Room correction, speaker correction anything above Schroeder a mistake?

[Blumlein 88]

Discuss.

The Schroeder frequency is the boundary between reverberant sound and a region dominated by widely space room modes. Obviously not a fine exact line. Below the Schroeder frequency the room responds with areas of resonance and nulling which will depend upon the location of the source, the listener and room dimensions. Above that is a reverberant sound with closely spaced responses that are diffuse in nature. For most domestic listening rooms this frequency will be 300 hz or less.

Room correction attempts to fix the speaker/room interaction so that you get proper response at the listening position. Some say doing this above the Schroeder frequency is a bad idea because you can't fix the combined result of room and reflections by altering just the speaker.

Other's think you should strive to correct for speaker response aberrations, not the room and speaker combined. Or in other words to make the speaker respond correctly for anechoic conditions. This with no attempt to fix the response found at the listening position when such a correct in anechoic conditions speaker displays non-flat response at the listening position.

 

[amirm]

One key issue is how the measurements are made. The resolution of our hearing progressively gets lower and lower as frequencies go up so the frequency response needs to be measured that way. Without it, you will be fixing troughs and peaks that are not audible.

The other point is that the speaker dominates the response in higher frequencies. So poor response there is best fixed using a better speaker.

 

[svart-hvitt]

I thought this old paper may be of interest:

https://www.genelec.com/sites/default/files/media/About Us/Academic_Papers/2001_makivirta_anet.pdf

Genelec found that their neutral speakers had very different in-room responses. No surprise, really, but science is much about documenting. After that finding, published in 2001, Genelec set out on a journey involving DSP room correction.

I find it intriguing that non-experts in the area of audio and perception use metaphors and philosophy 101 logic to entertain the position that Genelec (and other DSP producers) has followed a dead end for almost 20 years now.

I am a non-expert myself, so I am left to imitatate the behaviour of experts. I also find interest in understanding the process of competent products that implicitly contain intelligent behaviour in a standardised format. As such I believe there is a lot to learn from the DSP process of Genelec.

It leaves me wondering: What gives the best results for non-experts in technical areas:

1) Aping the behaviour of experts?
2) Using your own logic?

Back to the question in the OP: Genelec uses little correction in higher frequencies in their automated software. And they don’t fill out nulls.

 

[Wombat]

I don't see the room being corrected with DSP. The speaker is being equalised to give a more acceptable signal at a specific position in the room(see my avatar for preciseness).

Room correction involves modifying the room characteristics, not those of the speaker. Room 'compensation' is a better way to describe the DSP approach.

 

[Bjorn]

You can only correct for the response at a single point. And unless all points in coverage are similar, then correcting this single point will not correct the total sound field.
The speaker's directivity alone, which varies as we know, is reason enough for room correction not to work. Not to mention that most of the response from the room isn't minimum phase but varies in both time and strength.

No acoustic problem can be fixed with EQ/DSP unless it's minimum phase. You can't deal with a three dimensional problem using a one dimensional solution.

 

[oivavoi]

We've had this discussion before, I think

My own two cents from experience:
- The trueplay correction of Sonos really does wonders for my kithcen setup (two play:1 and a sub). I suspect most of this is about evening out large peaks in the bass, which became really disturbing when I added the sub. I don't know how much it corrects further up, but I assume it does some correction there as well.
- With dirac live in my previous setup in the living room, I thought it added a slight artificiality to correct above the schroeder frequency. Below the schroeder frequency it did wonderful, wonderful things.
- I auditioned the Grimm LS1 at an excellent room in a hifi shop. One of the best speaker setups I've heard. Tried with and without room correction with Trinnov. Sounded wonderful without Trinnov (even in the bass), sounded bizarre with Trinnov.
- I heard the Genelec 8351 at the place of a fellow forum member in Norway Tried both with and without Genelec's room correction. Sounded good both ways, but I thought I perceived a certain lack of punchiness or "freshness" with the room correction on, which disappeared with the room correction turned off. These differences were small though, so it might have been just biased perception.
- One of my best hifi experiences ever, bar none, was listening to the Beolab 50, in narrow mode, with B&O's room correction turned on.

But at the moment, is there a "right" answer as to whether it makes sense to correct above the schroeder frequency? I'm not sure. And I think this is one of the areas where listening tests - both systematic and subjective - actually are needed. The issue is philosophical, as @Cosmik points out. Do we listen through the room or not? The question is how much weight our perception gives to the direct sound, vs. the total power response and reflections in the room. This is not a question which can be answered by logic or measurement alone, it is a question about our psychoacoustic hearing.

My own tentative conclusion is that I think that the ideal is to do without room correction above the schroeder frequency, as any correction by definition will change the frequency response of the direct sound, and make the direct sound less fidelitous to the signal. This will necessitate having very good speakers which are corrected for frequency and time etc at the source, have even dispersion with frequency, and generally are as good as possible. It also necessitates a room which is large enough, I think ideally with high ceilings and interiors which are not too reflective (lots of carpeting, book shelves, etc will be good).

But in many cases, I think room correction will be subjectively advantageous. This will be the case when speakers are not excellent to begin with, and the room correction algorithm does a good job of correcting time and phase and frequency at all that. This means that a part of why the room correction works actually is that it does "speaker correction". I think room correction may also be advantageous if the room and loudspeaker placement are far from ideal, with reflections really messing things up.

Below the schroeder frequency, I'm very much pro room correction. It the room is large enough, and the room dimensions does that one lucks out with regards to standing waves, it may be that one can do without.

EDIT: But there really should be done more research into this!

 

[pastorbarrett]

I use Dirac via a DDRC 24 in my own setup. I hadn't read nor encountered the Schroeder effect before, however knew by ear that Dirac applied full range was doing some highly unpleasant things to the midband - it became glassy, reproduced timbre inaccurately and just sounded artificial to me. Just below 300hz, it's been positively transformative and I can't praise it enough. What a real, bona fide upgrade to my system it has been.

I think judicious application is key. A willingness to tinker and something of an ear helps too, as I feel Dirac's automated settings, even in the bass, were a little hot for me.

 

[Blumlein 88]

You can only correct for the response at a single point. And unless all points in coverage are similar, then correcting this single point will not correct the total sound field.
The speaker's directivity alone, which varies as we know, is reason enough for room correction not to work. Not to mention that most of the response from the room isn't minimum phase but varies in both time and strength.

No acoustic problem can be fixed with EQ/DSP unless it's minimum phase. You can't deal with a three dimensional problem using a one dimensional solution.

On the other hand, is that the problem imagined? I say not.

I have in the past Room Corrected a 2 way box speaker on the left channel, and full range electrostat on the right. I did make minor adjustments to the bass target curve to prevent problems for either speaker. I only measured one spot in the center listening position. It was amazingly good. As in you didn't get the sense of two different speakers. Imaging was good etc. You could move around some (about 3 persons wide) without the illusion falling apart. I wouldn't call it a perfect match, but it seems to belie the idea single point can't move you closer to your target response and that nearby the response will be way off.

Now I would say I wasn't really "fixing" the room. I was balancing the FR of the speakers at the listening position. You know those measurements are windowed. That doesn't fix effects from room reflection, but it helps. The device used two pulse for measurements one for lower frequencies and one for higher.

 

[svart-hvitt]

Bjorn, Cosmik,

you seem to be a couple of the guys that represent the position that for example Genelec is on a non-science route to harmful audio signal processing.

However, it seeems to me that the producers you criticise have scientific papers to support and document their R&D efforts while you are more opinion based?

Is this a classic audiophile discussion where opinion meets science?

 

[Wombat]

YES. More science please and contributors stop relying on just Toole, especially in a field that obviously has many variables related to listening environments and much uncertainty once the sound crosses the near-ear region and is subject to individual processing and perception.

Can there be one method that covers all situations?? I doubt it.

 

[Bjorn]

Bjorn, Cosmik,

you seem to be a couple of the guys that represent the position that for example Genelec is on a non-science route to harmful audio signal processing.

However, it seeems to me that the producers you criticise have scientific papers to support and document their R&D efforts while you are more opinion based?

Is this a classic audiophile discussion where opinion meets science?

I have just explained from a scientific point of view why you can't correct what's not minimum phase. Either you understand it or you don't, but it's certainly scientific based. However, introducing anomalies while perhaps fixing some sound better to some.

 

[Purité Audio]

It is really easy to test, just download Dirac’s Free trial offer, make a full range correction and then as many ‘partial’ FR corrections as you please, then A/B.
When I used Dirac I found just asking it to deal with the very lumpiest low bass worked best for me.
I was and to some extent remain a huge fan of EQ, but now I find less is more.
Keith

 

[svart-hvitt]

I have just explained from a scientific point of view why you can't correct what's not minimum phase. Either you understand it or you don't, but it's certainly scientific based. However, introducing anomalies while perhaps fixing some sound better to some.

Well, you also wrote that «...is reason enough for room correction not to work». That’s a pretty strong statement.

And please come up with peer-reviewed research to support your position. Otherwise, it will just be an opinion from an audiophile.

What I have in mind regarding supporting research is something like this:

https://www.genelec.com/academic-papers

It’s an example of how one producer has decided to support its design position, including DSP based room correction.

As a commercial provider of audio products and services you should have the academic material supporting your position readily available.

 

[Cosmik]

Bjorn, Cosmik,

you seem to be a couple of the guys that represent the position that for example Genelec is on a non-science route to harmful audio signal processing.

However, it seeems to me that the producers you criticise have scientific papers to support and document their R&D efforts while you are more opinion based?

Is this a classic audiophile discussion where opinion meets science?

It's a fascinating question in itself: can opinion trump 'science'? I think it can. Empirical science never answers a question without also being plastered with caveats, assumptions and conditions.

Regarding Geddes, I was trying to find the relevant papers regarding 'room correction' above the Schroeder frequency ('RCASF') but found this where he is quoted as saying:

"...once one has a well designed loudspeaker, there is no need for electrical EQ and in fact it is likely to make things worse. Acoustic problems can only be solved acoustically, and once that is done, nothing more is possible.”

Do you have some links to where he thinks different?

 

[Wombat]

Well, you also wrote that «...is reason enough for room correction not to work». That’s a pretty strong statement.

And please come up with peer-reviewed research to support your position. Otherwise, it will just be an opinion from an audiophile.

What I have in mind regarding supporting research is something like this:

https://www.genelec.com/academic-papers

It’s an example of how one producer has decided to support its design position, including DSP based room correction.

As a commercial provider of audio products and services you should have the academic material supporting your position readily available.

Any dissenting research is presentable, also.

Commercial/self interest requires evidential rigour. Scientific research, sales, marketing and product reviewing come to mind. These inputs are part of this forum and should always be held to firm scrutiny, otherwise we are merely a run-of-the-mill opinion forum.

 

[Bjorn]

s.
Regarding Geddes, I was trying to find the relevant papers regarding 'room correction' above the Schroeder frequency ('RCASF') but found this where he is quoted as saying:

Do you have some links to where he thinks different?

Geddes understands this very well. So does Floyd Toole. They understand that a directivity of a speaker changes and isn't 100% constant, thus you can only correct for a single point and not for the rest of the coverage from the speaker. It's actually very basic if one understand the fundamentals. And that's only looking at the speaker, not really including the impedance and time differences from surfaces of the room. It's really completely irrelevant what Genelec or any other companies have written, this is fundamental acoustic understanding.

By the way, what the thread starter said in the opening and quoted below is wrong.

The Schroeder frequency is the boundary between reverberant sound and a region dominated by widely space room modes. Obviously not a fine exact line. Below the Schroeder frequency the room responds with areas of resonance and nulling which will depend upon the location of the source, the listener and room dimensions. Above that is a reverberant sound with closely spaced responses that are diffuse in nature. For most domestic listening rooms this frequency will be 300 hz or less.

 

[svart-hvitt]

It's a fascinating question in itself: can opinion trump 'science'? I think it can. Empirical science never answers a question without also being plastered with caveats, assumptions and conditions.

Regarding Geddes, I was trying to find the relevant papers regarding 'room correction' above the Schroeder frequency ('RCASF') but found this where he is quoted as saying:

Do you have some links to where he thinks different?

People with opinions do science, but they are forced to back their opinions with empirical data and/or mathematical logic as soon as they enter the faculty doors.

In other audio forums people have strong statements without any support from academic articles. This place should be different. If you have a controversial claim, you should be able to support it with an academic article or two.

I use Genelec as an example of a producer that has both skin in the game and produce academic articles. Some people have academic experience, but no skin in the game. Others have skin in the game, but no academic experience. As a non-expert in this field I have respect for people who have both skin in the game and academic contributions. So I try and learn from the implicit behaviour in for example Genelec’s products.

Geddes is an interesting guy, but I cannot tell if your quote was from outside or inside of the «faculty doors». On a science site, it’s more interesting to look at statements that were made inside of the faculty.

 

[Cosmik]

People with opinions do science, but they are forced to back their opinions with empirical data and/or mathematical logic as soon as they enter the faculty doors.

In other audio forums people have strong statements without any support from academic articles. This place should be different. If you have a controversial claim, you should be able to support it with an academic article or two.

Some questions, though, are 'philosophical' in nature even if they look ostensibly scientific.

Should amplifier designers aim for zero distortion? 'Science' would make a dog's dinner of this question, finding that people seemingly preferred distortion in many cases. It would cloud the issue with questions of higher order versus lower order, feedback or no feedback, probably implying that it is a mistake to try to eliminate distortion because of possible side effects. You and I would probably just answer 'Yes'. Would we be wrong? Could science 'prove' we were wrong?

Does simulated stereo based on mono sound better than true stereo? Again, science would embark on an indefinite series of listening tests that might come to the conclusion that people 'prefer' simulated stereo. But they would, in effect, be attempting to answer the wrong question. The right question would be: Should a stereo system comprise two separate, discrete channels? The answer is simply 'Yes'. Science is not applicable, because we are defining a man-made system.

In this thread, you can phrase the question in many ways, too. "Can RCASF improve the frequency response at the listener's ears?". To many people the science - based on microphones and laptops - would say yes. This is based on an assumption that human hearing resembles a laptop and microphone. But to anyone thinking about it more deeply, the answer is much less simple. Even if the conclusion to this question was 'yes', it would not follow that the answer to "Will RCASF make my system sound better?" was also yes. Science is a pretty dumb instrument when it comes to working out what is going on in people's biological computers.

I quoted Toole earlier as saying:

Reflections within listening rooms are real and numerous. Some would argue that they all are problems to be eliminated. Others take a more philosophical view that they just provide information about the room, and the brain can figure it out. I’m somewhere in the middle, but leaning towards the latter. The science that has been done so far seems to be on my side.

It seems obvious to me that even he cannot answer the question using science, and is simply arguing rationally i.e. stating his 'opinion' if you like.

 

[svart-hvitt]

Some questions, though, are 'philosophical' in nature even if they look ostensibly scientific.

Should amplifier designers aim for zero distortion? 'Science' would make a dog's dinner of this question, finding that people seemingly preferred distortion in many cases. It would cloud the issue with questions of higher order versus lower order, feedback or no feedback, probably implying that it is a mistake to try to eliminate distortion because of possible side effects. You and I would probably just answer 'Yes'. Would we be wrong? Could science 'prove' we were wrong?

Does simulated stereo based on mono sound better than true stereo? Again, science would embark on an indefinite series of listening tests that might come to the conclusion that people 'prefer' simulated stereo. But they would, in effect, be attempting to answer the wrong question. The right question would be: Should a stereo system comprise two separate, discrete channels? The answer is simply 'Yes'. Science is not applicable, because we are defining a man-made system.

In this thread, you can phrase the question in many ways, too. "Can RCASF improve the frequency response at the listener's ears?". To many people the science - based on microphones and laptops - would say yes. This is based on an assumption that human hearing resembles a laptop and microphone. But to anyone thinking about it more deeply, the answer is much less simple. Even if the conclusion to this question was 'yes', it would not follow that the answer to "Will RCASF make my system sound better?" was also yes. Science is a pretty dumb instrument when it comes to working out what is going on in people's biological computers.

I quoted Toole earlier as saying:


It seems obvious to me that even he cannot answer the question using science, and is simply arguing rationally i.e. stating his 'opinion' if you like.

When a question is particularly difficult, like the perception of hi-res vs lo-res, one can resort to the meta-analysis. Having a good grip on the direction in research could be useful if you want to follow a weight-of-the-evidence approach.

I have the impression that academic research focuses more on room correction now than, say, 10 years ago. I stand corrected if you can provide support that acadmia is about to leave DSP based room correction.

 

[Thomas savage]

YES. More science please and contributors stop relying on just Toole, especially in a field that obviously has many variables related to listening environments and much uncertainty once the sound crosses the near-ear region and is subject to individual processing and perception.

Can there be one method that covers all situations?? I doubt it.

I endorse this, if we can present evidence along with ‘ what we thinks obvious ‘ it will help the quality of discussion. If for no other reason than it makes it easier to follow for the many reading . We are here to create educational content after all.

Cheers