• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Genelec on audio science

svart-hvitt

Major Contributor
Joined
Aug 31, 2017
Messages
2,375
Likes
1,253
I found this piece, by Siamäk Naghian of Genelec, to be well written:

https://www.genelec.com/blog/its-not-right-its-not-even-wrong-why-scientific-foundation-matters

It sums up what I believe to be the views of many people on ASR.

And I thought especially this sentence about Genelec founder Ilpo Martikainen is of great interest:

"Regardless of his engineering background, his mind often zoomed in on subjects of social, cultural and philosophical importance".

PS: Picture of late Ilpo Martikainen (1947-2017), founder of Genelec, at the bottom.

Copy of whole text:

It’s not right. It’s not even wrong. Why scientific foundation matters.
Several years ago, I attended our annual distributor dinner event in Frankfurt, where Ilpo Martikainen offered an interesting speech on marketing. I had long learned to relish the moments when Ilpo shared his thoughts in this way – so unlike a typical commercially or technology-oriented company leader. Regardless of his engineering background, his mind often zoomed in on subjects of social, cultural and philosophical importance.

On the evening in question, he raised the importance of maintaining a factual foundation in marketing. This is particularly true, he argued, in the world of audio, where subjective viewpoints and individual perception can prove highly influential when critically evaluating audio quality – whether consciously or otherwise. In his speech, he referred to Harry G. Frankfurt’s books, “On Truth” and “On Bullshit”, both of which discuss the background and motivation for the misrepresentation of facts – as well as the dangers to our civilization of becoming indifferent to such activity.

Now we are in the era of social platforms and the data economy, enabled by unprecedented developments in the fields of data collection, computing and networking, and indifference to mistruth has become a real danger. Thus, for the good of society, it is more than ever before the responsibility of individuals, companies and other organisations and communities, to represent themselves faithfully and honestly.

At Genelec, we believe that the objective and subjective assessment of reproduced sound should always go hand-in-hand. The subject is a human, the setting is planet Earth, and the question is “what are we able to hear, given enough time?”. Repeatability and falsifiability are equally important qualities in distinguishing facts from fiction, hence the headline of this blog, a quote from Austrian physicist, Wolfgang Pauli: “A claim may be worse than wrong if it neither can be proven correct, nor falsified”. Consequently, selecting the objective measurements that are most relevant in professional audio, then reporting them consistently and honestly is a guiding principle that Genelec has always followed.

There is significant research that demonstrates the direct inter-relation between the objective and subjective aspects of reproduced sound quality, but there is also often a tendency to remain favourably disposed towards the one or another. Too often, one may encounter claims such as “it doesn’t matter what a measured flat frequency looks like, it only matters how it sounds”. Such a statement might sound appealing, and it’s not even incorrect because it is based not on the purpose of flat frequency or acoustical calibration, but rather an alternative interpretation of “truth”, and thus it is misleading.

The starting point of any high-end monitor design is the scientific foundation, and the neutrality and standards that foundation requires to deliver quality assurance. The end quality of reproduced sound is shaped by several factors, including the audio material, the monitor design, the room in which the audio is played and the monitor’s interaction with that space, as well as perceptual influences. But without a foundation of facts, it would be impossible to find any reference for the quality perception which is, by definition, a subjective concept.

Genelec’s education-oriented marketing therefore reflects our philosophy of providing truthful monitoring systems with respect for the human user. Indeed, the truthful reproduction of sound has been the cornerstone of the Genelec design philosophy since the origin of the company in 1978. Accordingly, we maintain a harmonious approach towards electro-acoustic design, while our incorporation of digital technology is intended to further this goal. For example, the development of Genelec Loudspeaker Manager (GLM) based systems now enables us to contend with far more challenging acoustic environments. In addition, it enables the reproduced sound to be customised in a predictable way, but always while maintaining a flat frequency response as its foundation.

The emotional engagement between a brand and its users is important and desirable, but we believe that the foundation of such an engagement must always be based on facts related to the products or services offered to the users. The adoption of a scientific foundation has been an essential ingredient in the development of human civilisation. Science can’t cover all the facts around us but so far it has provided the most reliable ground upon which we can build.

As Ilpo understood, good marketing is all about being able to tell a powerful story, and the best story of all is always the truth. That’s the best way to educate, inspire and create real value for users, while avoiding the trap of indifference.


Siamäk Naghian

Managing Director



ILPO MARTIKAINEN (1947-2017)
28616937_10155898180003673_4167117947706031091_o.jpg
 
Last edited:

Cosmik

Major Contributor
Joined
Apr 24, 2016
Messages
3,075
Likes
2,180
Location
UK
To me, the place where articles like this fall down is:
Consequently, selecting the objective measurements that are most relevant in professional audio, then reporting them consistently and honestly is a guiding principle that Genelec has always followed.
"Truth" only exists within the framework that has been "selected", and it can always be argued that the other truth-seeker has selected the wrong framework; the selection is subjective.
 
OP
svart-hvitt

svart-hvitt

Major Contributor
Joined
Aug 31, 2017
Messages
2,375
Likes
1,253
To me, the place where articles like this fall down is:

"Truth" only exists within the framework that has been "selected", and it can always be argued that the other truth-seeker has selected the wrong framework; the selection is subjective.

@Cosmik , I don’t understand your point so please elaborate.

I take it as a matter of course that a scientist needs to make priorities, make judgements. Of course, these decisions need to be data driven, but at the end of the day one needs to make decisions under uncertainty.

This article is, naturally, no research article. So it should be read with that in mind.

If the article generates interest you could go to Genelec’s web site where they have picked out examples of 38 years worth of academic papers written by the Genelec team over the years:

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

I wonder, which other audio companies can show the same dedication to research and science over decades and reflecting a team based approach to research.

I always wondered why audio forums are so little curious about the Genelec way. I think I have the answer but I will not post it just yet ;)
 

Jakob1863

Addicted to Fun and Learning
Joined
Jul 21, 2016
Messages
573
Likes
155
Location
Germany
To me, the place where articles like this fall down is:

"Truth" only exists within the framework that has been "selected", and it can always be argued that the other truth-seeker has selected the wrong framework; the selection is subjective.

Scientific work is often about modelling and by definition a model is a simplified version of reality so i´d think your critique would fit to a lot of work in nearly all fields?!
 
OP
svart-hvitt

svart-hvitt

Major Contributor
Joined
Aug 31, 2017
Messages
2,375
Likes
1,253
Scientific work is often about modelling and by definition a model is a simplified version of reality so i´d think your critique would fit to a lot of work in nearly all fields?!

Exactly. Maybe @Cosmik was lost in translation here? I don’t understand what his point exactly is.
 
Last edited:

Cosmik

Major Contributor
Joined
Apr 24, 2016
Messages
3,075
Likes
2,180
Location
UK
Exactly. Maybe @Cosmik was lost in translation here? I don’t understand what his point exactly is.
I just meant that you could arrive at many different destinations through the choice of which measurements you 'believe in'. They talk about selecting the measurements most relevant to professional audio, but how did they come to decide which were the most relevant? Their "truth" and "facts" are based just as much on subjective choices as anyone else's. They may still be right in their choices, but they can't claim the mantle of "truth" so easily.

As an example, my particular bugbear is the widespread assumption that in-room frequency response measurements tell you how something will sound. Is this "falsifiable"? In theory, maybe, but if most people believe in it without question, any 'science' becomes twisted towards demonstrating it rather than trying to falsify it. When mysterious results surface anyway, they try to find explanations and interpretations that square the circle. The initial assumption survives unscathed regardless of the 'science'.

Here is an article (by Amir in fact!) on some science that shows that humans focus on direct sound regardless of reflections
http://www.madronadigital.com/perceptual-effects-of-room-reflecti

I read this and it seems to say that if electronic comb filtering sounds bad, but reflections sound good, and yet a microphone & FFT can't tell the difference, then electronic 'correction' of reflections based on measurements with microphones and FFTs is not valid - the listener will just focus on the now arbitrarily-modified direct sound regardless of the flatness of the resulting FFT. But does this revelation from 1983 do anything to dampen people's enthusiasm for full range 'room correction'? Of course not! In their heads, the idea has just got to be right, and that is that. (I once assumed it to be self-evidently obvious, too).
 
OP
svart-hvitt

svart-hvitt

Major Contributor
Joined
Aug 31, 2017
Messages
2,375
Likes
1,253
@Cosmik , you are very right in observing that "you could arrive at many different destinations through the choice of which measurements you 'believe in' ". That's how science works its way towards a consensus. Sometimes, in complex issues, a consensus may never be reached. You have certainly heard about meta-studies, which is a way of establishing a consensus on an issue (a metastudy could be regarded as a way of utilizing the "wisdom of crowds", delphi decision making etc. in complex decision making). To my surprise, I often encounter people with higher academic degrees who are very skeptical about meta studies. So I think one's view on science and what science is may also decide one's view on metastudies. Am I right?

(In hifi, audio, it seems like hi-res is an issue that is best solved through metastudies. One such study (http://www.aes.org/e-lib/browse.cfm?elib=18296) found that there indeed may be something going in hi-res, but that conclusion is only reached after merging sets of big data, with all the practical implications such a big data study has).

In my view, it's quite natural that science driven firms will follow different paths. Different paths need not mean that the paths are decided by emotions, hunches, myth and so on. But if there is something like truth in a field of science, like audio, I would expect the paths to merge on many issues. And I believe we are witnessing this, as much as we are witnessing examples where forces try and push us off the narrow path of insight and wisdom, cfr. "hi-end", MQA etc.

When it comes to your view on room compensation, it seems like your "no correction" view may be a bias of yours? Toole told us he prefers (back up by research) compensation up to 500 Hz, using an averaging algorithm on the way up to the mid frequencies. Olive showed us that some room compensations systems yielded better results than other systems - in a test where "no correction" fared a bit worse than "average".

For sure, Genelec's algorithms may be a bit different than other researchers', but again based on thoughtful analyses where physics and psychoacoustics play important roles. Again, that's quite natural and what you would expect from research-driven companies in a relatively novel and quite complex field.

Lastly, it seems like you believe that room compensation systems merge direct sound and reflections? FWIW, I have been told that modern room compensation systems indeed differ between direct sound and reflections.

PS1: Your link to Amir's old article, which refers to a much older article, is interesting (and we have discussed the 1983 article previously), but it is only one voice in a field which may be more suitable to meta-analysis. I don't think a decades old article can be regarded as the final truth on the complex issue of psychoacoustics, perception of sound. And a researcher who claims to have the final truth on a complex issue, is just that, i.e. an individual of lesser complexity.

PS2: You didn't answer my question if you had ever witnessed an audio company that lists research made by its group of researchers over 4 decades?
 
Last edited:

Jakob1863

Addicted to Fun and Learning
Joined
Jul 21, 2016
Messages
573
Likes
155
Location
Germany
@Cosmic,

the line that an "article like that falls down" was puzzling me and still is, as your citation merely said Genelec´s approach is to use the scientific toolbox to select the most important measurement (most important for professional work, which might be something different than most important for listening at home) .

That scientists can be wrong in their decisions isn´t hardly surprising and that it is quite difficult to work with human perception isn´t surprising either. Maybe i´m missing the main point, but i still don´t get why this specific paper should "fall down" because of a (justified) criticism that fits to a large proportion of scientifc work.

Something could co wrong? Sure, but isn´t that always a given that something could go wrong?
 

Cosmik

Major Contributor
Joined
Apr 24, 2016
Messages
3,075
Likes
2,180
Location
UK
the line that an "article like that falls down" was puzzling me and still is
What I meant was that in amongst the pure reason and objectivity and the suggestion that the article contains the way to "the scientific truth", lies a fatal flaw: that all the supposed objectivity that flows from the scientific approach is derived from subjectivity i.e. the initial selection of which measurements are 'believed in'.
 
OP
svart-hvitt

svart-hvitt

Major Contributor
Joined
Aug 31, 2017
Messages
2,375
Likes
1,253
What I meant was that in amongst the pure reason and objectivity and the suggestion that the article contains the way to "the scientific truth", lies a fatal flaw: that all the supposed objectivity that flows from the scientific approach is derived from subjectivity i.e. the initial selection of which measurements are 'believed in'.

@Cosmik , you cannot evaluate a blog article aimed at a general audience as if it were a scientific article, can you?

For evaluation of the company’s academic traditions and merits, you should look at the examples of academic articles over 4 decades on the company’s web site. Wouldn’t you agree?

Having said that, I think the article by Siamäk was pretty well written, given the article format’s intentional lack of «academic form».
 

Jakob1863

Addicted to Fun and Learning
Joined
Jul 21, 2016
Messages
573
Likes
155
Location
Germany
What I meant was that in amongst the pure reason and objectivity and the suggestion that the article contains the way to "the scientific truth", lies a fatal flaw: that all the supposed objectivity that flows from the scientific approach is derived from subjectivity i.e. the initial selection of which measurements are 'believed in'.

Ok, but why do you assume that it is just "believed in" ?
If they use the scientific toolbox for this choice of "most relevant measurements" - and that is imo what the article expressed - it is just the normal way of doing scientific work with human perception.
I hope, you understand why that conclusion puzzled me.

Furthermore we know that every work on models (i´d say that every human scientific work) is subject to subjective decisions. You have to simplify and and you have to be confident that you were throwing out the right parts, so it always contains a bit of "believe" , doesn´t it?
 
OP
svart-hvitt

svart-hvitt

Major Contributor
Joined
Aug 31, 2017
Messages
2,375
Likes
1,253
Currently working with a pair of Genelecs. They sound great (subjectively):)
https://www.genelec.com/studio-monitors/sam-studio-monitors/8260a-sam-studio-monitor

@Guermantes , you said «working», like doing studio work?

I never heard the 8260. Specs indicate they’re even smoother than «the (new) ones». But they are bigger and I would expect them to sound bigger/fuller than «the ones». But the sweet spot of 8260 may be a bit small than the SS of «the ones»?
 

Cosmik

Major Contributor
Joined
Apr 24, 2016
Messages
3,075
Likes
2,180
Location
UK
@Cosmik , you cannot evaluate a blog article aimed at a general audience as if it were a scientific article, can you?

For evaluation of the company’s academic traditions and merits, you should look at the examples of academic articles over 4 decades on the company’s web site. Wouldn’t you agree?

Having said that, I think the article by Siamäk was pretty well written, given the article format’s intentional lack of «academic form».
I was addressing purely what the article was saying. I was not concerned with who had said it, or which company they work for, or what their history is :)
 

Guermantes

Senior Member
Joined
Feb 19, 2018
Messages
484
Likes
561
Location
Brisbane, Australia
@Guermantes , you said «working», like doing studio work?

I never heard the 8260. Specs indicate they’re even smoother than «the (new) ones». But they are bigger and I would expect them to sound bigger/fuller than «the ones». But the sweet spot of 8260 may be a bit small than the SS of «the ones»?

Yes, they are at work, in a room dedicated to audio analysis. Unfortunately I don't own them myself . . .

I haven't heard The Ones but the engineering seems smart and the reviews are good but I can't comment on difference in sweet spots. I am using the SAM room correction which seems to clean up problems below 100 Hz (resonances) reasonably well. I can't say I hear too much change in character above the (probable) Schroeder frequency with the correction on or off. I think in this instance the Genelec SAM equalisation is targeting room modes while a much broader brush is used for reflections.
 

Floyd Toole

Senior Member
Audio Luminary
Technical Expert
Industry Insider
Forum Donor
Joined
Mar 12, 2018
Messages
367
Likes
3,893
As Ilpo understood, good marketing is all about being able to tell a powerful story, and the best story of all is always the truth. That’s the best way to educate, inspire and create real value for users, while avoiding the trap of indifference

I met Ilpo sometime in the 80s at AES conventions. He had been following my NRCC research and papers and was eager to demonstrate to me how well he had applied the findings. Then, and in following years, when we met at AES conventions, he would proudly take to to his demo room and play classical music through his range of speakers, from small to large, demonstrating how similar, and good, they sounded. He had the courage to publish fundamentally accurate and useful anechoic data on his speakers - something almost nobody did or does. I liked him. He is a loss to the audio industry.

That said, I assume for business reasons, Genelec followed the guidance of ITU and EBU standards requiring flat steady state in-room curves, and they provide means to achieve it. This is demonstrably wrong, but Europeans have a commitment to "standards" that is substantially missing in our culture. Many broadcasters and recording studios were misguided by these standards. See Section 13.2.2 "The wrong room curve" in my new book. Summary: they require a flat on-axis anechoic response, and then, after installation, a flat in-room steady-state curve. One cannot have both, except in an anechoic room!

Nevertheless, for anyone willing to exercise a modicum of independent, logical, thought can use his products to generate fine sound.
 

pirad

Active Member
Joined
Apr 22, 2018
Messages
178
Likes
61
I am curious how they achieved the flat room curve.
Did they use some room compensation mechanism in their active speakers?
Or their anechoic curve was rising?
 
Last edited:

Floyd Toole

Senior Member
Audio Luminary
Technical Expert
Industry Insider
Forum Donor
Joined
Mar 12, 2018
Messages
367
Likes
3,893
It has been a while since I reviewed their product specifications, but the answer to your question will be there. If the loudspeaker has a flattish on-axis frequency response, which theirs tended to have, then in a normally reflective room the steady-state room curve will tilt downwards. This is explained in detail in my book. To make that flat means equalizing the speaker, and this will yield an on-axis, direct sound, response that is tilted upwards.

It is a problem with the standards, not the speakers, but for those slavishly attached to the standards, less than ideal timbral balance results. When this is done in the process of program creation it contributes to the circle of confusion problem. Of course it makes a difference if the speaker is used as a "near-field" monitor, dominated by direct sound at a short listening distance. Some control rooms are very dead acoustically which is not recommended for recreational listening. When such speakers are used in normally reflective domestic spaces the condition I describe will be most obvious.

BTW those standards are erroneous in other respects as well. The ITU loudspeaker & room recommendations (which the EBU document duplicates) were part of a standard created for evaluating record/replay codecs, not for neutral listening to movies and music. These recommendations should be in a stand-alone document intended for wider use. The ITU is in the process of reviewing their standard and I have been involved. Whether it will be changed is anyone's guess because it must pass an international vote, and it is already clear that not everyone is in agreement. It is embarrassing to admit that for decades part of the industry and a lot of academics have taken guidance that was not in accord with existing science.
 

NorthSky

Major Contributor
Joined
Feb 28, 2016
Messages
4,998
Likes
937
Location
Canada West Coast/Vancouver Island/Victoria area
OP
svart-hvitt

svart-hvitt

Major Contributor
Joined
Aug 31, 2017
Messages
2,375
Likes
1,253

And this (the articles above are from 2004) was prior to their launching of any DSP in their speakers. As far as I understand, Genelec started with DSP on an R&D basis in 2000. In 2006 they launched their DSP, called SAM for Smart Active Monitoring, in commercially available products.
 
Top Bottom