Shinobiwan, Thanks for cutting to the nut of the thing! At the end of the day, all technical considerations aside, it,s a pretty good sounding speaker! Iv'e heard far worse at the Guru's price point and this was in a hotel room which is far from an ideal venu for showing off a speaker. Then again, given the Guru's design specifics, a hotel room with it's low celing and small size might just be ideal for the Guru
I´ve heard IngOehman´s speakers a few times and i agree with others who have positive experiences. In particular the bass and treble drivers appeared extremely well "sewn together"
They dont need to be marketed in scientifical manners with scientifical claims that isn´t possible to get verified in usual ways🙂 .
But thats maybe what everyone else in this business is doing also?
They dont need to be marketed in scientifical manners with scientifical claims that isn´t possible to get verified in usual ways🙂 .
But thats maybe what everyone else in this business is doing also?
Please note that all have ever done is ask for more precise information. I have said nothing against these speakers except to try to get clarification of details. (As Ingvar has said, some of the 'extra" explanations haven't helped the issue ;-)
If I was to hear some and really like them, that situation would not change, I would still seek better explanations of how the result was achieved !!!
This is a DIY site, remember.
If I was to hear some and really like them, that situation would not change, I would still seek better explanations of how the result was achieved !!!
This is a DIY site, remember.
ShinOBIWAN said:
I very much appriciated your above posting ShinOBIWAN. You are
a generous person, obviously. Respect.
I hope there will be an opportunity in the future, for me to prove
your friendly speculations to be correct. 😉
Best regards, Ingvar.
patrikf said:
I'm happy to hear that's your experience.
Well of coarse not. That's why they aren't.
Everything claimed can be verified by anyone who wants to do it. You can set up a study aimed at falsifying my claims at any time you like. I welcome it. And as I’ve told you earlier, that has already been done by others. Instead of thus falsifying my findings, the outcome of the studies is that my findings have been confirmed.
(But that does not necessarily mean that my conclusions on what is the best set of measures are correct, nor the only correct conclusions (nor conclusions relevant for every situation). Actually, there are many possible alternative measures that can be used to improve the fidelity, that does address the same stereo system colorations, but with different draw backs and benefits. Carlssons approach is one of them. Even though Stig Carlsson found his approach primarily by listening - the OA-world make good sense viewed through my findings. Research can lead to knowledge, leading to a solution that contains a lot of choices - but it can also lead to the understanding of choices already made…)
I’ll help explaining if you (or anyone else) do not want to research the claims yourself, of contemplate them thus trying them logically. It is quite difficult however, to verify something to a person who wants to have everything served by ordering it, and still reserves the right to say "I don't believe you" or "there is no studies" at any time - instead of asking questions regarding things unclear.
And sometimes, you might even have to accept the possibility that a positive impression from speakers you have listened to might actually be more than a fluke. Maybe, it is the consequence of design choices that is based on studies unpublished…
Why the “also”?
I don’t, and I've seen behaviours in that direction from others, only very seldomly. I believe that most serious loudspeaker manufacturers do their best to tell the truth, as they see and understand it. That is my experience.
Vh, iö
- - - - -
PS. Thank you for the change in attitude, I really appreciate it
I do apologize for my previous comments. Ingwar comes across actually as a nice guy and he’s taking time to express he’s opinion on some nut job forum. Most of the commercial designers would turn around and tell us to go f.. ourselves. And I know a couple. Granted that he’s trying to build up a company somewhat and we aren’t exactly costumers here.
The only suggestion that I have for Ing is to make his speakers look good. Although the shape is probably a necessity of the design, most buyers will ignore it and go to a more appealing speaker cabinet. From the business point of view, it shouldn’t cost much more.
Good luck.
The only suggestion that I have for Ing is to make his speakers look good. Although the shape is probably a necessity of the design, most buyers will ignore it and go to a more appealing speaker cabinet. From the business point of view, it shouldn’t cost much more.
Good luck.
R-Carpenter,
You my have a point, but....
I have seen these speakers first in, say 1980 (the Ino version). They have not changed a bit in appearence since then. Their appearance does not bother me any more, quite the contrary. I have seen multible dusins of shapes and forms over the years from other manufacurers, with no link to sound quality. Shape will degrade or upgrade sound quality, according to laws of physics. Why should the best looking speaker also perform the best, if their size and shape is in war with the laws of physics regarding sound dispersion? It is quite clear to me at least, that speaker design is as much importing as how they perform to the awerage manufacturer. The Öhman approuch is somewhat different. The laws of fysics does not change from 1980 to 2008. I think as a qoncequence of that, they look how they look. They are exacly the same. As a concequens of that, why should the appearance of the speaker change, if not to increase sales to people who does not care so much about the performance as they care about their ability to blend into the other furniture, including wives?
Construction and upgrading budgets are benefiting from the lack of need to spend the same money in more or less eye-catching design, which in this case only will degrade the sound compared to what the manufacters goals are , (IÖ`goal in this case). The marked is huge, and there are 13 to the dozen who has a approach on looks rather than performance, so people will have plenty of places elsewhere to look.
You my have a point, but....
I have seen these speakers first in, say 1980 (the Ino version). They have not changed a bit in appearence since then. Their appearance does not bother me any more, quite the contrary. I have seen multible dusins of shapes and forms over the years from other manufacurers, with no link to sound quality. Shape will degrade or upgrade sound quality, according to laws of physics. Why should the best looking speaker also perform the best, if their size and shape is in war with the laws of physics regarding sound dispersion? It is quite clear to me at least, that speaker design is as much importing as how they perform to the awerage manufacturer. The Öhman approuch is somewhat different. The laws of fysics does not change from 1980 to 2008. I think as a qoncequence of that, they look how they look. They are exacly the same. As a concequens of that, why should the appearance of the speaker change, if not to increase sales to people who does not care so much about the performance as they care about their ability to blend into the other furniture, including wives?
Construction and upgrading budgets are benefiting from the lack of need to spend the same money in more or less eye-catching design, which in this case only will degrade the sound compared to what the manufacters goals are , (IÖ`goal in this case). The marked is huge, and there are 13 to the dozen who has a approach on looks rather than performance, so people will have plenty of places elsewhere to look.
How fine you feelings about my attitude has changed IngOehman!
Its no point in evaluate claims about human hearing by trying if some statements are internally logical. It not maths we are talking about. Rather would it be interesting if you could explain your proposals in the context of current established research about hearing.
I´ve asked such questions earlier in this thread🙂 .
Its no point in evaluate claims about human hearing by trying if some statements are internally logical. It not maths we are talking about. Rather would it be interesting if you could explain your proposals in the context of current established research about hearing.
I´ve asked such questions earlier in this thread🙂 .
That reply really surprises me, Patrik. 
Should I interpret what you just wrote, that you are not clear over the connection between what I’m doing and HRTF?
Darn, I was going to say that!
You're to quick, Pan. 😉
Anyway, although my findings are slightly different from what I have seen published, I have no general objection regarding what is commonly accepted as the HRTF's of humans. The differences are not within the margin of error, but pretty close to it. It can all be explained by the fact that different peoples pinnas are studied every time attempts have been made to generally map the effects of the pinna. No two mappings will be identical, but they will be fairly similar. And for the purpose I'm using the information, I can not use aspects of it, that is to individual anyway. Only what I call "common codes" are usable.
But more important: The studies I’ve done have had a slightly different purpose than the typical (to get general knowledge on human hearing).
My focus has been on extracting knowledge regarding what can be done with acoustic coding/decoding techniques. And I'm not talking about MPEG now, but information reduction by coding to a finite number of channels, and decoding the same finite number of channels as accurate as possible* - i.e. I've been focusing on researching the angles of incidence that are of special interest for understanding the way that the stereo systems (there are many of them) are causing timbral and geometrical colorations.
Outside the "stereo window" - let’s say it’s 66 degrees (+/- 33, i.e 10 degrees outside the speakers) horizontally, and 37 degrees (+33/-4) vertically (but the exact size is not important) I have not gathered data with a higher resolution than what is commonly done (5, 10 or 22.5 degree intervals), but inside the window, I have been looking at the pinna#-effects with a much higher resolution.
However - one of the "verifying studies" that has been done has used commonly known HRTF's and did come to the conclusion that my findings are correct. So the matter of resolution in the analyses might not be of such a great importance, regarding the principles them selves, only regarding the precision of the compensation.
Best regards, Ingvar
- - - - -
*The ambition on what to accomplish, will give the limit for the precision obtainable. The ambition can differ in both size and character of the soundstage.
Is “only horizontal info within a +/- 22.5 degree opening” sufficient?
Or a larger opening?
Or is there an ambition to encode and decode depth also?
Maybe you want to be able to code both width and depth, and height as well?
If it is your opinion that not even a 3D-coding is sufficient, you might want to code and decode holophonically?
Or perhaps you do not think it is ok to have ANY restriction (though very large) to a window that opens up to the recording world? You may only accept an omni-opening.
Everything in the above is doable, but if only 2 channels is available for doing it, my experience is, that the precision (or other aspects of the reproduction) will suffer when attempting to go over the above mentioned 66/37 degree holophonic stereo window (which is a VERY large window as it is!). It does not make it a bad decision to try, but it is a subjective question. I can only defend my own choices in these matters.
But I can say this: My choices are done without any concern to old standards, nor to compatibility to movie theatres or home cinemas. But they coincide anyway! 🙂
Studies on practical listening situations in peoples homes (and choice of listening spots in recording studio control rooms by producers) has pointed at 23 degrees. Research done by NHT has suggested an optimum angle between the loudspeakers in the vicinity of +/- 21 degrees. My own studies regarding optimum placement of loudspeakers at centre front row (based both on HTRTF and subjective preferences) points at +/- 23 degrees. The film industry has a 45 degree (+/- 22.5 degree) listening angles for “best seats”.
The only “black sheep” is the old ITU-R BS.775, referring to an even older 2-channel standard, stating +/- 30 degree for the L and R speakers. But who put the speakers like that? Probably no one in their right mind (and with the ears intact).
#Actually, it’s wrong to call it pinna-effects, and also wrong (though better) to call it HRTF. The correct term would be HTRTF.
Let’s see who'll be the first to interpret it correctly… 😉
Should I interpret what you just wrote, that you are not clear over the connection between what I’m doing and HRTF?
Pan said:
Darn, I was going to say that!
You're to quick, Pan. 😉
Anyway, although my findings are slightly different from what I have seen published, I have no general objection regarding what is commonly accepted as the HRTF's of humans. The differences are not within the margin of error, but pretty close to it. It can all be explained by the fact that different peoples pinnas are studied every time attempts have been made to generally map the effects of the pinna. No two mappings will be identical, but they will be fairly similar. And for the purpose I'm using the information, I can not use aspects of it, that is to individual anyway. Only what I call "common codes" are usable.
But more important: The studies I’ve done have had a slightly different purpose than the typical (to get general knowledge on human hearing).
My focus has been on extracting knowledge regarding what can be done with acoustic coding/decoding techniques. And I'm not talking about MPEG now, but information reduction by coding to a finite number of channels, and decoding the same finite number of channels as accurate as possible* - i.e. I've been focusing on researching the angles of incidence that are of special interest for understanding the way that the stereo systems (there are many of them) are causing timbral and geometrical colorations.
Outside the "stereo window" - let’s say it’s 66 degrees (+/- 33, i.e 10 degrees outside the speakers) horizontally, and 37 degrees (+33/-4) vertically (but the exact size is not important) I have not gathered data with a higher resolution than what is commonly done (5, 10 or 22.5 degree intervals), but inside the window, I have been looking at the pinna#-effects with a much higher resolution.
However - one of the "verifying studies" that has been done has used commonly known HRTF's and did come to the conclusion that my findings are correct. So the matter of resolution in the analyses might not be of such a great importance, regarding the principles them selves, only regarding the precision of the compensation.
Best regards, Ingvar
- - - - -
*The ambition on what to accomplish, will give the limit for the precision obtainable. The ambition can differ in both size and character of the soundstage.
Is “only horizontal info within a +/- 22.5 degree opening” sufficient?
Or a larger opening?
Or is there an ambition to encode and decode depth also?
Maybe you want to be able to code both width and depth, and height as well?
If it is your opinion that not even a 3D-coding is sufficient, you might want to code and decode holophonically?
Or perhaps you do not think it is ok to have ANY restriction (though very large) to a window that opens up to the recording world? You may only accept an omni-opening.
Everything in the above is doable, but if only 2 channels is available for doing it, my experience is, that the precision (or other aspects of the reproduction) will suffer when attempting to go over the above mentioned 66/37 degree holophonic stereo window (which is a VERY large window as it is!). It does not make it a bad decision to try, but it is a subjective question. I can only defend my own choices in these matters.
But I can say this: My choices are done without any concern to old standards, nor to compatibility to movie theatres or home cinemas. But they coincide anyway! 🙂
Studies on practical listening situations in peoples homes (and choice of listening spots in recording studio control rooms by producers) has pointed at 23 degrees. Research done by NHT has suggested an optimum angle between the loudspeakers in the vicinity of +/- 21 degrees. My own studies regarding optimum placement of loudspeakers at centre front row (based both on HTRTF and subjective preferences) points at +/- 23 degrees. The film industry has a 45 degree (+/- 22.5 degree) listening angles for “best seats”.
The only “black sheep” is the old ITU-R BS.775, referring to an even older 2-channel standard, stating +/- 30 degree for the L and R speakers. But who put the speakers like that? Probably no one in their right mind (and with the ears intact).
#Actually, it’s wrong to call it pinna-effects, and also wrong (though better) to call it HRTF. The correct term would be HTRTF.
Let’s see who'll be the first to interpret it correctly… 😉
Its unfortunately not possible to draw any conclusions about human hearing from this mentioning of unknown studies. As i wrote in a precious post the connection between the unique mentioned conclutions to established knowledge about hearing, besides from mentioning of the concept of HRTF, is missing.
Is it possible to get any reference to "available" research done in this special topic (the influence of HRTF, depending on the incidence-angle of the sound-sources, on listening experience. Besides from being a spatial cue?)?
However, I don´t doubt that the speakers characteristic and placement has importance for the listening experience, but there is need for much more research before it´s possible to draw conclutions that some specific loudspeaker-characteristics results in "correct" stereophonic illusions for instance. If it ever will be.. 🙂
Is it possible to get any reference to "available" research done in this special topic (the influence of HRTF, depending on the incidence-angle of the sound-sources, on listening experience. Besides from being a spatial cue?)?
However, I don´t doubt that the speakers characteristic and placement has importance for the listening experience, but there is need for much more research before it´s possible to draw conclutions that some specific loudspeaker-characteristics results in "correct" stereophonic illusions for instance. If it ever will be.. 🙂
Well, I' not sure you have understood our ambition with the loudspeakers.
The ambition is to make speakers for people to listen to (and more than that - to be used in recording studios, for monitor listening). Not to talk about them. Thus - the studies has been aimed at generating knowledge, to enable making of the speakers, not to educate you.
If you do not regard the information I've tried to help you with, as something of value, I'm sorry about that, but you have been offered to get all the info you want, but at the time, your response was that the info I offered you did not exist.
Well fine, believe what you want. If you do not believe it exists, that fine, and does not really matter. Use the traditional HTRTF (yes Pan, torso it is!) and apply that to what I've said. The only difference is the lesser precision that can be accomplished with the cruder data. The general principles are still the same though.
Best regards, Ingvar
The ambition is to make speakers for people to listen to (and more than that - to be used in recording studios, for monitor listening). Not to talk about them. Thus - the studies has been aimed at generating knowledge, to enable making of the speakers, not to educate you.
If you do not regard the information I've tried to help you with, as something of value, I'm sorry about that, but you have been offered to get all the info you want, but at the time, your response was that the info I offered you did not exist.
Well fine, believe what you want. If you do not believe it exists, that fine, and does not really matter. Use the traditional HTRTF (yes Pan, torso it is!) and apply that to what I've said. The only difference is the lesser precision that can be accomplished with the cruder data. The general principles are still the same though.
Best regards, Ingvar
Ok,I understand.IngOehman said:
I hope you have understanding for my curiosity about details of your marketing since my interest is learning of audio and hearing in general and not bying your speakers. For that reason its important to sort out this type of extrapolition of accepted scientifical facts.
patrikf said:
This type of extrapolition of accepted scientifical facts?
Curiosity about marketing?
Because of your interest in audio and hearing?
Well, I'm not involved in the marketing of Guru speakers (other than giving tech-info to the people responsible for the marketing). I'm only the engineer.
And I'm sorry Patrik, but I have seen few signs of curiosity regarding learning of “audio and hearing” * from your side, during the years you have been more or less stalking me. I appreciate that you are behaving more civilized here than you have done historically on other foras, but I still se no attempts from your side to try to understand… anything.
You ask questions, then you deliberately misunderstand the answers, just ignore them, or claim that something is missing or just generally criticize the answers, and finally you now label them "marketing". 😱
Why do you behave like that?
I'm very aware that this is a DIY-forum, and that no one here is a potential customer! But if there is a serious interest in some thoughts that has gone into the design of speakers I’m responsible for, I'm happy to do my best in answering questions, given the time I have to spend on it. Charing information was a positive thing 2000 years ago, and it still is.
But answering questions from some one who’s only interest seems to be to misunderstand and distort the answers to use them against the person trying to answer, proves that answering those questions is a waist of time.
So if you have no other ambition: Torment some one else. Please.
Or better - start to try understanding.
If you understand the concept of HRTF (or rather HTRTF) and what it implies, and also can picture the acoustic differences between a real life sound source moving between the speakers, and a phantom source projected between the loudspeakers - then what I’ve written so far should not be that difficult for you to understand. After all, I’ve only written about a very minor part of the design complex I apply designing loudspeakers, a fairly straight forward and reasonable easy to understand part.
If the connection between HRTF and what Iäv written is still not clear to you, tell me what it is that you do not grasp regarding how the measures I’ve taken in acoustic design relates to the HTRTF’s you say you know about - and I’ll do my best to straighten out your question marks.
And if something is unclear regarding what HTRTF is, ask me about it and I'll try to answer it to for you. For example - are you clear over how HTRTF’s enable us (anyone with two ears (yes you need two to resolve the timbral matrix)) to hear direction vertically? You need to understand that before you can move forvard to understand more complex effects.
Ask questions, and I’ll try to help you.
By "ask questions", I do not mean "criticize the answers I’ve given", nor ordering me to service you with studies made by others. Searching for information you can do by your self, and also you have already rejected to even look at what I have offered you in that direction in the past, so I will not do it again. But I am sure you know what a question is.
Best regards, Ingvar
- - - - -
*Actually, Audio is Latin and means “to hear”. Auditorium is a hall made for listening. But maybe in English, the word Audio have glided somewhat, now meaning "sound"?
Ok, what would the effects upon the listening-experience be, by NOT using the compensations related to "HTRTF" that you mention, predicted by the results of your research?IngOehman said:
That is a good question (even though you have asked it before, and I have answered it before, but maybe the answer can be of interest to some one else).
The answer is: My research predicts it to be as it is perceived in real life, i.e: I’ve found no conflicts between information the can be gathered from the HRTF’s and what can be gathered by doing experiment to learn about how the effects are perceived in real life). But – “real life” can be many different things…
The outcome depends on many things, mostly regarding the recording technique, the set up of the stereo system, the behaviour of the room in which the speakers are playing, and not the least - the directivity of the speakers.
More than anything else the opening angle between the loudspeakers is a major factor, because without that piece of the puzzle, correlation can not be investigated. The strength of the correlation depends on many more things though.
To simplify the answer (but please do not forget that it is a simplification), I'll start with a situation where:
1. The two speakers have an infinitely linear frequency response (as per measurement by omni directional microphone placed on axis in an anechoic room),
2. the speakers are set up to widely apart (+/- 30 degrees),
3. in an anechoic room, and,
4. where a coincident recording technique have been used, and the recording room itself has an acoustically dry character.
Basically, the perceived flaws due to the stereo system can be divided into two categories - timbral and geometrical distortions (both linear distortions).
The geometrical errors are typically at their worst close to the centre of the soundstage, and they normally appear as what is often referred to as "the rainbow effect", i.e. things in the vicinity of the soundstage centre appears to be higher up than things at the extreme left and right.
. . . If you study the HRTF's and try to identify the timbral behaviour at the eardrum for at frontally arriving complex (wide band) sound wave, and compare to what is created when you sum up two sound waves arriving from 30 degrees left and 30 degrees right, you will be able to see the similarity (do not forget the comb filter effect - remember that the recording in the example was coincidental). Obviously, the differential part of the timbral coloration not coinciding with the HRTF for the perceived sound wave arrival angel (above actual speaker positions, if you remember) will be perceived as a smaller timbral coloration. Which leads us to the “non geometrically enterpred” colorations:
The timbral errors that will occur (compared to real life musical instruments playing) will typically be at their worst for phantom projection angles a little bit to the left and to the right of centre position in the soundstage. Typically they are perceived as some added “harshness”, like if the sound would be equalized with quite significant emphasis at a couple of “ear unfriendly bands”, i.e. frequency bands where the typical ear is sensitive to start with, and thus less tolerant for high sound pressures. The characters I’m trying to describe, refers to what a typical set of ears/brain-system will experience. It does not mean that everyone on the planet will describe them similarly, but the characters (stereo system caused colorations) will be there (acoustically) for everyone with a fairly normal shape of pinnas and torso). Naturally, it also depends on the program material used for the study. People will be less inclined to describe the very same coloration as harsh, on very soft/mellow program material.
. . . Again, if you study the HRTF-info you claim to know about, you will se the reason for this. If you do not do that, You'll simply have to do the experiment to learn for yourself, as I have done, or you can chose trust me (maybe supported by how your own experiences in which I’m sure you recognize the effects I’m describing to you? If so, what you are doing is using those very experiences as at “retrospective experiment”, that supports my theses!).
This said, I feel that I need to remind every one reading this, that this was not only a simplification, but also a description of what is the behaviour of a much “cleaner” chain (no or few room contributions, and a coincident recording) than what is typical at home. In the above scenario, the colorations measured in dB, will actually exceed 15 dB! The dominant one of the contributing factors (creating the comb filtering), will however be reduced to nothing in every well balanced loudspeaker/room-setup, thus leaving only a few dB* that needs to be compensated - but that is still well above the threshold of hearing!
(*which is the part left after having gone from adding L och R channel acoustically in the air with the phase-component preserved, to doing it in a random manner (above 300 Hz. Below 300 Hz, that simplification is not allowed).)
So, I believe it is now time to add the complicating (and beneficial) factors of real life.
(Next posting)
The answer is: My research predicts it to be as it is perceived in real life, i.e: I’ve found no conflicts between information the can be gathered from the HRTF’s and what can be gathered by doing experiment to learn about how the effects are perceived in real life). But – “real life” can be many different things…
The outcome depends on many things, mostly regarding the recording technique, the set up of the stereo system, the behaviour of the room in which the speakers are playing, and not the least - the directivity of the speakers.
More than anything else the opening angle between the loudspeakers is a major factor, because without that piece of the puzzle, correlation can not be investigated. The strength of the correlation depends on many more things though.
To simplify the answer (but please do not forget that it is a simplification), I'll start with a situation where:
1. The two speakers have an infinitely linear frequency response (as per measurement by omni directional microphone placed on axis in an anechoic room),
2. the speakers are set up to widely apart (+/- 30 degrees),
3. in an anechoic room, and,
4. where a coincident recording technique have been used, and the recording room itself has an acoustically dry character.
Basically, the perceived flaws due to the stereo system can be divided into two categories - timbral and geometrical distortions (both linear distortions).
The geometrical errors are typically at their worst close to the centre of the soundstage, and they normally appear as what is often referred to as "the rainbow effect", i.e. things in the vicinity of the soundstage centre appears to be higher up than things at the extreme left and right.
. . . If you study the HRTF's and try to identify the timbral behaviour at the eardrum for at frontally arriving complex (wide band) sound wave, and compare to what is created when you sum up two sound waves arriving from 30 degrees left and 30 degrees right, you will be able to see the similarity (do not forget the comb filter effect - remember that the recording in the example was coincidental). Obviously, the differential part of the timbral coloration not coinciding with the HRTF for the perceived sound wave arrival angel (above actual speaker positions, if you remember) will be perceived as a smaller timbral coloration. Which leads us to the “non geometrically enterpred” colorations:
The timbral errors that will occur (compared to real life musical instruments playing) will typically be at their worst for phantom projection angles a little bit to the left and to the right of centre position in the soundstage. Typically they are perceived as some added “harshness”, like if the sound would be equalized with quite significant emphasis at a couple of “ear unfriendly bands”, i.e. frequency bands where the typical ear is sensitive to start with, and thus less tolerant for high sound pressures. The characters I’m trying to describe, refers to what a typical set of ears/brain-system will experience. It does not mean that everyone on the planet will describe them similarly, but the characters (stereo system caused colorations) will be there (acoustically) for everyone with a fairly normal shape of pinnas and torso). Naturally, it also depends on the program material used for the study. People will be less inclined to describe the very same coloration as harsh, on very soft/mellow program material.
. . . Again, if you study the HRTF-info you claim to know about, you will se the reason for this. If you do not do that, You'll simply have to do the experiment to learn for yourself, as I have done, or you can chose trust me (maybe supported by how your own experiences in which I’m sure you recognize the effects I’m describing to you? If so, what you are doing is using those very experiences as at “retrospective experiment”, that supports my theses!).
This said, I feel that I need to remind every one reading this, that this was not only a simplification, but also a description of what is the behaviour of a much “cleaner” chain (no or few room contributions, and a coincident recording) than what is typical at home. In the above scenario, the colorations measured in dB, will actually exceed 15 dB! The dominant one of the contributing factors (creating the comb filtering), will however be reduced to nothing in every well balanced loudspeaker/room-setup, thus leaving only a few dB* that needs to be compensated - but that is still well above the threshold of hearing!
(*which is the part left after having gone from adding L och R channel acoustically in the air with the phase-component preserved, to doing it in a random manner (above 300 Hz. Below 300 Hz, that simplification is not allowed).)
So, I believe it is now time to add the complicating (and beneficial) factors of real life.
(Next posting)
In a real life situation, it will be a more complex exposure of our hearing mechanism, and depending on the exact situation, the colorations will be lessened or worsened. Typically they will be quite a bit lessened.
These effects will work to our benefit in a “normal” real life situation:
1. We will not hear nearly as much of the comb filtering effects as described early above, simply because it won’t be there. This is very good, since those effects are not only very unpleasant for many listeners (on top of what I’ve tried to the describe earlier, they will also regularly give birth to complaints of “phasy” sound, “in the head” sound, depth disorder, often creating an impression of “attaching sounds”, i.e. transients jumping out of the loudspeakers and extreme sensitivity to head movements of the listener) but they also makes it impossible to chose seating freely. The impression of the recording will be very different depending on where you are seated during the listening session. During live music listening, I’ve very seldomly (never, actually) seen people fight over the “sweet spot position”. Why’s that? Well - there isn’t any! In front of live musicians the difference between different seats are not that big (unless being very far apart) and there is not any sweet spot at all (hurray!). 😉
. . . The typical (mr Smith’s) listening situation for stereo listening 2008 (as was the case 1954) is taken place in a room where the listener is seated in an acoustic field where the arriving acoustic energy is reflected sound to a degree of 60 – 90 %. (Yes, really!) Since our sense of direction (answers “from where does the sound come?”) is heavily influenced by the HAAS-effect (the precedence effect), we will not perceive a 60-90% reflected sound-situation, perspective vice. But that is still the way it is! Timbrally however, the later arriving sounds will to a large degree serve to “smooth over” the geometric and timbral flaws of the stereo system (typically, it will smooth over the comb filtering problems). The price we pay is a degradation, resolution vice, and of coarse, that we have to listen to a combination of sounds that are most often intrinsically timbrally colored, due to most speakers lack of harmony between direct and reflected sound.
. . . So – what we want is a listening environment, in combination with the specific speakers in use regarding dispersion properties, that provides a better quote between direct and reflected sound. This is a delicate balance, and I’m surprised to see that so many audiophiles tend to reason from a (very) polarized reality idea – no one seem to be searching for the balance, every one is looking for the coffin filled with gold in one of the extremes. But my experience is, that point where one can find balance is the only one that seems to be avoided. Both schools are super polarized but in opposite directions. My guess is that people are running away from what they know is wrong, rather than taking the time needed to find out where to search for the balance point, so that a much more creative attempt to running towards the target, can be started! [Basic role of good science: Newer flee, hunt! 😉 (Untruths/truths)]
. . . Anyway: People realizing that to much reflected sound takes away resolution seems to decide that reflections must be minimized (instead of optimized) and people having experienced the problems with interference (phasing) and “in the head sound”, due to too directive loudspeakers in heavily damped rooms, seems to search for overcompensating it in the other direction. To me that’s a sad thing, seeing people running to change position in the matter, without realizing that in this case (but not in all cases) there actually is a Buddhas “happy medium”, since the sum of the benefits from each angle, create such an enormous maxima, when everything is in balance. (Is the laws of acoustics the final proof of Buddhism?)
. . . What I am saying is, that the amount of reflections from the room needed to take away the interference problems (comb filtering) is (if coming from the right directions) barely audible! So why not aiming at applying that exact amount of it? The only real problem I can see is the practical one: Most rooms in combination with most loudspeakers, results in to much reflected sound, and from the wrong directions. So in reality, most often, it is not a question of adding what is needed, but to take away the exact amount…
2. If the recordings used are made with time-difference techniques; that alone will reduce the sweet spot effect (creating a much bigger “sweet field”) since a recording that includes time difference will allow a room having an even lesser “happy medium” amount of reflections (from the right directions). The reason for that is, that a recording using time-difference techniques, will make sure that all frequencies above 300 Hz will add randomly and thus create a smooth comb filter free frequency response. (Frequencies below that will be a lesser problem no matter recording technique).
When all this is said, I’d like to add two new dimensions, or rather add two vindications of alternative viewpoints on home music reproduction (alternatives to my own). Two viewpoints that I can argue because they make sense to me, even if they will not be (are not) my personal choice.
(Next posting)
These effects will work to our benefit in a “normal” real life situation:
1. We will not hear nearly as much of the comb filtering effects as described early above, simply because it won’t be there. This is very good, since those effects are not only very unpleasant for many listeners (on top of what I’ve tried to the describe earlier, they will also regularly give birth to complaints of “phasy” sound, “in the head” sound, depth disorder, often creating an impression of “attaching sounds”, i.e. transients jumping out of the loudspeakers and extreme sensitivity to head movements of the listener) but they also makes it impossible to chose seating freely. The impression of the recording will be very different depending on where you are seated during the listening session. During live music listening, I’ve very seldomly (never, actually) seen people fight over the “sweet spot position”. Why’s that? Well - there isn’t any! In front of live musicians the difference between different seats are not that big (unless being very far apart) and there is not any sweet spot at all (hurray!). 😉
. . . The typical (mr Smith’s) listening situation for stereo listening 2008 (as was the case 1954) is taken place in a room where the listener is seated in an acoustic field where the arriving acoustic energy is reflected sound to a degree of 60 – 90 %. (Yes, really!) Since our sense of direction (answers “from where does the sound come?”) is heavily influenced by the HAAS-effect (the precedence effect), we will not perceive a 60-90% reflected sound-situation, perspective vice. But that is still the way it is! Timbrally however, the later arriving sounds will to a large degree serve to “smooth over” the geometric and timbral flaws of the stereo system (typically, it will smooth over the comb filtering problems). The price we pay is a degradation, resolution vice, and of coarse, that we have to listen to a combination of sounds that are most often intrinsically timbrally colored, due to most speakers lack of harmony between direct and reflected sound.
. . . So – what we want is a listening environment, in combination with the specific speakers in use regarding dispersion properties, that provides a better quote between direct and reflected sound. This is a delicate balance, and I’m surprised to see that so many audiophiles tend to reason from a (very) polarized reality idea – no one seem to be searching for the balance, every one is looking for the coffin filled with gold in one of the extremes. But my experience is, that point where one can find balance is the only one that seems to be avoided. Both schools are super polarized but in opposite directions. My guess is that people are running away from what they know is wrong, rather than taking the time needed to find out where to search for the balance point, so that a much more creative attempt to running towards the target, can be started! [Basic role of good science: Newer flee, hunt! 😉 (Untruths/truths)]
. . . Anyway: People realizing that to much reflected sound takes away resolution seems to decide that reflections must be minimized (instead of optimized) and people having experienced the problems with interference (phasing) and “in the head sound”, due to too directive loudspeakers in heavily damped rooms, seems to search for overcompensating it in the other direction. To me that’s a sad thing, seeing people running to change position in the matter, without realizing that in this case (but not in all cases) there actually is a Buddhas “happy medium”, since the sum of the benefits from each angle, create such an enormous maxima, when everything is in balance. (Is the laws of acoustics the final proof of Buddhism?)
. . . What I am saying is, that the amount of reflections from the room needed to take away the interference problems (comb filtering) is (if coming from the right directions) barely audible! So why not aiming at applying that exact amount of it? The only real problem I can see is the practical one: Most rooms in combination with most loudspeakers, results in to much reflected sound, and from the wrong directions. So in reality, most often, it is not a question of adding what is needed, but to take away the exact amount…
2. If the recordings used are made with time-difference techniques; that alone will reduce the sweet spot effect (creating a much bigger “sweet field”) since a recording that includes time difference will allow a room having an even lesser “happy medium” amount of reflections (from the right directions). The reason for that is, that a recording using time-difference techniques, will make sure that all frequencies above 300 Hz will add randomly and thus create a smooth comb filter free frequency response. (Frequencies below that will be a lesser problem no matter recording technique).
When all this is said, I’d like to add two new dimensions, or rather add two vindications of alternative viewpoints on home music reproduction (alternatives to my own). Two viewpoints that I can argue because they make sense to me, even if they will not be (are not) my personal choice.
(Next posting)
1. To much room contribution – let’s try to fix it by increasing speaker directivity!
. . . In most typical rooms the reverberation is far too long, and the amount of early reflections is to high to allow a high resolution music reproduction. That is a valid argument for introducing loudspeakers with a much higher directivity. However, one thing can not be perfectly balanced by the other, since the reverberation of a room have many components, but right now, I will concentrate on two of them: level and time. (A huge simplification, but it will have to do here, I can not write a whole book here.) The relative level (compared to the direct sound) can be changed by using loudspeakers with a high directivity, but the reverberation time will still be the same, tough that addition to the total sound will be lower in level. Therefore, there is a limit to what can be accomplished by compensating a too lively room by using high directivity speakers. There has to be an “audible logic” between the reverberation that can be heard, and the impact of the direct sound from the speakers, if a believable soundstage is to appear. It is difficult to fool the hearing.
. . . However, it has been very clear for me, from the experiment I’ve done, that when the reverberation time is long (RT60 >0.3 seconds (above f-schröder)) using a pair of loudspeakers with a higher than usual directivity, is often a choice that renders a preferred experience, even though doing something about the reverberation time is an even better solution. And don’t forget that many (or at least some) “normal living rooms” have reverberation times longer than 0.7 seconds.
. . . In rooms of larger size, using loudspeakers with a higher directivity index than usual is an absolute necessity. Actually, I’d say, that for every room and/or rather listening distance, there is an optimal directivity for the loudspeakers, in an optimal room regarding reverberation time (which depends on the size of the room). So, adapting the speaker design for the application is not only about compensating problems to reach a compromise – it is often about optimizing the performance without having to resort to compromises.
2. To much audibility of the stereo system flaws – let’s try to fix it by decreasing speaker directivity!
. . . So, another interesting idea is to make sure that the flaws of the stereophonic system can not occur at all. I have mentioned that the comb filtering effects (due to the acoustical mix of the sound from left and right loudspeakers) that will be at hand if coincident recordings are replayed in an anechoic room, is a part that can not be compensated by modifying the loudspeakers, since it manifests itself differently depending on where the listener are in the listening room.
. . . The only way of making that problem disappear that I know of (well, there are other methods, but…) is to let the room add some reflections (late enough (but not too late) and weak enough). That will make both the comb filtering and the reflections themselves close to inaudible. The refections are (seen from another perspective) what is needed to put the closest listener further from the speaker than the “acoustic room radius times the speaker directivity index”!
. . . Using time difference recordings makes the problem even smaller, but no matter what, we still need to have a listening situation that is helping the two speakers to blend together perfectly. When we have that – the comb filtering effects are no more, and all that remains is the timbral and geometrical errors from the stereo system (the problems I describes earlier), due to having all direct sounds emit from two directions only (as oppose to reality, where sound comes from the true direction within the soundstage, and is not phantom projections between two loudspeakers). These are the problems that I’m always addressing in my loudspeaker-designs (well, one group of the problems I’m addressing. There are many more that has to be dealt with…But let’s not complicate things more than is necessary. Each problem has it’s time).
. . . We can call the former group (i.e. the comb-filter problems) for Stereo system errors group R. R for “room”. And the latter we can call group L, and L of coarse is for “loudspeaker”. The letter is standing for the position in the chain where the compensation must be applied, due to the problems inherent physical nature.
. . . But what happened when we go further than that? Well that depends on how it is done. I believe that one of the more interesting ideas is Stig Carlssons. In his orthoacoustic world, he does two things. He diffuses each loudspeaker for itself (within the time frame than normally is regarded as tabu, for reflections!) and he then works with a large enough quantity of reflections to make sure the not even the group L-problems no longer need compensation! Why? Because the phantom image is no longer built up using phase and timing clues in our ear/brain system, but only energy. (Actually, Stig Carlsson never formulated his ambition thus, but applying some reverse engineering revels the mechanisms of it’s magic as doing just that.)
Now which method to use, to decode the encoded information on a recording, is a matter of preference! Personally, I can see a spectrum of possibilities ranging outside my own as well as Stig Carlssons thoughts. I do not believe that fighting over which metode is correct or better than the other proves anything else than that the person trying to make this subjective question more objective than it is, is very restricted in his (yes he, it’s never a her) thinking.
These truths are soft truths, since the whole concept of recording an replaying sound/music is not lossless. There are many ways to balance the replaying setup, and I’ve tried to do my best to leave my personal subjective preferences out (I never use myself as a listener other than for the final polishing) and research the weight between different flaws (always by using large group of listeners) so that the sum of them will be as small and harmless as possible. Also - I’ve been in a constant search for alternative methods to “compromise”. I do not like to compromise you see! What I’m always in search for is synergy. When one action does not drag another down, but the opposite(!) it is truly amazing how the flaws can be close to erased.
. . . There, I was being subjective, or at least kind of subjective. I have made experiment though, that indicates that a stereo chain (2 ch) can be close to transparent, recording, loudspeakers and room included! Transparent enough to fool people sitting inside a cotton tent, to believe that a real life string quartet is playing in front of them, when the speakers are doing the job. And I’m not talking of the old experiment having the loudspeakers substitute only the instrument/musicians in a room that is there in both cases. I’m talking about having the speakers reproduce both instruments and the hall around them. But that is another story.
So where are we? Well, I guess that what I am trying to say is that I can not do anything more than trying my best to tell about my own conclusions, how I came to them, and then leave it up to anyone interested to try them! Try by lending their ears to my speakers.
Best regards, Ingvar Öhman
- - - - -
PS. Or to quote Forrest Gump: That's what I have to say about it.
. . . In most typical rooms the reverberation is far too long, and the amount of early reflections is to high to allow a high resolution music reproduction. That is a valid argument for introducing loudspeakers with a much higher directivity. However, one thing can not be perfectly balanced by the other, since the reverberation of a room have many components, but right now, I will concentrate on two of them: level and time. (A huge simplification, but it will have to do here, I can not write a whole book here.) The relative level (compared to the direct sound) can be changed by using loudspeakers with a high directivity, but the reverberation time will still be the same, tough that addition to the total sound will be lower in level. Therefore, there is a limit to what can be accomplished by compensating a too lively room by using high directivity speakers. There has to be an “audible logic” between the reverberation that can be heard, and the impact of the direct sound from the speakers, if a believable soundstage is to appear. It is difficult to fool the hearing.
. . . However, it has been very clear for me, from the experiment I’ve done, that when the reverberation time is long (RT60 >0.3 seconds (above f-schröder)) using a pair of loudspeakers with a higher than usual directivity, is often a choice that renders a preferred experience, even though doing something about the reverberation time is an even better solution. And don’t forget that many (or at least some) “normal living rooms” have reverberation times longer than 0.7 seconds.
. . . In rooms of larger size, using loudspeakers with a higher directivity index than usual is an absolute necessity. Actually, I’d say, that for every room and/or rather listening distance, there is an optimal directivity for the loudspeakers, in an optimal room regarding reverberation time (which depends on the size of the room). So, adapting the speaker design for the application is not only about compensating problems to reach a compromise – it is often about optimizing the performance without having to resort to compromises.
2. To much audibility of the stereo system flaws – let’s try to fix it by decreasing speaker directivity!
. . . So, another interesting idea is to make sure that the flaws of the stereophonic system can not occur at all. I have mentioned that the comb filtering effects (due to the acoustical mix of the sound from left and right loudspeakers) that will be at hand if coincident recordings are replayed in an anechoic room, is a part that can not be compensated by modifying the loudspeakers, since it manifests itself differently depending on where the listener are in the listening room.
. . . The only way of making that problem disappear that I know of (well, there are other methods, but…) is to let the room add some reflections (late enough (but not too late) and weak enough). That will make both the comb filtering and the reflections themselves close to inaudible. The refections are (seen from another perspective) what is needed to put the closest listener further from the speaker than the “acoustic room radius times the speaker directivity index”!
. . . Using time difference recordings makes the problem even smaller, but no matter what, we still need to have a listening situation that is helping the two speakers to blend together perfectly. When we have that – the comb filtering effects are no more, and all that remains is the timbral and geometrical errors from the stereo system (the problems I describes earlier), due to having all direct sounds emit from two directions only (as oppose to reality, where sound comes from the true direction within the soundstage, and is not phantom projections between two loudspeakers). These are the problems that I’m always addressing in my loudspeaker-designs (well, one group of the problems I’m addressing. There are many more that has to be dealt with…But let’s not complicate things more than is necessary. Each problem has it’s time).
. . . We can call the former group (i.e. the comb-filter problems) for Stereo system errors group R. R for “room”. And the latter we can call group L, and L of coarse is for “loudspeaker”. The letter is standing for the position in the chain where the compensation must be applied, due to the problems inherent physical nature.
. . . But what happened when we go further than that? Well that depends on how it is done. I believe that one of the more interesting ideas is Stig Carlssons. In his orthoacoustic world, he does two things. He diffuses each loudspeaker for itself (within the time frame than normally is regarded as tabu, for reflections!) and he then works with a large enough quantity of reflections to make sure the not even the group L-problems no longer need compensation! Why? Because the phantom image is no longer built up using phase and timing clues in our ear/brain system, but only energy. (Actually, Stig Carlsson never formulated his ambition thus, but applying some reverse engineering revels the mechanisms of it’s magic as doing just that.)
Now which method to use, to decode the encoded information on a recording, is a matter of preference! Personally, I can see a spectrum of possibilities ranging outside my own as well as Stig Carlssons thoughts. I do not believe that fighting over which metode is correct or better than the other proves anything else than that the person trying to make this subjective question more objective than it is, is very restricted in his (yes he, it’s never a her) thinking.
These truths are soft truths, since the whole concept of recording an replaying sound/music is not lossless. There are many ways to balance the replaying setup, and I’ve tried to do my best to leave my personal subjective preferences out (I never use myself as a listener other than for the final polishing) and research the weight between different flaws (always by using large group of listeners) so that the sum of them will be as small and harmless as possible. Also - I’ve been in a constant search for alternative methods to “compromise”. I do not like to compromise you see! What I’m always in search for is synergy. When one action does not drag another down, but the opposite(!) it is truly amazing how the flaws can be close to erased.
. . . There, I was being subjective, or at least kind of subjective. I have made experiment though, that indicates that a stereo chain (2 ch) can be close to transparent, recording, loudspeakers and room included! Transparent enough to fool people sitting inside a cotton tent, to believe that a real life string quartet is playing in front of them, when the speakers are doing the job. And I’m not talking of the old experiment having the loudspeakers substitute only the instrument/musicians in a room that is there in both cases. I’m talking about having the speakers reproduce both instruments and the hall around them. But that is another story.
So where are we? Well, I guess that what I am trying to say is that I can not do anything more than trying my best to tell about my own conclusions, how I came to them, and then leave it up to anyone interested to try them! Try by lending their ears to my speakers.
Best regards, Ingvar Öhman
- - - - -
PS. Or to quote Forrest Gump: That's what I have to say about it.
- Status
- Not open for further replies.