Hello,
Stable phantom imaging in a stereo triangle at 8kHz will certainly not happen. If you pour wax into your pinnas you might have a slight change
Easy to test: put a 7kHz high pass to your speakers and listen to some stereophonic sounds.
In this case I hear only two tweeters at +/-30 degrees, if direct to reflection ratio is high. In case direct to reflection ratio is low sound is just 'spacious'. But no phantom in either case
Phantoming happens at lower freq ranges.
Due to this reason I avoid speakers with high directivity tweeters, like horns. I think it's better to have 'spacious' treble than contradicting imaging.
- Elias
Stable phantom imaging in a stereo triangle at 8kHz will certainly not happen. If you pour wax into your pinnas you might have a slight change
Easy to test: put a 7kHz high pass to your speakers and listen to some stereophonic sounds.
In this case I hear only two tweeters at +/-30 degrees, if direct to reflection ratio is high. In case direct to reflection ratio is low sound is just 'spacious'. But no phantom in either case
Phantoming happens at lower freq ranges.
Due to this reason I avoid speakers with high directivity tweeters, like horns. I think it's better to have 'spacious' treble than contradicting imaging.
- Elias
Hello Elias,
maybe i did not make clear enough what i was talking
about when writing "presence to brillance region" ...
Presence region is often assumed starting at 2Khz and is
correlated with the curves of equal loudness:
http://www.sengpielaudio.com/ISO-KurvenANeu.gif
A presence filter used in studios has often significant
influence on frequencies even lower than 1Khz
http://www.elektroniktutor.de/analog/an_pict/filter5.gif
If we go up in frequency including the brillance range and
also including the directional band for 'above' localization
- which was the starting point for my thoughts on
balance between presence and 'sparkle' range -
we end up in a frequency range from 2Khz to 8Khz or
even larger i was talking about.
Kind Regards
Last edited:
...
That range could also be called "the highs" if you like,
when adding that uppemost octave for convenience
and the younger listeners. My goal was to describe my
observation, that a change in vertical directivity pattern
occuring between presence and the 'above presence'
region affects the sense of "sparkle" and the
- subjective- integration of the presence region
experienced over different listening positions in the room,
even with on axis frequency response being the same.
Kind Regards
That range could also be called "the highs" if you like,
when adding that uppemost octave for convenience
and the younger listeners. My goal was to describe my
observation, that a change in vertical directivity pattern
occuring between presence and the 'above presence'
region affects the sense of "sparkle" and the
- subjective- integration of the presence region
experienced over different listening positions in the room,
even with on axis frequency response being the same.
Kind Regards
Last edited:
Vertical/horizontal directivity
I am yet to see a sensible 'quantization' of vertical and horizontal directivity patterns of speakers and their effects on reproduction and realism.
But much before such rarefied arguments became fashionable, particularly in regard to 'popular' two-way bookshelf speakers, many of us had noticed tangible differences in imaging and what we call 'realism' when speakers were oriented vertically or horizontally.
Hi-fi audio sometimes is unfortunately a confused attempt at reproducing a ruler-flat FR at one moment, while at the other we bewail the loss of 'imaging' and root for that, and there are surely many other 'schools' of thought, in fact as many as there are enthusiasts!
I personally go for a situation where the reproduction is able to "emotionally" recreate what the original audio event was capable of--irrespective of absolute FR, particularly the LF and HF limits. And I have found that a simple system like an OB with a passable response from about 100 Hz to 10 kHz, set up in a 'free-field' condition was able to give more subjective satisfaction than systems that were in most ways "superior" to the above in many ways. "Accurate beaming" HF, in my humble opinion, never went with an "accurate" reproduction. I agree with LineArrays thoughts in so far as that they point to the "need" for the brilliance region to contribute to "realism", and also the need for a 'wider', uniform vertical dispersion, with as few confusing colourations as possible, in order to achieve better realism and less 'bitchiness'.
It is of paramount importance that more knowledgeable minds should devote time and effort to identify these factors and contribute to our better understanding of these factors vis-a-vis our pursuit of "reproduced realism".
May realism reign!
I am yet to see a sensible 'quantization' of vertical and horizontal directivity patterns of speakers and their effects on reproduction and realism.
But much before such rarefied arguments became fashionable, particularly in regard to 'popular' two-way bookshelf speakers, many of us had noticed tangible differences in imaging and what we call 'realism' when speakers were oriented vertically or horizontally.
Hi-fi audio sometimes is unfortunately a confused attempt at reproducing a ruler-flat FR at one moment, while at the other we bewail the loss of 'imaging' and root for that, and there are surely many other 'schools' of thought, in fact as many as there are enthusiasts!
I personally go for a situation where the reproduction is able to "emotionally" recreate what the original audio event was capable of--irrespective of absolute FR, particularly the LF and HF limits. And I have found that a simple system like an OB with a passable response from about 100 Hz to 10 kHz, set up in a 'free-field' condition was able to give more subjective satisfaction than systems that were in most ways "superior" to the above in many ways. "Accurate beaming" HF, in my humble opinion, never went with an "accurate" reproduction. I agree with LineArrays thoughts in so far as that they point to the "need" for the brilliance region to contribute to "realism", and also the need for a 'wider', uniform vertical dispersion, with as few confusing colourations as possible, in order to achieve better realism and less 'bitchiness'.
It is of paramount importance that more knowledgeable minds should devote time and effort to identify these factors and contribute to our better understanding of these factors vis-a-vis our pursuit of "reproduced realism".
May realism reign!
The ceiling reflections hardly abate the Interaural Cross Correlation coefficient. That's correctly. Thus those reflections are not causing the effect of "more space". However, there are very important in terms of "correct space".
In the recording room the reflections arrive from all possible directions. Reducing the playback onto the horizontal level of the listener is a significant loss. In the second chapter of the linked website is described a simple mirror source model of the spatial sonic field in the recording room. It describes a way for restoring the concert hall acoustics on a snowy meadow. Theoretically, of course, however, becomes graspable, on what score, we cannot dispense reliable ceiling reflections.
H.
Last edited:
Vertical/horizontal directivity
@Syntheticwave
The trouble, to my largely 'uneducated' mind appears to be the plethora of directions from which the needed signals as well as the reflections come to bombard the poor listener's ears. Coherence of any sort suffers, obviously, the soundfield being one big spaghetti of mashed up signals, direct and reflected.
Dont you think we have been stuck with the "two ears, two speakers" setup for long enough? Its simplicity and "quick results" have lulled us into virtual inaction in chasing the Holy Grail of reproduced realism, with the "emotional content" intact.
My own as yet incomplete and imperfect experiments in reproducing M-S signals as such using two full-range (100 Hz -- 10 kHz) closely spaced OB speaker drivers have given me some encouraging results. Good or bad, I dare say they are "way different" from what you perceive with the 'conventional' stereo speaker arrangement.
Who knows what tomorrow shall bring...
@Syntheticwave
The trouble, to my largely 'uneducated' mind appears to be the plethora of directions from which the needed signals as well as the reflections come to bombard the poor listener's ears. Coherence of any sort suffers, obviously, the soundfield being one big spaghetti of mashed up signals, direct and reflected.
Dont you think we have been stuck with the "two ears, two speakers" setup for long enough? Its simplicity and "quick results" have lulled us into virtual inaction in chasing the Holy Grail of reproduced realism, with the "emotional content" intact.
My own as yet incomplete and imperfect experiments in reproducing M-S signals as such using two full-range (100 Hz -- 10 kHz) closely spaced OB speaker drivers have given me some encouraging results. Good or bad, I dare say they are "way different" from what you perceive with the 'conventional' stereo speaker arrangement.
Who knows what tomorrow shall bring...
Interesting way to frame the dichotomy. Too often arguments occur when the participants use the same word as one another, but assign different meanings to it. The technically oriented use "accurate" in its technical sense - output exactly follows input. The experience oriented use "accurate", when they probably mean "realistic" (as in; makes me believe "they are here" or "you are there").
The problem with "accurate", though it is easy to define, is that we can typically define it only for a subset of the signals that reach our ears. So, while it may be accurate with respect to some well defined parameters, our choice of parameters is critical to the what we perceive.
On the other hand, "realistic", is a matter of what we perceive, not necessarily complete accuracy of all (or even any) audio content. That's more complicated, because it is the sum total or our individual perception preferences and all the information we process at the time (and yes, including whether we like the look of special stones on our cables).
Seems like research is moving more into the complicated area of which components of audio reproduction are important to creating a certain kind (or different kinds) of "realism". In the meantime if we don't confuse "accurate" with "realistic", we might avoid some unproductive disagreement.
Sheldon
Hello,
Something like this?
http://www.diyaudio.com/forums/multi-way/10962-stereolith-loudspeakers-question-3.html#post1633904
- Elias
Something like this?
http://www.diyaudio.com/forums/multi-way/10962-stereolith-loudspeakers-question-3.html#post1633904
- Elias
....let me coming back to the main page of the linked site. I believe, that describes exactly your thougths:
"In the recording room, the sound source is originating a complex pattern of reflections. Truly spatial audio reproduction has to restore the wave fronts of the source itself, as well as its reflections, from its correct starting point – irrespective of the listener's position. That is impossible in the traditional, phantom source based procedures. Assigning all spatial information to some channels, as we have performed since Alain Blumlein's time, cannot truly reconstruct the genuine impression of space and size of the recording room. The complex structure of the sound field is not simply describable as a signal difference between some single points.
By attempting to increasing the immersion with the number of channels, we have lost the substance of the sources itself. We are not really immersed into the sonic field. The sound sources are only around us. Ghostlike now, somehow dematerialized, and disappearing, in case we want draw near. This site describes a different procedure, aiming at the opposite. During playback becomes complemented the purely, dry in mono recorded sound of the source itself by the correct reflections. A synthesis creates those wave fronts in the same manner, as the recording room creates all of them from the pure signal of the sound source itself. That's shown in the little animation above. The following chapters will delineate the way. First we will be describing the complex problems of the traditional audio procedures."
As become described, would it be possible, establish a virtual, three dimesional copy of the genuine sonic field. All we need is leave the worn- out traditional procedures.
H.
I don't think stereo is a bad thing. What good examples of beaming may have shown us is the problem with reflections on intelligibility and the stereo signal.
I first noticed this with car twin-cones and later HiFi full-rangers. As perfect as the concept is not, there is something about the limited dispersion of the higher frequencies that is revealing to the whole process of in room reproduction.
You may have discovered it with dipoles. I am exploring it with waveguides. We want the sound to be dispersed evenly, and not necessarily off to the sides where we would be encouraging the earlier reflections.
Reproduced realism
@Sheldon
Thanks, Sheldon. I guess we are beginning to see the problem now.
Broadly I think if we go for realism and take the example suggested by LineArray of ancient man's hearing mechanism having evolved along a path of practicality, we would accept that if the hunter doesnt want to be the hunted, his priority in listening would be imaging and accurate localization. If you were the type who bothered more about FR and other niceties, you'd perhaps earn a place as pre-history's first audiophile, but IT would have got you and you wouldnt be alive to tell the world about your superior audio experience.
I am no musician, but witnessing and live musical event with "emotional content", and listening to its reproduction over various chains, with some attention paid to 'musicality', would tell you what sound more 'real'--of course, to you. But surely there are factors that go beyond measurements. This is NOT to denigrate those devoted souls whose art and science of measurement has advanced the state of our hobby very much.
Perhaps it is time for us to shed our obsession with 'stereo' as we know it and move on to something else that would better mimic the real.
@Elias:
Yes, sir, something like that, But my setup is a feed of the M-S signals as such into the amp and then to a pair of stacked OBs at right angles, located comparatively free-field, a-la the M-S mic arrangement. (I am not very keen about matrixing L and R.) I think it is one the easiest experimental setups that every enthusiast should lash up and experience. Like I said earlier, you would agree it is something way beyond what you are accustomed to. The near-field experience is something that is pleasantly different, and I think more 'live', with less of the confusing smear.
@Syntheticwave:
Yes, I get you. I join you whole-heartedly in that wish/prayer that somebody should find a way to encode at least minimally what all we need to have in order to reproduce 'realism' with not much loss of flavour! It is, I think, a lot like MP3 encoding. If you know how much you can 'throw away' without affecting quality, that would give you a method for encoding the essentials of the sonic field.
@AllenB:
Sir, within my experience, vehicle environments and their EQ and reproduction are a totally different kettle of fish. It is entertainment, simple and good, sometimes great, but nobody is going to imgine that it is going to reproduce 'reality' in a grand manner.
Let us remember that every little step is good and that matters; but it surely is not the last. We need to take more steps forward before we could come to realize our dreams. The moment we think we have perfection with our two speakers and the 'sweet spot', that is likely to spell trouble. Let us be more bothered with the 'ends' than the 'means'--we want to experience 'realism', as real as it can get. Period.
@Sheldon
Thanks, Sheldon. I guess we are beginning to see the problem now.
Broadly I think if we go for realism and take the example suggested by LineArray of ancient man's hearing mechanism having evolved along a path of practicality, we would accept that if the hunter doesnt want to be the hunted, his priority in listening would be imaging and accurate localization. If you were the type who bothered more about FR and other niceties, you'd perhaps earn a place as pre-history's first audiophile, but IT would have got you and you wouldnt be alive to tell the world about your superior audio experience.
I am no musician, but witnessing and live musical event with "emotional content", and listening to its reproduction over various chains, with some attention paid to 'musicality', would tell you what sound more 'real'--of course, to you. But surely there are factors that go beyond measurements. This is NOT to denigrate those devoted souls whose art and science of measurement has advanced the state of our hobby very much.
Perhaps it is time for us to shed our obsession with 'stereo' as we know it and move on to something else that would better mimic the real.
@Elias:
Yes, sir, something like that, But my setup is a feed of the M-S signals as such into the amp and then to a pair of stacked OBs at right angles, located comparatively free-field, a-la the M-S mic arrangement. (I am not very keen about matrixing L and R.) I think it is one the easiest experimental setups that every enthusiast should lash up and experience. Like I said earlier, you would agree it is something way beyond what you are accustomed to. The near-field experience is something that is pleasantly different, and I think more 'live', with less of the confusing smear.
@Syntheticwave:
Yes, I get you. I join you whole-heartedly in that wish/prayer that somebody should find a way to encode at least minimally what all we need to have in order to reproduce 'realism' with not much loss of flavour! It is, I think, a lot like MP3 encoding. If you know how much you can 'throw away' without affecting quality, that would give you a method for encoding the essentials of the sonic field.
@AllenB:
Sir, within my experience, vehicle environments and their EQ and reproduction are a totally different kettle of fish. It is entertainment, simple and good, sometimes great, but nobody is going to imgine that it is going to reproduce 'reality' in a grand manner.
Let us remember that every little step is good and that matters; but it surely is not the last. We need to take more steps forward before we could come to realize our dreams. The moment we think we have perfection with our two speakers and the 'sweet spot', that is likely to spell trouble. Let us be more bothered with the 'ends' than the 'means'--we want to experience 'realism', as real as it can get. Period.
Sound & Realism
With your permission I would point at least some of you to Chapter-2 of my blog posts at < Sound... unboxed ! >, in which I try and explain some of these concepts for our little group of youngsters, who sometimes lose sight of the aim of the 'war' in the heat of the 'skirmishes'!
Advance apologies for any disappointments!
With your permission I would point at least some of you to Chapter-2 of my blog posts at < Sound... unboxed ! >, in which I try and explain some of these concepts for our little group of youngsters, who sometimes lose sight of the aim of the 'war' in the heat of the 'skirmishes'!
Advance apologies for any disappointments!
That's a very interesting approach, abandoning the traditional ways. However, the system approach ends at the loudspeakers. Nevertheless,the most mistakes ar arising if the loudspeakers have done its work. We have to include the playback room properties into the system approach, thus.
The Holophony approach subtract the supplemental detours of the sound waves in playback room during synthesis, separately for the direct wave front and her reflections. Loudspeakers and room are really one system in that procedure.
H.
3DAA allows you to do just that. The idea is to separate sounds and their spatial information from rendering (playback). If you want your renderer to include aspects of the playback room then go ahead and build one.
Hi,
To me it appears the same? M (=L+R) and S (=L-R) signals fed to two dipoles.
- Elias
To me it appears the same? M (=L+R) and S (=L-R) signals fed to two dipoles.
- Elias
Ambisonics has been doing the same since 70's. To separate the spatial sounds from the rendering.
- Elias
- Elias
Hello,
Also many things can be concluded by observing HRTF.
For example to emulate the spectral balance of a side or a rear sound source by a frontal real source one can put a notch at 3kHz (BBC dip !) and boost at 8kHz.
http://homepage.mac.com/cooperbauck/cbpublications/documents/prospectsfortransaural.pdf
- Elias
Also many things can be concluded by observing HRTF.
For example to emulate the spectral balance of a side or a rear sound source by a frontal real source one can put a notch at 3kHz (BBC dip !) and boost at 8kHz.
An externally hosted image should be here but it was not working when we last tested it.
http://homepage.mac.com/cooperbauck/cbpublications/documents/prospectsfortransaural.pdf
- Elias
When it comes to the physical aspects of sound production (or reproduction), the answer to that is no. Unless, that is, we want to invoke magic. Historically that hasn't held up too well to examination. Just because we don't know which measurements to take doesn't invalidate the principle.
When it comes to the perceiver, we're not as far along in measuring or generalizing aspects of perception. And there remains the issue of preference, or at least priorities.
And it's not just the sound field. How many live performances have you gone to where you can actually close your eyes and place the performers? None, that are amplified, I dare say, because they aren't produced that way. And in a concert hall? Good luck with "pinpoint" imaging. To replace the visual cues (yes our ancestors used those too) that come with the live experience you need a sort of hyper realism to create the same emotional reaction.
Something like this? Recombinant Media Labs at UCSD Roundup | barry threw
Sheldon
3DAA is an object based procedure. That separates content ( the dry recorded, pure audio signals ) and form ( the information regarding the recording room properties ). During playback, the synthesis merge together both components, much more sophisticated as at all conventional approaches. The result is very close the genuine event, however the playback room acustics have to be supress for true perception. In other case the signal would shaped from two rooms, the recording room and the playback room.
However, living in an anechoic chamber may result in trouble with some unregenerate spouses. Much more clever seem, construct a common model from recording room and playback room, as in the Holophony procedure. By those object based procedure, a combination of the model based and data based approach allows, constitute an common approach. In the near field of an large WFS loudspeaker field behind the silver screen, the playback room acoustics become subordinated matter. Thus, the common model allow subtracting playback room acoustic. Is easily comprehensible in this little animation:
http://www.holophony.net/pictures/WFS_transformation_principle.swf
We are at the work to build that. I hold the patent for the procedure since more as five years. The result will be the physikal reconstruction of the spatial sonic field.
Regards H.
Last edited: