• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). Come here to have fun, be ready to be teased and not take online life too seriously. We now measure and review equipment for free! Click here for details.

An Enticing Marketing Story, Theory Without Measurement?

Joined
Jul 10, 2019
Messages
15
Likes
16
Would the same principle apply to, say, a solo performer?
Imagine a band playing outside. The bass player adjusts the volume of his bass amp to match the loudness of the singer. The overall sound is great.

Now the band moves into a room. The singer sings as loud as before. The bass amp has the same volume adjustment but the bass sounds boomy and to loud because of standing waves in the room.
So what is the natural behaviour of the talented bass player? He will turn down the volume of the amp. He does not wear headphones instead or sit closer to his bass amp and he even does not change the room to an anechoic chamber :) He simply turns down the playback level of the amp and the bass fits to the singer.

Can you imagine this? If yes, you learn that it is possible to improve the sound in a room by adjusting the speaker (here via the bass amp gain).
 

Cosmik

Major Contributor
Joined
Apr 24, 2016
Messages
2,968
Likes
1,800
Location
UK
Imagine a band playing outside. The bass player adjusts the volume of his bass amp to match the loudness of the singer. The overall sound is great.

Now the band moves into a room. The singer sings as loud as before. The bass amp has the same volume adjustment but the bass sounds boomy and to loud because of standing waves in the room.
So what is the natural behaviour of the talented bass player? He will turn down the volume of the amp. He does not wear headphones instead or sit closer to his bass amp and he even does not change the room to an anechoic chamber :) He simply turns down the playback level of the amp and the bass fits to the singer.

Can you imagine this? If yes, you learn that it is possible to improve the sound in a room by adjusting the speaker (here via the bass amp gain).
The difference being that the bass player has access to the individual 'feed'. The would-be room corrector only has access to the composite signal. Any single dimensional EQ adjustment that doesn't tally with the other dimensions may (will) sound false.
 
Joined
Jul 10, 2019
Messages
15
Likes
16
Any single dimensional EQ adjustment that doesn't tally with the other dimensions may (will) sound false.
You are not forced to listen to "corrected" systems.

Any conductor knows that the orchestra has to adopt to the hall they are playing in. With new halls this can even take some time. But they do it.

Of course we cannot change the mix during playback like a mastering engineer. But we can apply an overall filtering if there are benefits. If there is no benefit for you you can simply leave everything untouched. By the same logic I can also deduct that I do not need for any room treatments. But if someone e.g. starts to move a speaker to find a good position he starts to adopt to the room sound despite this is a single dimensional operation too. Is this truly wrong?
 

Cosmik

Major Contributor
Joined
Apr 24, 2016
Messages
2,968
Likes
1,800
Location
UK
You are not forced to listen to "corrected" systems.

Any conductor knows that the orchestra has to adopt to the hall they are playing in. With new halls this can even take some time. But they do it.

Of course we cannot change the mix during playback like a mastering engineer. But we can apply an overall filtering if there are benefits. If there is no benefit for you you can simply leave everything untouched. By the same logic I can also deduct that I do not need for any room treatments. But if someone e.g. starts to move a speaker to find a good position he starts to adopt to the room sound despite this is a single dimensional operation too. Is this truly wrong?
It's a free country for the moment, so anyone can do as they like. Although some people are doing what they think they should, having been advised badly. It may contaminate the way recordings are made, affecting my listening pleasure :).

Question: have you heard of the idea of blind deconvolution? Do you think humans may do something similar? 'How much' of it do you think they do? And when do they do it, and when do they not? Might those answers not be essential before you even start thinking about room correction?
 
Joined
Mar 27, 2019
Messages
24
Likes
19
Location
Greece
A 'target curve' makes no sense - it is the tail wagging the dog.
I would like to see DBT studies that prove your arguments.
Tests similar like the ones that Sean Olive has presented, with hundreds of listeners stating their preference for various loudspeakers, house curves etc.

I am not aware of any DBT study proving that in a modern living space and using well designed speakers, listeners will prefer unequalised sound for listening to music reproduction, especially under 250-300Hz.

And furthermore, I would like to see a DBT of preference comparison of various Eq software, including non Eq. Like the small study done by S.Olive back in 2009 using ancient software versions. Even back then, the study showed listeners preference for the Eq.

F389FDAE-C19D-4894-BAE8-9DF8F646A4B9.png

http://seanolive.blogspot.com/2009/11/subjective-and-objective-evaluation-of.html

Until proven otherwise by DBT studies, preferably larger, I can’t accept your arguments. You are presenting an “expert opinion” which has very low level of evidence. And lots of Bias.

8E39EDF0-0E8C-4E69-A64E-6CE600993ABF.jpeg
 

Cosmik

Major Contributor
Joined
Apr 24, 2016
Messages
2,968
Likes
1,800
Location
UK
I would like to see DBT studies that prove your arguments.
Tests similar like the ones that Sean Olive has presented, with hundreds of listeners stating their preference for various loudspeakers, house curves etc.
But that doesn't disprove that the tail is wagging the dog: merely that in these cases the dog just about met up with the tail.

It can't be understood by experiment; only logic.

Experiment may seem to validate a faulty hypothesis if the hypothesis is, itself, just a regurgitation of an observation of an experiment, and the future experiments more-or-less duplicate that experiment! And this 'target curve' thing is exactly that. Highly intelligent people stating without any further thought or explanation that it is better to reproduce audio signals with a drooping frequency response! Does that not seem 'odd'? Is it not more likely that the listener is hearing the frequency response of the speaker, not the speaker and room combined, especially when you find, by calculation, that a neutral speaker in a room like the one they're listening in would produce that drooping frequency response? But a speaker with different dispersion characteristics, or in a different room, wouldn't.

The extra complication is that most speakers aren't neutral in terms of dispersion. So the results of this target curve thing are just a crapshoot: if you happen to have an average speaker in an average room, then maybe the result will be about right: it will sound OK and seem to give you your target curve (and if you were to test the speaker in an anechoic chamber you would find that it was set up 'sensibly' for that room). But if you have the wrong speaker and the wrong room and shoehorn in the adjustments to get the perfect combined target curve you will be mystified that it sounds rubbish.

And you will wonder why speakers in the 1970s sounded so much better. It must be your imagination of course....
 
Last edited:

Cosmik

Major Contributor
Joined
Apr 24, 2016
Messages
2,968
Likes
1,800
Location
UK
No. You have to prove your arguments by DBT. Always.
So you would actually conduct double blind tests in order to argue against this?!
https://www.binauralbeatsmeditation.com/432-hz-truth-behind-natures-frequency/
https://www.audiosciencereview.com/forum/index.php?threads/schiit-gadget-music-tuning-box.1980/
According to product designer, Schiit’s Number 2 Mike Moffat, music’s tuning pitch has slowly crept upwards over time. The Gadget corrects for this drift with digital signal processing (DSP). Changes can be applied in real time with the twist of a knob with users finding their own unique “aah” point – one where, according to Moffat, music will sound more fluid and more immersive.
Now, I'm quite prepared to believe that it might be possible that I and everyone else here really does like music at lower pitch - for a while, as a novelty.

But if someone argues instead that it is simply stupid, I am prepared to accept that argument without going to the lengths of conducting a DBT - that might actually appear to show that I would prefer my music tuned to 432Hz for all I know. Nor do I want my amp to have a built-in phaser even though I love the sound of a phaser. And I don't want it to have spatialiser effects or a 'phat' setting even though I might like it for an hour or two.
 
Joined
Mar 27, 2019
Messages
24
Likes
19
Location
Greece
So you would actually conduct double blind tests in order to argue against this?!
https://www.binauralbeatsmeditation.com/432-hz-truth-behind-natures-frequency/
https://www.audiosciencereview.com/forum/index.php?threads/schiit-gadget-music-tuning-box.1980/

Now, I'm quite prepared to believe that it might be possible that I and everyone else here really does like music at lower pitch - for a while, as a novelty.

But if someone argues instead that it is simply stupid, I am prepared to accept that argument without going to the lengths of conducting a DBT - that might actually appear to show that I would prefer my music tuned to 432Hz for all I know. Nor do I want my amp to have a built-in phaser even though I love the sound of a phaser. And I don't want it to have spatialiser effects or a 'phat' setting even though I might like it for an hour or two.
I am not convinced.
Try again.
 

Cosmik

Major Contributor
Joined
Apr 24, 2016
Messages
2,968
Likes
1,800
Location
UK
I am not convinced.
Try again.
I think it's your turn to make the case for room correction.

But let me ask you a question. Have you ever made a tape recording of a live concert from a distance, but then noticed that when you listen to it at home it sounds very different: 'hollow', 'echoey'. If so, why do you think that is? Does the resemblance between a tape recorder and a mic & laptop set off any alarm bells?
 

Blumlein 88

Major Contributor
Joined
Feb 23, 2016
Messages
6,497
Likes
6,229

Cosmik

Major Contributor
Joined
Apr 24, 2016
Messages
2,968
Likes
1,800
Location
UK
Ι have already quoted the relevant studies. You haven’t. Just philosophical views
Yes, but logic tells us that an experiment can be based on a faulty hypothesis and still come up with the right answer*. The results and hypothesis may fail to be successfully extrapolated to all speakers and all rooms, and the others where it seems to 'work' are probably not optimal and could be better.

What should an amplifier do? What should a DAC do? I don't think experiments are needed to show what these should do. I am extending the same logic to saying what a speaker should do. You may call it philosophy.

Do you need to do an experiment to show that an audio system shouldn't shift the pitch of the signal? It is logical that it should not. What will you do if a DBT finds that it should? *Logic* tells you that it may just be the novelty that has caused people to prefer it temporarily, but you would still plough ahead and build it into your DSP system..? That would be cutting your nose off.
____________________________________________________________________
* Neutral speaker in real room gives drooping in-room response, but the listener (being led astray by evolution and philosophy) is hearing the direct sound of the speaker. A non-neutral speaker (dispersion-wise) in a real room needs some baffle step type compensation to sound OK, giving a slightly different in-room response. A correctly set up 'average' speaker in an an average room will give an average drooping response.

Experiment mistakenly assumes that the in-room response alone is what the listener is hearing and duplicates said in-room response using average speaker in average room. Bingo! The listener prefers it! But it was never the in-room response alone that the listener was hearing; it was (ideally) a neutral speaker or, if neutral not available, an average speaker with some EQ compensation for its dispersion deficiencies.

The experiment merely reproduced an 'average' setup. Attempts to duplicate the experiment with other speakers and other rooms would be shoehorning a target in-room response into a system that should have a different in-room response. It may not be miles out, but it's still wrong, and could be set up so much better by doing it 'philosophically'.
 
Last edited:
Joined
Jul 10, 2019
Messages
15
Likes
16
Let's assume there is a neutral speaker in a well done setup. Of course you are happy to listen to it.
Now move the neutral speaker close to front wall. Listening to the result will result in more bass. But the neutral speaker has not been changed it is still "neutral". Obviously at least one of 27 dimensions has changed to our disadvantage.

Now let's further assume that for some weird reason we are not allowed to move the speakers back.
So what can we do?
Pure logic tells us that exchanging the neutral speaker is not a solution as then we may hurt the neutrality?
Changing the room to an anechoic chamber also does not seem very logical beside the effort it will take.

But if now the amplifer is equipped by a bass control the average critical listener will turn down the bass until the result is pleasing to him. He may not seriously think about that the bass control changes the content of the music track. He will not think about the 26 remaining dimensions. He simply turns down the bass and is way more happy (listeners infected by audiophilia nervosa should be excluded here).

By this experiment there is no microphone in game. No 'room correction' software. Just the listener who does a correction by himself within his 27 dimensions of hearing.
BUT:
A correction takes place in fact. There are two mandatory prerequisites for a correction: 1. you have to know the current state and this requires a measurement, 2. you need to have a target. Without the two conditions you cannot correct.
By the given example 1. the listener uses his ears instead of the microphone 'to measure' and 2. stops turning the bass control when the sound matches his imaginations = personal target.
The final result is now by principle: the listener applies a correction for deficiencies caused by moving the neutral speakers close to the front wall. If you call it deficiencies for speaker-room-setup or simplify it to 'correction for room deficiencies' (remember: the neutral speaker has not changed) then indeed you can now speak of 'room EQ' or 'room correction'.

Please think about: there is no microphone in the game, no software, no discussion about drooping frequency responses (btw. a good room correction does not change the drooping or it allows to tweak the drooping until the final correction system = listener with ears and target in mind is happy).

Now experience tells us that adjusting and optimizing 27 dimensions just by listening with ears is a challenge where the average listener fails. Only a few gifted talents may be sucessful. I do not belong to the latter group. Thus I prefer to combine both ears and microphones and a well-thought room correction in a peaceful coexistence.
 
Joined
Mar 27, 2019
Messages
24
Likes
19
Location
Greece
@Cosmik

In my line of work, if my decisions are taken on “logic” and not hard-proven evidence, people get crippled or dead, and I end up in prison.
This should be the standard.

You have to properly design a DB study, with the variables that your hypothesis require, and prove that listeners prefer non-roomEq music reproduction.

Till then, you are just being philosophical, not scientific.

832DEAF1-1A86-4DE3-801B-7F761577A2AB.jpeg
 

Cosmik

Major Contributor
Joined
Apr 24, 2016
Messages
2,968
Likes
1,800
Location
UK
Let's assume there is a neutral speaker in a well done setup. Of course you are happy to listen to it.
Now move the neutral speaker close to front wall. Listening to the result will result in more bass. But the neutral speaker has not been changed it is still "neutral". Obviously at least one of 27 dimensions has changed to our disadvantage.

Now let's further assume that for some weird reason we are not allowed to move the speakers back.
So what can we do?
Pure logic tells us that exchanging the neutral speaker is not a solution as then we may hurt the neutrality?
Changing the room to an anechoic chamber also does not seem very logical beside the effort it will take.

But if now the amplifer is equipped by a bass control the average critical listener will turn down the bass until the result is pleasing to him. He may not seriously think about that the bass control changes the content of the music track. He will not think about the 26 remaining dimensions. He simply turns down the bass and is way more happy (listeners infected by audiophilia nervosa should be excluded here).

By this experiment there is no microphone in game. No 'room correction' software. Just the listener who does a correction by himself within his 27 dimensions of hearing.
BUT:
A correction takes place in fact. There are two mandatory prerequisites for a correction: 1. you have to know the current state and this requires a measurement, 2. you need to have a target. Without the two conditions you cannot correct.
By the given example 1. the listener uses his ears instead of the microphone 'to measure' and 2. stops turning the bass control when the sound matches his imaginations = personal target.
The final result is now by principle: the listener applies a correction for deficiencies caused by moving the neutral speakers close to the front wall. If you call it deficiencies for speaker-room-setup or simplify it to 'correction for room deficiencies' (remember: the neutral speaker has not changed) then indeed you can now speak of 'room EQ' or 'room correction'.

Please think about: there is no microphone in the game, no software, no discussion about drooping frequency responses (btw. a good room correction does not change the drooping or it allows to tweak the drooping until the final correction system = listener with ears and target in mind is happy).

Now experience tells us that adjusting and optimizing 27 dimensions just by listening with ears is a challenge where the average listener fails. Only a few gifted talents may be sucessful. I do not belong to the latter group. Thus I prefer to combine both ears and microphones and a well-thought room correction in a peaceful coexistence.
For sure, we may need to compensate for a system's defects, and this should be by ear - certainly I adjusted the depth of the baffle step compensation for my speakers by ear. It isn't a correction, merely a better compromise. And your example listener may well feel the need to change the bass later as the slightly unnatural sound begins to annoy him.

Take your example further: start at random EQ settings and give a listener the ability to adjust it to his liking by saying "Alexa 300Hz, up 2dB" without any knowledge of the current settings and see where it gets to!

Your example still implies that the listener enjoys the sound of a stream of "frequency response matter" and is completely ambivalent on the issue of modification of transients in the time domain, for example (another one of the 27 dimensions). As though music is flavoured paste and its texture is unimportant.

In reality, a source and room all hangs together, and the room's effect on the frequency response (which looks different depending how you measure it, of course) also results in highly-specific 1:1 corresponding effects on transients in the time domain. It cannot be 'played with' arbitrarily otherwise it no longer hangs together and the listener will know.

Another analogy: do some experiments to find the optimum height for a table using rigorous double-blind testing methods. It comes out as 1.2m. Then look at your own table. Oh no, it's 1.3m. One of the legs is very easy to get to without moving the table, so you decide to saw it down to 1.2m. It isn't perfect, but it is a partial correction, proven by science.
 
Joined
Jul 10, 2019
Messages
15
Likes
16
In reality, a source and room all hangs together
Great news, you seem to start to agree that a neutral speaker in an arbitrary room does not guarantee a perfect listening result :)

It cannot be 'played with' arbitrarily otherwise it no longer hangs together and the listener will know.
Is there anyone out there who says that playing arbitrarily playing with room correction is the proper way? Of course it makes sense to seriously think about the nature of sound perception.

At the end it is always the same. As long as a playback system does not satisfiy the listener will start to change it. He will notice the "fault" by measurement (by ears, by microphone, by ruler [setup of speaker position], by voltmeter ...), he will define his own personal target and apply a correction (tuning, equalizing, speaker exchange, room treatment) until the result meets his expectations. It does not even matter if he does this in an arbitrary way until he reaches the desired result. Though the probability is not very high in this case.

You are talking about a neutral speaker: when is a speaker neutral and how do you EXACTLY know it is? Wouldn't it make sense to use a "neutral" measurement device (despite this introduces the discussion about calibration)? Your ears are not mine. So I expect differences in the definition of neutral, ears and brains are pretty difficult to calibrate. You can only tell that a speaker is neutral (your target) when you measure it (e.g. by your ears) and if it not you can modify = correct it (in a scientific way or arbitrarily) until the speaker matches your target. So what's the difference between adjusting a speaker for neutrality and correcting for room deficiencies except the number of variables?
 

Cosmik

Major Contributor
Joined
Apr 24, 2016
Messages
2,968
Likes
1,800
Location
UK
Great news, you seem to start to agree that a neutral speaker in an arbitrary room does not guarantee a perfect listening result :)
Obviously a ridiculous room will give a ridiculous result. But this cannot be fixed by changing the source. If a room is "fine" then nothing needs changing.
Is there anyone out there who says that playing arbitrarily playing with room correction is the proper way?
Of course they don't say it, but they do it anyway.
...he will define his own personal target and apply a correction (tuning, equalizing, speaker exchange, room treatment) until the result meets his expectations. It does not even matter if he does this in an arbitrary way until he reaches the desired result. Though the probability is not very high in this case.
The point about the neutral speaker in the ordinary room is that it is all 27 variables in perfect alignment. If it doesn't please the discerning listener, it must be because of the room and its layout. As you note, playing with stuff randomly applying correction to achieve a personal target is age-of-the-universe stuff.
You are talking about a neutral speaker: when is a speaker neutral and how do you EXACTLY know it is? Wouldn't it make sense to use a "neutral" measurement device (despite this introduces the discussion about calibration)? Your ears are not mine. So I expect differences in the definition of neutral, ears and brains are pretty difficult to calibrate. You can only tell that a speaker is neutral (your target) when you measure it (e.g. by your ears) and if it not you can modify = correct it (in a scientific way or arbitrarily) until the speaker matches your target. So what's the difference between adjusting a speaker for neutrality and correcting for room deficiencies except the number of variables?
I think you need to reduce the number of variables as much as possible. A speaker close to neutral would give you the easiest situation:
No voicing required. Other loudspeakers usually require voicing. Based on listening to a lot of recordings, the tonal balance of the loudspeaker is changed so that most recordings sound good. Voicing is required to balance differences between direct and off-axis sound. The 8c has very even dispersion. It is the first loudspeaker I ever designed that did not benefit from voicing. The tonal balance is purely based on anechoic measurements.
For the rest of us, a two-way speaker with a huge discrepancy between dispersion of drivers at the crossover is creating problems straight away. So use a three-way. The baffle step frequency can be calculated or modelled and the curve applied with only the depth requiring setting in situ. That kind of thing. It will sound great. No in-room measurements needed.
 
Joined
Jul 10, 2019
Messages
15
Likes
16
The point about the neutral speaker ...
Can you please give me a scientifically or mathematically EXACT definition of a neutral speaker? And how do you determine that a real speaker is neutral? And if you talk about
then how much 'close' is necessary to allow talking about neutral?

BTW the 6moons report you are linking to shows up a frequency response chart (without your required drooping). Is that allowed? Did they measure the activity of the listeners ears? And they use a DSP, oh my god.
6moons said:
Farther at the back of our 14 meter long room where our dining table sits, the sound still wanted a little extra. With the app at hand, we dialed the bass level up by just 3dB. Wow, that was all the sound needed.
[sarcasm on]Clearly they hurt all scientific rules about the inviolability of 27 dimensions[sarcasm off] :)
 

Cosmik

Major Contributor
Joined
Apr 24, 2016
Messages
2,968
Likes
1,800
Location
UK
Can you please give me a scientifically or mathematically EXACT definition of a neutral speaker? And how do you determine that a real speaker is neutral? And if you talk about

then how much 'close' is necessary to allow talking about neutral?

BTW the 6moons report you are linking to shows up a frequency response chart (without your required drooping). Is that allowed? Did they measure the activity of the listeners ears? And they use a DSP, oh my god.

[sarcasm on]Clearly they hurt all scientific rules about the inviolability of 27 dimensions[sarcasm off] :)
I'm not claiming that you can have a perfect system, but I think that the person who plans ahead can get it down to just one or two variables they adjust by ear in situ. With my speakers it was literally one variable that I adjusted by ear. The rest was automatic based on the anechoic measurements. I have no doubt that they might be optimised further through really complex Spin-o-Rama measurements, or even better, modelling, but I think they are pretty good anyway.

I can do a demonstration that suggests they're not too bad: I can radically reconfigure the crossovers in real time while playing music with virtually no change in the sound - in fact I can't hear it even when listening for it. But if we listen to a single driver and do the same thing, we are left in no doubt that the crossovers have changed drastically e.g. 250Hz and 5kHz changed to 500Hz and 2.5 kHz or any such permutations, slopes going from 2nd order to 8th; that kind of thing.

To be neutral, the speaker would require a flat frequency response (and phase, and time alignment) at all points in an anechoic chamber. The choice of angle of dispersion (uniform at all frequencies) would be one of those variables that the designer/listener would have to choose, as explained here:
https://www.grimmaudio.com/site/assets/files/1088/speakers.pdf

And the Grimm people also address the baffle step frequency by the power of thought alone rather than DBTs. I accept their arguments, anyway.
 
Top Bottom