Hello,
Directional speakers yes, especially at low freqs and midrange is essential.
If you are refering to Summa, I see no constant directivity there. It's just a box with omnipolar bass and midrange and a high directivity tweeter.
- Elias
gedlee said:
Directional speakers yes, especially at low freqs and midrange is essential.
If you are refering to Summa, I see no constant directivity there. It's just a box with omnipolar bass and midrange and a high directivity tweeter.
- Elias
Elias said:
Well we clearly don't agree on the psychoacoustics and even read the same literature and conclude completely different things from it. Above about 800 Hz the Summa is as Constant Directivity as any speaker that you can find. While below 800 Hz all data points to localization being dominated by the signal content ABOVE this frequency not the "bass and midrange". You have provided no evidence to say otherwise.
Distortions by HRTFs (head related transfer functions) that deliver spatial cues (Elias uses "pinna localization" what I've never read in literature and is inaccurate) occur at 500 Hz and above caused by the head, at 1500 Hz and above caused by the pinna and at 3000 Hz and above caused by the combination of pinna and ear canal (see Blauert (!) or Begault). There is NO distortion beyond 1500 Hz caused by the pinna that could contribute to localization. This simply is a proven fact.
Best, Markus
Best, Markus
Markus
I completely agree that the term "Pinna localization" is not defined in the literature. What is quantified as contriibutions by the Pinna to localization is only in effect above about 2 kHz. The contention of bass or midrange localizations being significant to sound imaging from two speakers in a small room is simply not supported by the data.
Elias seems to have a strong bias for "dipole line arrrays" and appears to have a hidden agenda to support his prior beliefs about their effectiveness. While I do not argue about such arrays having a greater directivity at LFs I do argue with this approach being ideal, optimum or even practical to impliment. What advantages they may have at LFs are more than offset by disadvantages at HFs, the region critical to imaging.
I completely agree that the term "Pinna localization" is not defined in the literature. What is quantified as contriibutions by the Pinna to localization is only in effect above about 2 kHz. The contention of bass or midrange localizations being significant to sound imaging from two speakers in a small room is simply not supported by the data.
Elias seems to have a strong bias for "dipole line arrrays" and appears to have a hidden agenda to support his prior beliefs about their effectiveness. While I do not argue about such arrays having a greater directivity at LFs I do argue with this approach being ideal, optimum or even practical to impliment. What advantages they may have at LFs are more than offset by disadvantages at HFs, the region critical to imaging.
Hello,
Turning of the head problem is due to fact that 2 speaker stereo triangle is not capable reproducing even a simple plane wave correctly. You can do the analysis of the interference field and see that. Hint: compare the distance of constant phase contours to the wave length.
- Elias
el`Ol said:
Turning of the head problem is due to fact that 2 speaker stereo triangle is not capable reproducing even a simple plane wave correctly. You can do the analysis of the interference field and see that. Hint: compare the distance of constant phase contours to the wave length.
- Elias
mige0 said:
Here is a link:
http://www.uni-heidelberg.de/presse/news05/2508waru.html
There has been a test in the Audio magazine and some following discussions at the Hifi-Forum.
Hello,
I don't know why do you use the word 'distortion' in this context because sound diffraction around the head is natural thing to happen.
The term 'pinna localisation' I use to refer to the localisation mechanism provided by HRTF at high freqs (say >3kHz), the directional dependent dips and valleys at the freq response pressure inside of the ear canal of the single ear (of course both ears are involved too, but one ear emphasised because pinna localisation allows spatial localisation with one ear only).
Finally you are begining to follow my point I've been trying to make: pinna localisation will be effective at high freqs (~>3kHz).
- Elias
markus76 said:
I don't know why do you use the word 'distortion' in this context because sound diffraction around the head is natural thing to happen.
The term 'pinna localisation' I use to refer to the localisation mechanism provided by HRTF at high freqs (say >3kHz), the directional dependent dips and valleys at the freq response pressure inside of the ear canal of the single ear (of course both ears are involved too, but one ear emphasised because pinna localisation allows spatial localisation with one ear only).
Finally you are begining to follow my point I've been trying to make: pinna localisation will be effective at high freqs (~>3kHz).
- Elias
Did we all miss something? In all of your last posts you argued the converse!?
The term "distortion" is correct as the HRTFs distort the original signal. There is linear distortion (doesn't add new frequencies) and non-linear distortion (adds new frequencies that were not in the original signal).
The term "distortion" is correct as the HRTFs distort the original signal. There is linear distortion (doesn't add new frequencies) and non-linear distortion (adds new frequencies that were not in the original signal).
markus76 said:
certainly not all
I didn't notice any discrepancy in Elias' arguments
perhaps because I have hidden agenda
oh well, anyway 😉
best!
graaf
Hello,
Show me the posts I argue pinna localisation is not high freq phenomena?
- Elias
markus76 said:
Show me the posts I argue pinna localisation is not high freq phenomena?
- Elias
Hello,
Why would it be? He argued I gave contradictory information in my posts and he didn't show any particular post he is refering when making that claim. I'm asking he to clarify the claim he made.
- Elias
graaf said:
Why would it be? He argued I gave contradictory information in my posts and he didn't show any particular post he is refering when making that claim. I'm asking he to clarify the claim he made.
- Elias
Elias said:
of course I was ironically joking
anyway, I was labelled in this thread as disrecpectful troll
and You have been already labelled as person with hidden agenda
so better BEWARE! 😉
my opinion I have already expressed above:
best!
graaf
Graaf,
Yes I understood your wink 😉
I just want Markus to show the inconsistency he is claiming I made.
- Elias
Yes I understood your wink 😉
I just want Markus to show the inconsistency he is claiming I made.
- Elias
Hello,
I'm thinking all you need to do in recording session is to capture the soundfield, i.e. the pressure and velocity of X, Y and Z directions.
Only 4 channels needed (B format).
Commercial version of such a mic:
http://www.soundfield.com/
- Elias
syntheticwave said:
I'm thinking all you need to do in recording session is to capture the soundfield, i.e. the pressure and velocity of X, Y and Z directions.
Only 4 channels needed (B format).
Commercial version of such a mic:
http://www.soundfield.com/
- Elias
Why can't we all show some respect and quite being sarcastic.
Markus, many people do associate the term "distortion" with nonlinear effects. I am always careful to say "linear distortion" when using the term as you have used it. You are of course correct and the context makes it clear what you meant, but confusion will almost always prevail when we are not clear.
Elias, I either misunderstand you or I agree with Markus. You seem to be saying that "pinna localization" is factor at LFs and this is clearly NOT the case. You have repeatedly made the case for directivity at LF as an aid to "imaging" and Markus (I believe) and I don't see it that way.
Markus, many people do associate the term "distortion" with nonlinear effects. I am always careful to say "linear distortion" when using the term as you have used it. You are of course correct and the context makes it clear what you meant, but confusion will almost always prevail when we are not clear.
Elias, I either misunderstand you or I agree with Markus. You seem to be saying that "pinna localization" is factor at LFs and this is clearly NOT the case. You have repeatedly made the case for directivity at LF as an aid to "imaging" and Markus (I believe) and I don't see it that way.
Elias, I re-read your postings and found that I misunderstood what you were trying to say about "pinna localization". But nonetheless you overstate the influence of HRTFs in stereophony. Interaural time and phase differences form very stable phantom images despite the fact that the HRTF of the phantom source doesn't match the HRTF of a real source at the position of the phantom source. This can lead to coloration of the phantom image - that's why there are recording and mixing engineers 🙂
There's an interesting experiment performed by Klensch (1948!) - you can find it in Blauert's book: The effect of the HRTF was eliminated by connecting each ear canal to a cone through a flexible tube. Of course the sound coming from a speaker in front of the proband was localized within the head. But when rotating the head and the cones accordingly, the sound was localized in front. So there are stronger mechanisms that deliver spatial cues then the distortions our head, pinna and ear canal produce - especially in the horizontal pane.
Best, Markus
There's an interesting experiment performed by Klensch (1948!) - you can find it in Blauert's book: The effect of the HRTF was eliminated by connecting each ear canal to a cone through a flexible tube. Of course the sound coming from a speaker in front of the proband was localized within the head. But when rotating the head and the cones accordingly, the sound was localized in front. So there are stronger mechanisms that deliver spatial cues then the distortions our head, pinna and ear canal produce - especially in the horizontal pane.
Best, Markus
Earl:
I went back and looked at the previous posts, (yes, actually READ THEM), and without to much difficulty I could tell that:
1. Any bit of sarcasm was only expressed by one person recently, and it was exceedingly modest. Nor was it in context to show disrespect to anyone, rather it was designed to *highlight* the disrespect that this person endured here. (i.e. it was *poignant*.)
2. Elias has NOT ONCE argued that pinna localization cues were a factor at lower freq.s.. Frankly I do not understand HOW you could come to such a conclusion AT ALL - except that perhaps you are so focused on Markus's previous correspondence that you have some how adopted this representation.
As for #1 - it doesn't have any real relevance in this thread: its NOT A TOPICAL ISSUE. Nor is questioning someone's respect. IF you find that someone has personally "flamed" another, then by all means please alert the moderators.
with that hopefully out of the way, and on to the thread's purpose..
With #2 - Elias has argued for directive low freq.s, but at least within the last 4 pages of the thread I haven't seen a significant reason *WHY* (..though I do remember a "hint" of a reason).
I'd like to "hear" a well thought out explanation why (..in particular because I've come to a similar conclusion). We already know that you haven't found any compelling reason for low freq. directivity, but if Elias expands on this issues then you might have something more meaningful to discuss (..other than it isn't practical).
I went back and looked at the previous posts, (yes, actually READ THEM), and without to much difficulty I could tell that:
1. Any bit of sarcasm was only expressed by one person recently, and it was exceedingly modest. Nor was it in context to show disrespect to anyone, rather it was designed to *highlight* the disrespect that this person endured here. (i.e. it was *poignant*.)
2. Elias has NOT ONCE argued that pinna localization cues were a factor at lower freq.s.. Frankly I do not understand HOW you could come to such a conclusion AT ALL - except that perhaps you are so focused on Markus's previous correspondence that you have some how adopted this representation.
As for #1 - it doesn't have any real relevance in this thread: its NOT A TOPICAL ISSUE. Nor is questioning someone's respect. IF you find that someone has personally "flamed" another, then by all means please alert the moderators.
with that hopefully out of the way, and on to the thread's purpose..
With #2 - Elias has argued for directive low freq.s, but at least within the last 4 pages of the thread I haven't seen a significant reason *WHY* (..though I do remember a "hint" of a reason).
I'd like to "hear" a well thought out explanation why (..in particular because I've come to a similar conclusion). We already know that you haven't found any compelling reason for low freq. directivity, but if Elias expands on this issues then you might have something more meaningful to discuss (..other than it isn't practical).
Scott
The comment on "respect" was not limited to just this post. And I don't really understand
"Nor was it in context to show disrespect to anyone, rather it was designed to *highlight* the disrespect that this person endured here. (i.e. it was *poignant*.)"
My claims here have been simple: that localization, which I hold synonimous with "image" in this context (although it is not in general) is dominated by HF ques. We have to get the HFs (> 1 kHz) right if we want good "imaging". This seems intuitively obvious to me and is supported by all the data that I have seen.
So let's assume that we have gotten the HFs right with a CD waveguide of narrow directivity oriented such that it minimizes early reflections. Now we need to consider the LFs. To do this the directivity of the LF source has to match that of the HF source at the crossover. This dictates a rather large speaker that is round since all my waveguides are round (this is a debatable point, but not pertinent here). This situation will hold good directivity control down to about 500-700 Hz. Below this there are some things that would improve upon the directivity, such as an unbaffled woofer.
But an unbaffled woofer adds a lot of problems with lose of sensitivity and seriuos EQ requirements and basically dictates an active crossover and multiple power amps. When I consider what we gain for what we loose, it appears to me that these are not good tradeoffs. We gain directivity, but is this going to produce a better image? I think that data is clear that this is not likely to be significant since the image is dominated by the higher frequencies (i.e. above 700 Hz. where the directivty is well controlled).
So I do not see how a directional LF source is a good design tradefoff to make.
If someone else believes that it is then I suggest that they do the designs, build the prototypes, test them and then get several independent reviews of the results (all things that I have done to support my contensions) and we can compare all of this to what I have shown. But I haven't seen any of that as yet, although I would certainly be happy to look at it if it were available.
What has happened is a lot of quick criticisms about one design approach without offering up alternative design approaches that do not suffer from new problems that are worse than those they claim to solve.
I have considered open baffled woofers on several occasions and always come away with exactly this situation - more problems created than solved.
The comment on "respect" was not limited to just this post. And I don't really understand
"Nor was it in context to show disrespect to anyone, rather it was designed to *highlight* the disrespect that this person endured here. (i.e. it was *poignant*.)"
My claims here have been simple: that localization, which I hold synonimous with "image" in this context (although it is not in general) is dominated by HF ques. We have to get the HFs (> 1 kHz) right if we want good "imaging". This seems intuitively obvious to me and is supported by all the data that I have seen.
So let's assume that we have gotten the HFs right with a CD waveguide of narrow directivity oriented such that it minimizes early reflections. Now we need to consider the LFs. To do this the directivity of the LF source has to match that of the HF source at the crossover. This dictates a rather large speaker that is round since all my waveguides are round (this is a debatable point, but not pertinent here). This situation will hold good directivity control down to about 500-700 Hz. Below this there are some things that would improve upon the directivity, such as an unbaffled woofer.
But an unbaffled woofer adds a lot of problems with lose of sensitivity and seriuos EQ requirements and basically dictates an active crossover and multiple power amps. When I consider what we gain for what we loose, it appears to me that these are not good tradeoffs. We gain directivity, but is this going to produce a better image? I think that data is clear that this is not likely to be significant since the image is dominated by the higher frequencies (i.e. above 700 Hz. where the directivty is well controlled).
So I do not see how a directional LF source is a good design tradefoff to make.
If someone else believes that it is then I suggest that they do the designs, build the prototypes, test them and then get several independent reviews of the results (all things that I have done to support my contensions) and we can compare all of this to what I have shown. But I haven't seen any of that as yet, although I would certainly be happy to look at it if it were available.
What has happened is a lot of quick criticisms about one design approach without offering up alternative design approaches that do not suffer from new problems that are worse than those they claim to solve.
I have considered open baffled woofers on several occasions and always come away with exactly this situation - more problems created than solved.
- Status
- Not open for further replies.