective they are. And yes, definitely need your help to figure out the PEQ parameters as I am too lazy to compute them.

 

[Dennis Murphy]

Toole doesn't say EQ is necessary for speaker comparisons, the Harman double blinds don't use room EQ for their evaluations either. I'm sure you're also aware of Dr. Toole's study that showed the same speakers were preferred in the same order in various rooms and is the basis for the phrase that we "hear through the room".

I don't think anyone would argue that EQ below the transition frequency won't make our systems sound better but I think it's just adding unnecessary variables to a speaker comparison and is pointless since it won't make a difference in preference.

Thanks for clearing that up. I was really scratching my head at to what Floyd's reasoning might be. I would be interested in seeing the exact quote QMuse was referring to. My guess is that it relates to Toole's (and my) opinion that it's legitimate and very useful to apply EQ in the bass region for long-term listening purposes since the woofer's's output won't change significantly off axis, whereas EQing can be problematic higher up where the on and off-axis response can be very different.

 

[andreasmaaan]

Thanks for clearing that up. I was really scratching my head at to what Floyd's reasoning might be. I would be interested in seeing the exact quote QMuse was referring to. My guess is that it relates to Toole's (and my) opinion that it's legitimate and very useful to apply EQ in the bass region for long-term listening purposes since the woofer's's output won't change significantly off axis, whereas EQing can be problematic higher up where the on and off-axis response can be very different.

My understanding of Toole's position is that it is legitimate/useful to apply EQ in the bass region not specifically because the woofer's output won't change off-axis, but rather because the steady-state in-room response in this region correlates highly with perceived tonal balance, whereas at higher frequencies the direct sound (and to a lesser extent first reflections) correlate more strongly with perceived tonal balance than the steady-state in-room response.

To use a practical example to illustrate this: EQing of a cardioid subwoofer to correct in the steady-state in-room response would (if I understand Toole correctly) be considered just as legitimate/useful as EQ of an omnidirectional subwoofer, i.e. the relevant factor is not the subwoofer's directivity, but rather the (frequency-dependent) perceptual importance (or otherwise) of the in-room steady-state response.

 

[amirm]

I don't think anyone would argue that EQ below the transition frequency won't make our systems sound better but I think it's just adding unnecessary variables to a speaker comparison and is pointless since it won't make a difference in preference.

I have found an exception to this. To the extent you test one speaker with better low frequency extension than other, this speaker will excite room modes and muddy its sound. This is why I now perform all of my listening tests with one room mode PEQ active.

If speakers are equal in this regard, then it doesn't matter. Otherwise it actually penalizes the speaker with more bass!

 

[aarons915]

I have found an exception to this. To the extent you test one speaker with better low frequency extension than other, this speaker will excite room modes and muddy its sound. This is why I now perform all of my listening tests with one room mode PEQ active.

If speakers are equal in this regard, then it doesn't matter. Otherwise it actually penalizes the speaker with more bass!

Yes that's actually true and is also where the subwoofer myth started that ported Subs are sloppier and have "slow" bass relative to sealed subs since they generally have more output. I'm a bit different than you and many others when comparing speakers since I always use dual subs crossed over at 100Hz, I compare with the crossovers in place without the subs running and the bass isn't much of a factor in my listening tests. I realize that is a flawed method for people who don't use subs or want a more full range sound with towers but it makes sense for me.

 

[Dennis Murphy]

My understanding of Toole's position is that it is legitimate/useful to apply EQ in the bass region not specifically because the woofer's output won't change off-axis, but rather because the steady-state in-room response in this region correlates highly with perceived tonal balance, whereas at higher frequencies the direct sound (and to a lesser extent first reflections) correlate more strongly with perceived tonal balance than the steady-state in-room response.

To use a practical example to illustrate this: EQing of a cardioid subwoofer to correct in the steady-state in-room response would (if I understand Toole correctly) be considered just as legitimate/useful as EQ of an omnidirectional subwoofer, i.e. the relevant factor is not the subwoofer's directivity, but rather the (frequency-dependent) perceptual importance (or otherwise) of the in-room steady-state response.

I haven't run into this argument before, and unfortunately I left my Toole bible in a Santa Fe hotel, so I can't research it. But I do know as a fact that Toole argues in his frequently posted video that it's the variations in off-axis response that can make EQing higher up dicey, and I believe he even thinks it's an inherently flawed approach for most speakers. I have no practical experience in this department, but it's on my to-do-list. I guess I agree with Amir that if your listening and testing room has a big room mode, it might be worth EQing out when making speaker comparisons. But a complete make-over of the bass room response just doesn't make sense to me when reviewing, and I would be surprised if Floyd advocates that approach.

 

[andreasmaaan]

But I do know as a fact that Toole argues in his frequently posted video that it's the variations in off-axis response that can make EQing higher up dicey, and I believe he even thinks it's an inherently flawed approach for most speakers.

I completely agree with this, yeh.

It's just that I understand the argument he makes about EQing the lower frequencies to be based on the premise I mentioned above rather than woofer/subwoofer directivity per se.

I guess I agree with Amir that if your listening and testing room has a big room mode, it might be worth EQing out when making speaker comparisons. But a complete make-over of the bass room response just doesn't make sense to me when reviewing, and I would be surprised if Floyd advocates that approach.

On the same page here too

 

[Dennis Murphy]

I completely agree with this, yeh.

It's just that I understand the argument he makes about EQing the lower frequencies to be based on the premise I mentioned above rather than woofer/subwoofer directivity per se.

Now i'm getting a little confused (my usual state). If it's the on-axis and early arrival response that shapes our perception of tonality, why does Floyd stress the importance of even response even fairly far off axis?

 

[QMuse]

Yes. For now, I want to just keep my one PEQ for the room mode and no more for that. For the rest, we just correct based on Klippel data. I will then test each one of the filters individually to see how effective they are. And yes, definitely need your help to figure out the PEQ parameters as I am too lazy to compute them.

I'll be glad to help.

 

[QMuse]

Toole doesn't say EQ is necessary for speaker comparisons, the Harman double blinds don't use room EQ for their evaluations either. I'm sure you're also aware of Dr. Toole's study that showed the same speakers were preferred in the same order in various rooms and is the basis for the phrase that we "hear through the room".

Toole definitely advises room EQ but I'm done quoting him. There has been numerous quotes on that so you won't have trouble finding them.

I don't think anyone would argue that EQ below the transition frequency won't make our systems sound better but I think it's just adding unnecessary variables to a speaker comparison and is pointless since it won't make a difference in preference.

Unnecessary variables? What variables??? Room EQ is about taking room out of the comparison so you can compare speakers more accurately. Does your post mean you don't use room EQ in normal listening? Have you ever measured response of your speaker to find out exactly what response you're listening?

 

[thewas]

Most agree here that sensible EQing loudspeakers at the modal room region improves their perceived sound, but since there is no unique approach on how to do it, we would introduce another "random" variable. Of course we could agree to a standard doing it (for example which target curve, matched to which rest frequency level, do or don't increase extension, fill dips up to xx dB or just use the default correction of Audyssey, Dirac etc) but again those wouldn't help much people which use other systems or approaches. Also imho good listening tests take quite some time and any additional variation would multiply their time which in this stage seems very limited. On the other hand it would be more fair to a loudspeaker which "accidentally" has a dip or raise which compensated the idiosyncrasies of Amirs listening room and loudspeaker placement. These also show the problems of listening tests which at high loudspeaker levels are imho still quite subjective and room dependent in the end (just look at the famous M2 vs Salon 2 blind test, depending if Amir would belong to the 65% or 35% of the population he would give a different preference) , so if you ask me, ASR should rather concentrate on objective measurements and leave listening comments to others like members in the comment section.

 

[QMuse]

Most agree here that sensible EQing loudspeakers at the modal room region improves their perceived sound, but since there is no unique approach on how to do it, we would introduce another "random" variable.

Here is how I see it: "random" variable has already been introduced with acoustic characteristics of his room. Room EQ would only make that random variable less random and more controlled. With proper room EQ done for random rooms same speaker will still have in-room response similar to one predicted by Spinorama (PIR).

As same room and same room EQ would be used for all speakers I believe that all speakers would get a better chance of presenting themselves.

 

[thewas]

The problem I see is that not the exact same room EQ could be used for all speakers so it would be also partially a loudspeaker correction.

 

[QMuse]

The problem I see is that not the exact same room EQ could be used for all speakers so it would be also partially a loudspeaker correction.

I don't understand. Are we talking about room in which Amir is performing listening tests or..?

 

[thewas]

Yes, about Amir's room.

 

[QMuse]

Yes, about Amir's room.

Why do yu think same room EQ cannot be used for all speakers?