Best Room Response

[levimax]

When Toole says a properly designed speaker doesn't need EQ above 400 Hz I believe he means a speaker which is flat on axis. It seems reasonable to conclude if the speaker isn't flat on axis then using EQ to make it flat should make it sound better. I have found this to be true with the LS50 bump at 2200 Hz.

I suppose a post EQ response curve could be run through Olive's formula and the result would be a higher preference score.

I think the question is can you accurately enough measure a speakers FR response at the listening position separate from the room in order to make the "speaker adjustment". For speakers with published accurate on axis data you can just use that .... for the vast majority of speakers you would need to rely on measurements with REW or similar. As far as I know no one has verified that EQ'ing these type of in room measurements above Schroeder results in preferred sound quality.

 

[andreasmaaan]

P.S. In your simple formula RT60 is acting as a frequency related transfer function, which is also related to walls characteristics (primary material density), and it is certainly not a constant

In Schroeder's "simple formula" both V and T are variables.

Where did I say otherwise?

You said (my bold):

The ratio of absorption vs reflection vs transmission within typical room changes very much as you climb the frequency range. And that is the only reason why we have differnt room influence on speaker's response in the 20-300 vs 300-900 vs 900-20kHz ranges

This falsely implies that T is a variable but V is not.

I posted Schroeder's formula to correct that.

 

[andreasmaaan]

Hi @QMuse, I've finally had the time today to gather up some of the evidence you asked me for earlier.

Firstly, Olive in et al in 2003 (my bold):

Finally, Fig. 19 shows the measured in-room response of the Revel F208 (black) equalized to the preferred target response curve. This measured response is very similar to the in-room loudspeaker target (dotted red) that was preferred by listeners in two previous studies where they evaluated different loudspeaker-room correction products [5], and different headphone target response curves [6]. Also shown in Fig. 19 is the predicted in-room response of the loudspeaker based on anechoic measurements (see Fig. 1). Above 200 Hz there is good agreement between the predicted in-room response of the loudspeaker (based on anechoic measurements with no room equalization), and the measured in-room response of the loudspeaker equalized to the preferred target response. What this tells us is that a well-designed loudspeaker shouldn’t require much equalization above the transition frequency where the room no longer dominates the quality of sounds heard. However, below 100-300 Hz, the loudspeaker will likely need equalization to deal with room mode and boundary effects, and possibly some bass enhancement to satisfy the tastes of individual listeners, and accommodate variations in the quality of program material.

You'll see that the preferred in-room response in that study tracked the predicted in-room response of a neutral loudspeaker more or less perfectly above 250Hz.

Here are relevant graphs Toole's Sound Reproduction (my added annotations in red):

And most importantly for the present discussion:

Note the worst-case scenario transition region shown for a small domestic room of 70m3. It is far lower in frequency (200-300Hz) than the 400-800Hz you're suggesting room correction should be carried out in.

Toole repeats many times across his body of work the following basic idea, which is ofc familiar by now:

Above the transition frequency we hear a combination of the direct sound and many reflections. Steady-state measurements in a room are not definitive because the physical interaction of sounds at a microphone is very different from the perceptions arising in two ears and a brain. Visually alarming irregularities in measurements are frequently not heard.

Consistently, he refers to this transition frequency as typically being in the range of 100 to 300Hz, as you know.

Look, I think overall we are really just in disagreement about (1) what you're measuring and (2) where the transition region typically is.

With regard to (1), my view is that, if you are consistently finding that a resonance is appearing above 400Hz across a large number of averaged MMS measurements, it is most likely a resonance in the loudspeaker itself, and not a room effect.

I further believe that, if it's a loudspeaker resonance, it's obviously best to correct it. If it's a room resonance, despite the direct sound from the speaker being flat, best leave it alone.

With regard to (2), perhaps you could explain why in your view the transition region extends up to 800Hz or so in a typical room?

 

[QMuse]

I think the question is can you accurately enough measure a speakers FR response at the listening position separate from the room in order to make the "speaker adjustment".

If you take a look at the measurements I posted on previus page you can see that it demonstrates that it is possible to measure accurately speaker's response at LP separate from the room even at frequencies at 650Hz. Below 300Hz you can measure nierfield drivers and port response. Upward from 900Hz you can make gated measurement from 50-120cm, but the 300-800Hz region is problematic and that is why I am trying to shed some light on it with using precise MMM.

 

[QMuse]

I further believe that, if it's a loudspeaker resonance, it's obviously best to correct it. If it's a room resonance, despite the direct sound from the speaker being flat, best leave it alone.

Sorry, but none of the things you posted so far proves it is best to leave alone room resonance at say 500Hz. Controllled listening tests are needed to support such statement and I have no information that something like that has ever been done.

With regard to (2), perhaps you could explain why in your view the transition region extends up to 800Hz or so in a typical room?

Because room resonances can be found up to 800Hz. The only difference is that they are smaller. I'll try to make some measurements to prove it in the next few days.

 

[QMuse]

An interesting chart from SynAudCon back in the day when I was dealing with studio control room acoustics. As @QMuse has mentioned, the transition between wave acoustics and specular reflections is not a switch, but a gradual transition:

View attachment 60181

@andreasmaaan I noticed you conviniently ignored this post although it is essential for our discussion.

 

[andreasmaaan]

Sorry, but none of the things you posted so far proves it is best to leave alone room resonance at say 500Hz.

Sorry, but nothing you've said so far "proves" that a loudspeaker response should be made nonlinear at frequencies as high as 800Hz in the pursuit of correcting a room nonlinearity.

Why don't you try to prove that?

@andreasmaaan I noticed you conviniently ignored this post although it is essential for our discussion.

I did not comment on this because I have already agreed, from the beginning, that it is correct.

 

[QMuse]

Sorry, but nothing you've said so far "proves" that a loudspeaker response should be made nonlinear at frequencies as high as 800Hz in the pursuit of correcting a room nonlinearity.

Obviously room EQ filters below 300Hz are also making loudspeaker nonlinear. I don't see a point in stating something obvious like that.

Why don't you try to prove that?

As it cannot be done without controlled listening tests where 2 room EQ strategies would be compared I obviously can't do it as well.

Let's me just clear one detail with you: are you saying that room modes don't exist above 300Hz or you're saying that they exist but should be ignored and not EQ'd?

 

[andreasmaaan]

Let's me just clear one detail with you: are you saying that room modes don't exist above 300Hz or you're saying that they exist but should be ignored and not EQ'd?

The latter. I've been saying they exist (indeed all the way up to 20kHz+ if surfaces are reflective at those frequencies) from the begining.

 

[andreasmaaan]

As it cannot be done without controlled listening tests where 2 room EQ strategies would be compared I obviously can't do it as well.

Then perhaps the standards of proof you demand for my position are similarly too high?

 

[QMuse]

Then perhaps the standards of proof you demand for my position are similarly too high?

Oh, nothing personal - I was merely pointing out that a controlled listening test where different room EQ strategies were compared hasn't been done.

Which is actually a pity, as even if we invest in a speaker with nice looking spinorama we still have to put it in our room where 2 of them become single acoustic instrument, as acoustic guitar and the strings. And so often we are left with very little choices of room treatment and speaker's positioning leaving room EQ as our only weapon to make a good sound. I have an impression that this area (how to make proper measurements, what to EQ in which range etc.) would benefit from further research.

 

[andreasmaaan]

Oh, nothing personal - I was merely pointing out that a controlled listening test where different room EQ strategies were compared hasn't been done.

Which is actually a pity, as even if we invest in a speaker with nice looking spinorama we still have to put it in our room where 2 of them become single acoustic instrument, as acoustic guitar and the strings. And so often we are left with very little choices of room treatment and speaker's positioning so room EQ is left as our own weapon to make a good sound. I have an impression that this area (how to make proper measurements, what to EQ in which range etc.) would benefit from further research.

I certainly agree with that

 

[QMuse]

The latter. I've been saying they exist (indeed all the way up to 20kHz+ if surfaces are reflective at those frequencies) from the begining.

That 300Hz hard limit is the thing I'm not ready to accept. If you agree with the stuff Mitch posted which basically states that room gently transform it's behaviour from modes to diffusion than one would expect that if we perceive room EQ in the 20-300Hz region than in the next region we would partially perceive it while later we won't perceive it at all. In other words, I would expect our perception of room EQ effects to gradually fade instead of being brutally switched off at 300Hz.

 

[levimax]

If you take a look at the measurements I posted on previus page you can see that it demonstrates that it is possible to measure accurately speaker's response at LP separate from the room even at frequencies at 650Hz. Below 300Hz you can measure nierfield drivers and port response. Upward from 900Hz you can make gated measurement from 50-120cm, but the 300-800Hz region is problematic and that is why I am trying to shed some light on it with using precise MMM.

I have had the same results as you with MMM and RTA with consistent and repeatable results. I also subjectively prefer the results of wide band EQ based on these measurments. Mr Toole , by limiting his listening tests to "well designed" speakers, has created an interesting situation where his work can be taken out of context to say 'EQ above Shroeder should not be done". I think the answer is not so cut and dry and more studies need to be done.... not sure who would do these though.

 

[andreasmaaan]

That 300Hz hard limit is the thing I'm not ready to accept. If you agree with the stuff Mitch posted which basically states that room gently transform it's behaviour from modes to diffusion than one would expect that if we perceive room EQ in the 20-300Hz region than in the next region we would partially perceive it while later we won't perceive it at all. In other words, I would expect our perception of room EQ effects to gradually fade instead of being brutally switched off at 300Hz.

I totally agree there is no hard limit and that there's a transition region.

My main argument is that, at say 600Hz, the distance between the peak of a room mode and its null will be around 30cm.

If you use EQ to knock down the peak of that mode, this can only possibly be valid over a listening position range of less than 30cm.

30cm from the listening position, your room correction is going to have the opposite effect to that intended.

TBH, given that, I can't see what else there is to discuss

IIUC, you're saying that in your room you measured peaks in the midrange that, even when averaged across a large number of different positions in the room, including positions well away from anywhere you normally listen, were clearly still present. I know you know what you're doing when it comes to MMS measurements.

My hypothesis is that perhaps what you've measured is a nonlinearity in the speaker itself. Of course that's only speculative. But a 600Hz mode will have around 16 peaks and 16 nulls just across one length of a room of 4m.

Given that, my two questions for you are:

(1) If what you've measured at 600Hz is a mode, is changing the (let's assume for argument's sake) already-linear response of the speaker a good idea given the mode's peaks and dips are closer together than your normal head movements are likely to be when listening?

(2) If the peak you've measured is not a mode and not a speaker nonlinearity, what is it?

And assuming you have an idea of an answer to (2), why then should it be corrected?

 

[andreasmaaan]

To both @QMuse and @levimax, I don't interpret Toole as saying "no EQ above 300Hz".

What IMO he's saying, and the point I've been trying rather unsuccessfully to make, is essentially "no room EQ above (around) 300Hz; only speaker EQ (if warranted) in this region."

 

[QMuse]

My hypothesis is that perhaps what you've measured is a nonlinearity in the speaker itself. Of course that's only speculative. But a 600Hz mode will have around 16 peaks and 16 nulls just across one length of a room 4m wide.

I don't think that is a point. The point is that amplitude of the room's response to ideally super-linear speaker would be say 2dB higher at 600Hz than at say 450 and 750Hz.

Ok, 600Hz is a little stretched, I choosed that frequency only to demonstrate I can detect introduced resonance in speaker with MMM, but I don't really have any room modes there. The highest room mode I am able to measure in my room for which I'm sure it is not coming from speaker is at 400Hz, 3dB with Q=5.

 

[QMuse]

To both @QMuse and @levimax, I don't interpret Toole as saying "no EQ above 300Hz".

What IMO he's saying, and the point I've been trying rather unsuccessfully to make, is essentially "no room EQ above (around) 300Hz; only speaker EQ (if warranted) in this region."

Sure, and that is how I understood you. I am only advocating that 300Hz is IMHO too low and that this limit should be moved to 400 or maybe even 450Hz.

 

[andreasmaaan]

I don't think that is a point. The point is that amplitude of the room's response to ideally super-linear speaker would be say 2dB higher at 600Hz than at say 450 and 750Hz.

Ok, 600Hz is a little stretched, I choosed that frequency only to demonstrate I can detect introduced resonance in speaker with MMM, but I don't really have any room modes there. The highest room mode I am able to measure in my room for which I'm sure it is not coming from speaker is at 400Hz, 3dB with Q=5.

Sure, and that is how I understood you. I am only advocating that 300Hz is IMHO too low and that this limit should be moved to 400 or maybe even 450Hz.

Well that's interesting, as the original post of mine that you objected to and which started this whole discussion was:

I was talking only about using EQ to correct your system, but what I was saying was that you should either:

1. correct only below 300-400Hz (i.e. correct only for the room) or
2. correct the room below 300-400Hz and correct the speakers above 300-400Hz

PS. I made a mistake in post #115 which I've just corrected. Don't think the basic concept is any different, though.

 

[levimax]

To both @QMuse and @levimax, I don't interpret Toole as saying "no EQ above 300Hz".

What IMO he's saying, and the point I've been learning trying rather unsuccessfully to make, is essentially "no room EQ above (around) 300Hz; only speaker EQ (if warranted) in this region."

I understand that you understand Toole but many "internet sources" quote him as " no EQ above 300 Hz" which is oversimplified.