Toole page 359:
"Dissecting the data and looking at the results for listeners of different genders and level of experience, Figure 17.13 shows that experienced males distinguished themselves by delivering lower scores for all of the loudspeakers. This is a common trend among experienced listeners. Otherwise, the pattern of the ratings was very similar to those provided by inexperienced males and females.
...
Lack of experience in both sexes shows up mainly in elevated levels of variability in responses (note the longer error bars), but the responses themselves, when averaged, reveal patterns similar to those of more experienced listeners. With experienced listeners, statistically reliable data can be obtained in less time."
So experienced listeners would provide better absolute ratings but a listening / preference test is relative anyway.
page 361:
"Olive (2003), some of whose results are shown in Figure 17.6, compiled data on 268 listeners and found no important differences between the ratings of carefully selected and trained listeners and those from several other backgrounds, some in audio, some not, some with listening experience, some with none. There were, as shown in Figure 17.6, huge differences in the variability and scaling of the ratings, so selection and training have substantial benefits in time savings. Rumsey et al. (2005) also found strong similarities in ratings of audio quality between naive and experienced listeners, anticipating only a 10% error in predicting ratings of naive listeners from those of experienced listeners."
Also, I believe naive listeners can very well differentiate what sounds natural (G, yet another attribute
) and what does not. When I refer to naive listeners then I am talking about educated people, who are able to explain differentiatedly in their mother tongue what they hear and these people were certainly not born with iPoo in their ears as a novel body part but they know nothing about early, late and what not reflections etc.
It is time to review how stereo works at freq below about 1kHz. The two wavefronts from the speakers form an interference field at the listening position that will generate ITD at the ear canals through cross talk. It is due to the cross talk that amplitude panning is transformed into phase differences in ear canals. BUT, the formed interference field, even it propagates energy, is a fake field because it does not propagate at the natural speed, but the velocity vector is less than unity. This is one of the reasons why turning of the head at the stereo field makes the image to shift and collapse, because turning the head at a given amount generates ITD that is higher than natural ITD in a natural field.
Few people understands this, and propably in their lack of knowledge think stereo is such a great invention. sigh.
I think it's the beast range to use dipole line array. I'm using in 100Hz - 2 kHz. I'd propably like to cross at 1kHz but suitable tweeter is hard to find.- Elias
Yes, and what sounds normal people usually get exposed during all days? Natural environmental sounds of course ! Not sound of the damped studio !
It should be "they all voted for the response with lots of reflections because that is what they're familiar with from real world everyday life, the reality"
This is simply too loose definition and does not tell anything. Directional at what frequency range? Let me guess, omni at the bass, high directional at the treble
But it should be just the opposite, highly directional at the bass and only little directional at the treble. If you think it through and take into account the room and human perception, you'll agree with me.
- Elias
6.283,
If you have a very large number of naive listeners you'll get more or less the same outcome as what you'd get with a few experienced listeners. Does this really mean naive listeners are equally good at rating speakers as experienced listeners? I don't think so. Most naive listeners rate a speaker either a lot higher or a lot lower than experienced listeners. The average listener rating of naive listeners is more or less the same as that of experienced listeners, so apparently there is no significant systematic error. But the random error makes the listening experience of naive listeners unreliable.
EDIT: wrong spelling corrected
If you have a very large number of naive listeners you'll get more or less the same outcome as what you'd get with a few experienced listeners. Does this really mean naive listeners are equally good at rating speakers as experienced listeners? I don't think so. Most naive listeners rate a speaker either a lot higher or a lot lower than experienced listeners. The average listener rating of naive listeners is more or less the same as that of experienced listeners, so apparently there is no significant systematic error. But the random error makes the listening experience of naive listeners unreliable.
EDIT: wrong spelling corrected
Last edited:
And one solution is not to use that much 'phantom' imaging but discreet sources. Multichannel is here, every ht receiver has an upmixer integrated. Well SFS is great also, but has anyone seen/heard one outside universities?
Good quotes.
In conclusion the only reason to use professional listeners is that it will be cheaper for the researcher because with them you can find statistical significance faster wth smaller N.
Usually, and generally, it is rare to see high N in any research paper from any field. Typically N is less than 10. Normally N is 4, including the author, coauthor, the professor and maybe the author's wife
- Elias
After listening to an extremely high quality recording of the Portland Oregon Chamber Orchestra through the Linkwitz Orions, I decided I wanted a dipole of Seas Millenium domes on each side for 1440HZ on up. Not sure I've ever heard better treble.It is time to review how stereo works at freq below about 1kHz. The two wavefronts from the speakers form an interference field at the listening position that will generate ITD at the ear canals through cross talk. It is due to the cross talk that amplitude panning is transformed into phase differences in ear canals. BUT, the formed interference field, even it propagates energy, is a fake field because it does not propagate at the natural speed, but the velocity vector is less than unity. This is one of the reasons why turning of the head at the stereo field makes the image to shift and collapse, because turning the head at a given amount generates ITD that is higher than natural ITD in a natural field.
Few people understands this, and propably in their lack of knowledge think stereo is such a great invention. sigh.
- Elias
Your comment about imaging below 1kHZ leaves me confused. I don't know what ITD is, and if it's inter-aural crosstalk, the rest still confuses me. But I'd like to understand your explanation. Where I used the word "timing", I probably should have used the word "phase", which is the same thing in this context. Above about 1kHZ, the ear-brain mechanism won't be able to know which period it's comparing, since the half wavelength gets shorter than the distance between the ears, so above that frequency it switches over to using amplitude comparisons. In a conventional setup, embedded imaging cues are only useful in the uppermidrange, unless this second set of inter-aural crosstalk is somehow reduced or eliminated. Any perceived imaging or spacial cues outside of the upper-midrange will be a product of listening room reflections.
Except in this case the two groups came to different conclusions, not the same conclusion in different amounts of time.
David S.
Hi Keyser,
I agree and what you state was imbedded/implied in both of my posts: "naive listeners" are less efficient and as Elias also says they are thus more expensive.
My whole point was that only because some paper is talking about "naive listeners" you should not conclude that their rating and impression is useless as questioned below.
That's all.
I agree and what you state was imbedded/implied in both of my posts: "naive listeners" are less efficient and as Elias also says they are thus more expensive.
My whole point was that only because some paper is talking about "naive listeners" you should not conclude that their rating and impression is useless as questioned below.
That's all.
It is time to review how stereo works at freq below about 1kHz. The two wavefronts from the speakers form an interference field at the listening position that will generate ITD at the ear canals through cross talk. It is due to the cross talk that amplitude panning is transformed into phase differences in ear canals. BUT, the formed interference field, even it propagates energy, is a fake field because it does not propagate at the natural speed, but the velocity vector is less than unity. This is one of the reasons why turning of the head at the stereo field makes the image to shift and collapse, because turning the head at a given amount generates ITD that is higher than natural ITD in a natural field.
Few people understands this, and propably in their lack of knowledge think stereo is such a great invention. sigh.
- Elias
How would a dipole line array help with speaker crosstalk?
different - yes but not so different
it seems that there are two conclusions from the study:
1) strong early reflections don't lead to degraded stereo imaging - so ideal directivity-pattern-for-stereo-speakers directional<->omni dilemma is NOT a question of accuracy of stereo image
2) some professional listeners (roughly half of the group in the study) have a kind of occupational disease - they have become oversensitized to reverberation and because of that they prefer RFZ approach and directional speakers - but it is NOT a question of accuracy but just of preference
so, in the specific context of our discussion here I want ask all participants to refrain from judging other participants preferences - it is perfectly all right to prefer highly directional horn et al. or omni because both can be equally accurate as far as accuracy of the stereo image is concerned
it would be nice to leave this pointless quarrel behind and to discuss other issues in the context of - for example:
of ideal directivity pattern for stereo speakers vs factors of realism of sound reproduction other than accuracy of stereo image
or
of ideal directivity pattern for stereo speakers vs factors of high quality sound reproduction other than accuracy of stereo image
Yes, (Speaker) Dave, good paper.
Here is another one about the subject.
Effects of loudspeaker directivity on perceived sound quality - a review of existing studies
The authors plan an extensive test about directivity. The paper was presented in Munich at the 126th AES convention in 2009.
Anyway, there seems to be no answer to the question (also of this thread), yet.
Hopefully, my Omni Demokrit is able to provide some answers for me at least.
Here is another one about the subject.
Effects of loudspeaker directivity on perceived sound quality - a review of existing studies
The authors plan an extensive test about directivity. The paper was presented in Munich at the 126th AES convention in 2009.
Anyway, there seems to be no answer to the question (also of this thread), yet.
Hopefully, my Omni Demokrit is able to provide some answers for me at least.
Have You read the paper?
Isn't it that for now the answer is simply NO. I mean "no" to the question whether one directivity pattern (narrow) is better than the other (wide/omni) from the perspective of perceived sound quality? At least as far as quality of stereo image is concerned?
back to the main topic
is the so called strong early spatial impression (see Markus' post below) desirable from perspective of high quality(/realistic) sound reproduction?
if yes then - what directivity pattern for stereo speakers is capable of creating of such impression?
it seems that early reflection free narrow directivity pattern is not:
I'm with Markus, and I also think that this kind of spaciousness is a factor of high quality sound reproduction
so it looks like some sort of wide pattern creating the appropriate pattern of early reflections is needed
what's You opinion?
is the so called strong early spatial impression (see Markus' post below) desirable from perspective of high quality(/realistic) sound reproduction?
if yes then - what directivity pattern for stereo speakers is capable of creating of such impression?
it seems that early reflection free narrow directivity pattern is not:
I'm with Markus, and I also think that this kind of spaciousness is a factor of high quality sound reproduction
so it looks like some sort of wide pattern creating the appropriate pattern of early reflections is needed
what's You opinion?
Last edited:
This thread seems to partially change topic into
"drawbacks of stereophonic reproduction".
But the topic was
"What is the ideal directivity pattern for stereo speakers?"
That makes stereo a fixed point IMO.
Maybe it would be helpful to distinguish between frequency bands in
answering that question.
My proposal would be distinguishing at least 3 bands as a first attempt:
- low frequencies < Schroeder frequency of the listening room
- low to mid band, where ITD is dominant in natural spatial hearing
- high band, where ILD (and envelope) is dominant in natural spatial hearing
"drawbacks of stereophonic reproduction".
But the topic was
"What is the ideal directivity pattern for stereo speakers?"
That makes stereo a fixed point IMO.
Maybe it would be helpful to distinguish between frequency bands in
answering that question.
My proposal would be distinguishing at least 3 bands as a first attempt:
- low frequencies < Schroeder frequency of the listening room
- low to mid band, where ITD is dominant in natural spatial hearing
- high band, where ILD (and envelope) is dominant in natural spatial hearing
Last edited:
I agree!
but dominant under what conditions? do we really know what spatial cues are dominant under reverberant conditions, especially in a small room, especially in case of virtual sound sources, especially multiple, especially in presence of two real sound sources (stereo loudspeakers)?
furthermore the duplex theory that You refer to is being questioned even on case of localization of one real sound source in the free field
a small repost with additional underlines may be useful:
therefore I think it is scientifically unfounded to try to design stereo loudspeakers in order to follow assumptions of the duplex theory
Last edited:
difficult to tell, here's what I have found:
Introduction to Psychoacoustics - Module 08B
I also tend to think that onset ITD (as opposed to ongoing time/phase ITD) must be important if we assume that precedence effect is indeed something real (though perhaps something more complicated than we tend to think when we overgeneralize results of Haas' exepriments) because this effect relies on things that happen <1 ms where we have just onsets/wavefronts arriving at our ears
I also tend to think that if we assume that precedence effect is indeed something real then our hearing must be able to identify the onsets as similiar (and as such falling under the effect as reflected copies of one original sound) or dissimilar (heralding a possible new sound event)
the only thing that allows for identification of the particular onset is it's shape therefore I tend to think that our hearing analyses shapes of the oncoming onsets, this shape is basically determined by energy in frequencies below say ~1600 Hz where most of the energy of natural/musical sounds is concentrated
in conclusion I hypothesize that the freq band crucial for spatial sound reproduction is roughly below 1.6 kHz and it is important to preserve the basic onset shape in the direct and all reflected onsets - not to slice it's timeline in filters, speaker drivers time misalignments etc.
OTOH I think that our hearing is especially happy when it gets strong and congruous (supporting coherently the ATF and onset ITD cues) duplex kind of spatial cue like ILD/IID in the highs - perhaps a point for all setups producing low IACC like Elias' "stereolits" for example?
what do You think?
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.