Room correction, speaker correction anything above Schroeder a mistake?

[Cosmik]

Yes. Wherein lies the fallacy?

I'm not trying to catch you out!

What I wanted to do was to suggest that a possible way to meet the above criterion would be, for example, "Turn up the volume by 0.5dB". Even though it is 'scientifically proved to work', rational arguments against it could be made that would be superior to the science.

 

[svart-hvitt]

I'm not trying to catch you out!

What I wanted to do was to suggest that a possible way to meet the above criterion would be, for example, "Turn up the volume by 0.5dB". Even though it is 'scientifically proved to work', rational arguments against it could be made that would be superior to the science.

I see your point. Science plus common sense is an unbeatable proposition.

 

[jhaider]

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.

The Toole position has some empirical support.

"In these [Olive] experiments, trained listeners were able to draw curves of spectral trends— crude frequency responses— describing what they heard in the listening room. All of the high-scoring loudspeakers were described as having flattish spectra
(above the transition frequency), a trend that matches the flattish on-axis/ listening window curves for all of the corresponding anechoic measurements...Below the transition frequency, in the gray shaded area in Figure 5.15, it is evident that listeners responded to something closer to the low-frequency response in the room than to the anechoic frequency response."

Toole, Floyd E.. Sound Reproduction: The Acoustics and Psychoacoustics of Loudspeakers and Rooms (Audio Engineering Society Presents) (Kindle Locations 3355-3365). Taylor and Francis. Kindle Edition.

 

[svart-hvitt]

The Toole position has some empirical support.

"In these [Olive] experiments, trained listeners were able to draw curves of spectral trends— crude frequency responses— describing what they heard in the listening room. All of the high-scoring loudspeakers were described as having flattish spectra
(above the transition frequency), a trend that matches the flattish on-axis/ listening window curves for all of the corresponding anechoic measurements...Below the transition frequency, in the gray shaded area in Figure 5.15, it is evident that listeners responded to something closer to the low-frequency response in the room than to the anechoic frequency response."

Toole, Floyd E.. Sound Reproduction: The Acoustics and Psychoacoustics of Loudspeakers and Rooms (Audio Engineering Society Presents) (Kindle Locations 3355-3365). Taylor and Francis. Kindle Edition.

I thought there was some agreement that Toole is in group 2? (See post #35 for definition).

Besides, what’s meant by «crude frequency responses»? It’s recommended that speakers have a deviation of less than 0.5 dB from a flat curve in the mid to higher frequencies to be used in professional applications (broadcasting, film, I think it is).

My point, and provoking hypothesis, is this:

=> Toole & Co. were the first vintage of audio researchers, whose research program was focussed on documenting that people prefer «flat» speakers. This documenting exercise was used to convince speaker producers to build «flat» speakers. I guess this period can be traced back to the 1980s and 1990s?

=> The next vintage of audio researchers were not convinced that «flat» speakers is everything. They started to reflect upon the fact that frequency response in practical life deviated a lot from «flat» (more than 2 dB in all or most cases). So they wanted to compensate for the room interaction of the speaker. Genelec has been a pioneer in this respect.

=> Today’s vintage: First, it’s a fact that many speaker producers are stuck in the pre-Toole area. They build big beasts without documenting anything; no documentation of frequency response, radiation...nothing. Then comes the ongoing battle between those who own the Toole paradigme to the extent that they cannot buy into the «Genelec paradigm» of room compensation. They claim that neutrality is not only a safe choice; it’s the only choice! And you dare not use EQ to compensate for room interaction!

To me, it seems like today’s researchers are way beyond both Toole and the first «Genelec paradigm» of room compensation. Today’s resesarchers work on immersive audio and how to fool the senses into believing a story where audio is one of the ingredients. Maybe what describes today’s paradigm is fragmentation and specialisation? Serious researchers don’t talk about boxes anymore. Flat speakers? That’s a battle won in the 1980s. This is a bad period for audiophooles, but a golden period for audiophiles (I use the old definition of audiophile).

I don’t know what the future looks like, but I guess systems integration will be even more important.

Maybe a bit off-topic...I just got carried away...

 

[amirm]

=> Toole & Co. were the first vintage of audio researchers, whose research program was focussed on documenting that people prefer «flat» speakers. This documenting exercise was used to convince speaker producers to build «flat» speakers. I guess this period can be traced back to the 1980s and 1990s?

Just to be very correct , their research points to needing "smooth" frequency response, not flat. The in-room response can be sloping down for example. Dips in frequency response was found to be especially bad.

 

[svart-hvitt]

Just to be very correct , their research points to needing "smooth" frequency response, not flat. The in-room response can be sloping down for example. Dips in frequency response was found to be especially bad.

You’re correct, of course.

Maybe the Toole era should be called «the era of targeted speaker response», or something like that.

 

[Cosmik]

Just to be very correct , their research points to needing "smooth" frequency response, not flat. The in-room response can be sloping down for example. Dips in frequency response was found to be especially bad.

I see a distinction between "flat speakers" (what svart-hvitt was talking about) and flat "in-room response". Part of the confusion here?

 

[svart-hvitt]

I see a distinction between "flat speakers" (what svart-hvitt was talking about) and flat "in-room response". Part of the confusion here?

Elaborate, please.

 

[amirm]

I see a distinction between "flat speakers" (what svart-hvitt was talking about) and flat "in-room response". Part of the confusion here?

I don't think that is the confusion but yes, we are talking about in-room measurements where reflections are taken into account.

 

[DonH56]

A speaker with a ruler-flat response will generally exhibit a slight downward slope from LF to HF in an average room due to HF absorbers (chairs, people) that are present.

 

[Cosmik]

Elaborate, please.

An ideal speaker (in my and some other people's view) is one that would have a flat frequency response anywhere within its on- and off-axis 'window'. In an anechoic chamber this would measure as flat. In a real room, though, if the reflections were part of the measurement it probably wouldn't be flat - in terms of a simple frequency response measurement.

This is central to the "Steinway still sounds like a Steinway no matter which room it is in" argument. Place the genuinely neutral speaker in any room, and the listener will still hear it as a neutral speaker. The room may sound 'dead', 'live', 'bright', etc. but the speaker itself will still sound neutral.

 

[Fitzcaraldo215]

I think it is also at least partially true that many speaker guys don't want people to mess with the "natural" sound of their creations and as as the buyer heard them during audition. It may be a belief and a bias, rather than something they have experimented with. They cannot help but to acknowledge the issues in the bass below Schroeder, so EQ there might be OK. But, otherwise, they are not on board with full range EQ, some without regard to theory or empirical data, just mindset. Audio pundits often take the same view, which makes sense during speaker reviews, but, again, it is just mindset, except for Kal, of course.

In fairness, though, more and more speaker guys are designing DSP-enabled speakers, some designed from the ground up that way. Though still few in number, they are increasing. But, I am not sure if there is consensus about full range vs. just bass EQ.

Of, course, the EQ guys take the opposite view, and they promote the ability of their tool creations to benefit all speakers everywhere, some even forcing full range EQ all the time or no EQ at all. But, the better ones allow for customization of the target curve and/or EQ nullification in selected frequency ranges, at the risk of making the tool more complex. Some tools also default to leaving higher frequencies unEQed, but often well above Schroeder. And, of course, some cheapie tools are just half-baked. So, ideally, the choice is yours, as it should be.

Many audiophiles try full range, and they like it, though that opinion might be based on expectation bias. But, it is their system, so who really cares? It is not snake oil, and it makes a sonic difference, for better or for worse, but that is up to the beholder.

I like it full range, myself. In my system, I don't like the measurements or the sound without it. And, I am not ready to buy new speakers to try to fix the issue, even if that were easy and inexpensive to do. How much incremental improvement would I hear by replacing the speakers, anyway? That is a tough and risky road to follow.

 

[Rodney Gold]

Ive tried both ways with various systems , full range or low bass limited it's really all about subjective taste
Ive just got a trinnov ST2 in my grubby paws and am keen to give it a whirl over DIRAC .. its a much more complex and comprehensive system..it will still all boil down to taste however

 

[Cosmik]

Coming at it from a different angle.

Some people may be familiar with the concept of convolution.

In a room, the signal from the speaker is being convolved with the room's impulse response to give a result at a particular point in space. Mathematically, if the room's impulse response at that point is known, the signal can be deconvolved with that impulse response to get back to the original signal. If the room's impulse response is not known, the original signal can still be retrieved by a process of blind deconvolution aided by familiar 'cues' in the signal.

The standard audiophile view of 'correction' is to effectively play with a graphic equaliser in order to flatten the smoothed frequency response at the listening position. This is not the same as deconvolution; it does not give the same result.

I, and my fellow 'no correction above Schroeder' people, are suggesting that the human brain performs deconvolution on the signal in order to hear back to the speaker's output regardless of the room - it is an ability that has grown out of the evolutionary need to 'read' arbitrary acoustics and create a stable audio scene in the brain.

Therefore the almost information-free smoothed magnitude frequency response of the convolved signal does not tell you how the convolved signal will sound - as highlighted in Amir's comb filter article earlier. Counteracting the measured frequency response with a graphic equaliser will not make the speaker sound neutral at the listening position.

 

[Blumlein 88]

Just to be very correct , their research points to needing "smooth" frequency response, not flat. The in-room response can be sloping down for example. Dips in frequency response was found to be especially bad.

But isn't the downward sloping in room response an artifact of measurement technique? The anechoic measurement of a speaker with the right downward sloping in room measurement is flat is it not? Though yes they found smoothness of response surprisingly important.

 

[oivavoi]

I think it is also at least partially true that many speaker guys don't want people to mess with the "natural" sound of their creations and as as the buyer heard them during audition. It may be a belief and a bias, rather than something they have experimented with. They cannot help but to acknowledge the issues in the bass below Schroeder, so EQ there might be OK. But, otherwise, they are not on board with full range EQ, some without regard to theory or empirical data, just mindset. Audio pundits often take the same view, which makes sense during speaker reviews, but, again, it is just mindset, except for Kal, of course.

In fairness, though, more and more speaker guys are designing DSP-enabled speakers, some designed from the ground up that way. Though still few in number, they are increasing. But, I am not sure if there is consensus about full range vs. just bass EQ.

Of, course, the EQ guys take the opposite view, and they promote the ability of their tool creations to benefit all speakers everywhere, some even forcing full range EQ all the time or no EQ at all. But, the better ones allow for customization of the target curve and/or EQ nullification in selected frequency ranges, at the risk of making the tool more complex. Some tools also default to leaving higher frequencies unEQed, but often well above Schroeder. And, of course, some cheapie tools are just half-baked. So, ideally, the choice is yours, as it should be.

Many audiophiles try full range, and they like it, though that opinion might be based on expectation bias. But, it is their system, so who really cares? It is not snake oil, and it makes a sonic difference, for better or for worse, but that is up to the beholder.

I like it full range, myself. In my system, I don't like the measurements or the sound without it. And, I am not ready to buy new speakers to try to fix the issue, even if that were easy and inexpensive to do. How much incremental improvement would I hear by replacing the speakers, anyway? That is a tough and risky road to follow.

Is it possible that multichannel systems benefit even more from wideband correction than 2 channel systems? No idea whether it's indeed the case, just a hunch. Feels like all those channels can easily create a mess in a room, and that a central DSP/EQ command center could do wonders for keeping it tidy? That said, I have never listened to a dedicated multichannel system for music, so I don't know.

 

[svart-hvitt]

Coming at it from a different angle.

Some people may be familiar with the concept of convolution.

In a room, the signal from the speaker is being convolved with the room's impulse response to give a result at a particular point in space. Mathematically, if the room's impulse response at that point is known, the signal can be deconvolved with that impulse response to get back to the original signal. If the room's impulse response is not known, the original signal can still be retrieved by a process of blind deconvolution aided by familiar 'cues' in the signal.

The standard audiophile view of 'correction' is to effectively play with a graphic equaliser in order to flatten the smoothed frequency response at the listening position. This is not the same as deconvolution; it does not give the same result.

I, and my fellow 'no correction above Schroeder' people, are suggesting that the human brain performs deconvolution on the signal in order to hear back to the speaker's output regardless of the room - it is an ability that has grown out of the evolutionary need to 'read' arbitrary acoustics and create a stable audio scene in the brain.

Therefore the almost information-free smoothed magnitude frequency response of the convolved signal does not tell you how the convolved signal will sound - as highlighted in Amir's comb filter article earlier. Counteracting the measured frequency response with a graphic equaliser will not make the speaker sound neutral at the listening position.

You elaborate quite well.

It seems like you and AES fellow John Watkinson are in agreement?

 

[amirm]

But isn't the downward sloping in room response an artifact of measurement technique?

It is and that was one of the first points I made in this thread . Off-axis response of all speakers has less highs than the direct due to beaming (and furnishings as Don correctly noted) so when we measure and include reflections, then we don't want flat response. Excluding the reflections creates other problems in that we are not measuring what is heard.

 

[Cosmik]

It seems like you and AES fellow John Watkinson are in agreement?

I didn't know the name, but I realise that I have certainly read this article by him in the past.
http://www.theregister.co.uk/Print/2014/07/02/feature_the_future_loudspeaker_design/

 

[amirm]

Is it possible that multichannel systems benefit even more from wideband correction than 2 channel systems?

The research shows the opposite. In that the more the channels, the less picky we become in how the individual speakers sound.

Here is research that Dr. Olive et. al. performed in effectiveness of room equalization as the number of channels is increased: https://secure.aes.org/forum/pubs/conventions/?elib=14622

Comparison of Loudspeaker-Room Equalization Preferences for Multichannel, Stereo, and Mono Reproductions: Are Listeners More Discriminating in Mono?

X axis is different Equalization methods. I have highlighted the surround ratings in yellow. We see there that there is very little difference there between No Eq and the other methods. Whereas in mono (blue), No Eq rated far lower than other methods.

I think the surround presentation is so captivating and effective that we forego other problems with fidelity.