Okay, summing up what was said (relative to thread's subject that the Danley video posted above about loudspeaker depth location):
1) Danley pointed out that for his MEH work in the fixed install professional audio market, as the MEH loudspeaker designs integrated linear phase into the design to greater and greater degrees, the ability of listeners to discern apparent distance to (stereo) loudspeakers becomes increasingly difficult (a good thing to achieve for home hi-fi and professional fixed install applications).
2) My own experience shows that as stereo loudspeakers having full-range directivity control in a controlled early-reflection listening room achieve a threshold phase and SPL flatness performance, apparent stereo and multichannel depth of field increased suddenly (which is desirable hi-fi capability to enhance).
3) Griesinger has written and lectured on the subject of clarity as related to phase fidelity of harmonics in large listening spaces (mostly driven by large venue early reflections both inside and outside listener Haas intervals). This is interesting but not coincident with enhanced perception of acoustic depth--not including visual/auditory cues and phenomena (another subject outside of the present topic for this thread). Perhaps more on this subject in another thread.
4) There is speculation that low diffraction loudspeakers (referencing horn mouth termination in particular) is likely a source of loudspeaker/room interference which gives listeners clues to where the loudspeakers actually are. I tend to agree that this is likely, but no relevant near field or intermediate (before transition) field measurements have been brought to bear on the problem.
5) Other visual/aural effects and a modified-headphone related source was cited, but are not in the domain of subject that Danley broached in his video. In order to keep this thread focused, this other information will not presently be discussed in this thread (but may later be raised if relevance to the current subject can be tied more closely.
6) There seems to be various attributions that don't actually relate to the subject at hand, but rather are more like "pep rallies" around certain psychoacoustic authors or perhaps some form of popularity of certain users--but no direct references to studies on apparent acoustic depth perception factors. I don't believe I have a lot of time for "pep rally" activity, and will continue to discourage that which is not directly related to the subject at hand.
So the choices for the source of this effect thus far:
a) "acoustic summing" within the horn aperture, which not only controls lobing effects, but also artificial cues to loudspeaker location.
b) phase and directivity fidelity (avoiding early reflections in-room), combined with SPL fidelity.
c) avoidance of diffraction effects (horn termination or cabinet edges re-radiating)
Nowhere have we been talking about late reflections from the listening space to enhance the sense of apparent source depth. Late reflections can be artificially introduced through techniques like multichannel loudspeaker arrays (5.1, etc.) and mixing/mastering electronically added reverberation. Selection of the recording space acoustics and microphone positions are functions not controllable by listeners or audio editors. This is surprising--I think.
Chris
1) Danley pointed out that for his MEH work in the fixed install professional audio market, as the MEH loudspeaker designs integrated linear phase into the design to greater and greater degrees, the ability of listeners to discern apparent distance to (stereo) loudspeakers becomes increasingly difficult (a good thing to achieve for home hi-fi and professional fixed install applications).
2) My own experience shows that as stereo loudspeakers having full-range directivity control in a controlled early-reflection listening room achieve a threshold phase and SPL flatness performance, apparent stereo and multichannel depth of field increased suddenly (which is desirable hi-fi capability to enhance).
3) Griesinger has written and lectured on the subject of clarity as related to phase fidelity of harmonics in large listening spaces (mostly driven by large venue early reflections both inside and outside listener Haas intervals). This is interesting but not coincident with enhanced perception of acoustic depth--not including visual/auditory cues and phenomena (another subject outside of the present topic for this thread). Perhaps more on this subject in another thread.
4) There is speculation that low diffraction loudspeakers (referencing horn mouth termination in particular) is likely a source of loudspeaker/room interference which gives listeners clues to where the loudspeakers actually are. I tend to agree that this is likely, but no relevant near field or intermediate (before transition) field measurements have been brought to bear on the problem.
5) Other visual/aural effects and a modified-headphone related source was cited, but are not in the domain of subject that Danley broached in his video. In order to keep this thread focused, this other information will not presently be discussed in this thread (but may later be raised if relevance to the current subject can be tied more closely.
6) There seems to be various attributions that don't actually relate to the subject at hand, but rather are more like "pep rallies" around certain psychoacoustic authors or perhaps some form of popularity of certain users--but no direct references to studies on apparent acoustic depth perception factors. I don't believe I have a lot of time for "pep rally" activity, and will continue to discourage that which is not directly related to the subject at hand.
So the choices for the source of this effect thus far:
a) "acoustic summing" within the horn aperture, which not only controls lobing effects, but also artificial cues to loudspeaker location.
b) phase and directivity fidelity (avoiding early reflections in-room), combined with SPL fidelity.
c) avoidance of diffraction effects (horn termination or cabinet edges re-radiating)
Nowhere have we been talking about late reflections from the listening space to enhance the sense of apparent source depth. Late reflections can be artificially introduced through techniques like multichannel loudspeaker arrays (5.1, etc.) and mixing/mastering electronically added reverberation. Selection of the recording space acoustics and microphone positions are functions not controllable by listeners or audio editors. This is surprising--I think.
Chris
On the subject of diffraction, I did enjoy to be able to see the test results of 16 different horns, done by J.M. Le Cléac’h (may he rest in peace).
Linked here: http://unepassionaudiophile.fr/wp-content/uploads/2018/01/Horns_measurements_ETF2010d-1.pdf
If reflections are detrimental to perception, than this test may show some valid results in favor of proper care of horn mouth termination.
Linked here: http://unepassionaudiophile.fr/wp-content/uploads/2018/01/Horns_measurements_ETF2010d-1.pdf
If reflections are detrimental to perception, than this test may show some valid results in favor of proper care of horn mouth termination.
Chris,
The effect of loudspeaker distance location being harder to detect with a single full range speaker, a co-axial, or multiple entrant horn (MEH) compared to loudspeakers with separate components spread over a longer vertical distance is fairly simple to explain.
Sound waves arriving at the ears are reflected from the shoulders and pinnae, ("outer ears" as Tom mentioned around 2:20) and those reflections modify the frequency response as it enters the ear canal. The peaks and troughs in response are located in different frequency ranges depending on arrival angle, allowing localisation of the sound source in the vertical plane, the angle of incidence between them corresponding to a location cue.
Since a MEH reduces the vertical height of the apparent source compared to separate components, our hearing has less vertical locational cues, so the distance to the speaker is harder to determine.
Art
Since high frequencies in multiple aperture loudspeakers almost always lead the lower frequencies (due to passive or active IIR crossover filter all-pass phase growth--at least), where is the Haas effect (precedence effect) in masking the location of the lower frequencies relative to higher frequencies in your explanation?
How would you explain the imaging depth performance of electrostatic or dynamic film-type planar dipole loudspeakers (head in a vice performance)--especially large ones?
Chris
Last edited:
Where are the measurements of the effects of diffractions (higher horn mode generation)? You can't quantify "diffraction amount" if you're not measuring it. You need multiple measurements across the throat or horn mouth span in at least the vertical and horizontal directions vs. frequency to quantify "diffraction" vs. differing horn profile geometries. I believe that just discussing diffraction without quantifying via measurement is a fruitless exercise in guessing. (Impulse response measurements do not measure diffraction, but only their gross effect at a certain point in space in front of the horn mouth(s)).
Chris
Last edited:
Assuming the HF is above the MF and LF, the Haas effect would localize the speaker at an upwards angle, rather than to the acoustic center.
Too many years have passed since I've heard an electrostatic for any imaging memories, though their polar response explains the "head in a vice" performance.
Measurements are useful, as are the results of controlled listening tests. I watched both Danley videos which have been referenced in this thread, and not one single measurement or other empiirical evidence was presented to support their assertions. I "get it" that these are sales/promotional videos, but it would have been useful if they had presented some actual data.
Regarding diffraction at the horn mouth, I'm not familiar with any data about that, although it's probably out there somewhere. What I am familiar with is sound wave energy being reflected from the horn mouth back into the horn. One important factor is how the mouth meets the outside world. The more abrupt the interface, the more reflection there is. The smoother the interface, the less reflection there is. This is because of the acoustical impedance "mismatch" between the horn and the open air.
Last edited:
Danley Sound Labs is a professional PA manufacturer and as such heavily based on having precise specs and tons of measurements, otherwise they wouldn't sell and survive. Therefore I'm inclined to trust them although I would agree some real data would have been helpful, in the videos. Yeah, and a bit less of marketing blurb. We know that Danley knows there have true been point sources previously even though his horns are actually dual-domain point sources, in space (geometry) and in time which makes them somewhat unique (as far as passive loudspeakers go).
----:----
As for the thread topic, my personal little findings are pretty much in line with what Tom Danley is saying.
It's all a question of preferences in the end.
Many ways of projecting stereo signals to a listeners are valid, including omnis (monopoles) like Linkwitz Plutos in a rather reverberant normal living room which give away their location as good as possible... which is basically the antipode of a Synergy in a well damped and diffused studio ambience. The latter is good for a "I am there" style of projection whereas omnis give the "They are here" kind of impression, regarding the auditory scene relation between listener and artists.
The data of how each Synergy horn will measure will be available at Danley's commercial site, as he shares those files (as measured by a third party) for commercial purposes.
Not for a minute I would regard the things he said in these video's to be purely commercially driven, as if you look up his posts on this forum, he already shared that same info many times over, if I had to guess I'd say it was to share with us how "he sees these matters". Certainly not for bragging rights though, he has been sharing good ideas and ditto info for years.
Like this quote (dating back 10 years):
He's been sharing a lot of info and his views on many audio related topics, which has been inspirational for many members here, to say the least. Always in a kind, humble and civilized way. The video, to me, looks like an honest talk of a man that is very passionate about his work.
He touched on a few other subjects in these video's, just look at his posts on this forum to find it in written language long before those video's were created.
The diffraction of a horn mouth I wrote about, just look up the papers and posts from Dr. Geddes. From a quick search on Google:
Original post: https://audioroundtable.com/forum/index.php?t=msg&th=3347&goto=18213&#msg_18213Earl Geddes on AudioRoundTable said:
More info:
[1] Geddes, E., R.: "Sound Radiation from Acoustic Apertures." JAES., vol. 41, pp. 214-23 (April 1993)
Source: https://audioroundtable.com/forum/index.php?t=msg&th=3347&prevloaded=1&&start=0
Thanks for reposting Danley. I couldn't quite understand the short bit about one ESL63 (my first system after graduation) -- often cited for "palpable presence".
Danley: "One can only localize the depth location of a speaker IF there is enough difference in the speakers radiation between what arrives at your right and left ears."
There's a lot to infer (learn) from this statement. While working on crossfeed headphone stereo sound I tried a 30-sec experiment: (eyes closed) easily localize direction and distance of sound source in front or from behind (can just turn around); then try with, flat palm over nose, two or three fingers behind skull as fake nose. Localization weakened or even destroyed.
OP might point out this is head-shape induced difference in sound not speaker-related. In fact head-shape-psychoacoustics magnifies speaker-related issues such as non-point-source, beyond measurement mic limitations (incidence angle to pinna) that Danley pointed out in the first quote above.
Danley: "One can only localize the depth location of a speaker IF there is enough difference in the speakers radiation between what arrives at your right and left ears."
There's a lot to infer (learn) from this statement. While working on crossfeed headphone stereo sound I tried a 30-sec experiment: (eyes closed) easily localize direction and distance of sound source in front or from behind (can just turn around); then try with, flat palm over nose, two or three fingers behind skull as fake nose. Localization weakened or even destroyed.
OP might point out this is head-shape induced difference in sound not speaker-related. In fact head-shape-psychoacoustics magnifies speaker-related issues such as non-point-source, beyond measurement mic limitations (incidence angle to pinna) that Danley pointed out in the first quote above.
Last edited:
Interesting video, but nothing new for us old Danley fans. Curious: I'm surprised that no one has mentioned the problems of diffraction and closely related distortions like HOM. Geddes' HOM are somewhat elusive, but ordinary waveguide cavity resonance as well as the problems of diffraction of sharp edges, are AFAIK very well-known phenomena. You can have the best transducer in the world (perhaps a Danley), but if you are creating diffraction at the mouth or elsewhere in the wave's path, it generates "ricochets" that, it is claimed, act as secondary sound sources. The ear is not easily fooled. As good as Danley's speakers are (I've only very briefly heard them) I fail to see why, unmodified, they would not suffer -- at least in theory -- the problems mentioned.
The good news is that these problems are easily soluble using tricks like the Geddes foam plug, the towel mod or equivalents.
The good news is that these problems are easily soluble using tricks like the Geddes foam plug, the towel mod or equivalents.
The second Tom Danley Youtube video on this subject has been released, this time he talks about stereo imaging:
So the effect that Tom is apparently talking about is the psychoacoustic effect dealing with very short delays (he mentions "...on the order of two inches..."), which correspond to the regime of time delays less than 0.7 milliseconds, as shown in the figure below. This corresponds to acoustic delays of physically less separation than 8-10 inches. The figure below illustrates this as a distinct psychoacoustic effect from the precedence effect (a.k.a., Haas effect):
Dual source time delays less than 8-10 inches apart from the listener (in distance or time) produce a perceived localization of sound that moves to between the two sources of sound, with the latter sound source delayed by less than 0.7 ms. This is what Tom is talking about in terms of in-ear cancellations that combine to give hearing cues to source distance/delay. Only one sound is perceived (just like the precedence effect) but the perceived source of the sound is moved to a position between the two sources, very unlike the precedence effect (longer delays between sources greater than 0.7 ms) in which the sound source is localized to the position of the first source arrivals.
So there are two psychoacoustic effects in play that affect our perception of distance, one at less than 8 inches of (time-based) separation, the other at greater than 8-10 inches of (time-based) separation.
In multiway loudspeakers using multiple horn apertures or drivers separated vertically and horizontally on a flat baffle, there is another effect on stereo imaging that is related to a phase difference (time-based separation) of 1.0 to 1.5 wavelengths:
For typical tweeter-midrange crossover frequencies that are used in loudspeakers --crossing typically at 4-6 kHz, the physical separation (time misalignments) of 0.25 ms (0..00025 seconds) to 0.16 ms (0.00016 seconds) are audible. For even higher frequencies approaching the limit of human hearing, the phase-based time delays are even shorter.
So for perceived distance perception, the human hearing system seems to be less sensitive to estimating distance to the sources than in the case two-source phase separation due to crossed multiway loudspeakers, as much as 10x less sensitive.
In my own experience, once the loudspeaker drivers are within 1/4 wavelength at the crossover frequency interference band in x,y,z space (i.e., including depth) as they are in a time aligned MEH (such as Danley produces), they will appear as a single source to the human hearing system as close as putting your head inside the aperture of the horn mouth.
When you think about the physical size of multiway loudspeaker drivers mounted on a flat baffle (i.e., multiple drivers physically separated--not coaxial), there is no way to achieve time alignment off-axis of perpendicular to the plane of a flat loudspeaker baffle. The more off-axis you go, the greater the time misalignments.
Chris
So the effect that Tom is apparently talking about is the psychoacoustic effect dealing with very short delays (he mentions "...on the order of two inches..."), which correspond to the regime of time delays less than 0.7 milliseconds, as shown in the figure below. This corresponds to acoustic delays of physically less separation than 8-10 inches. The figure below illustrates this as a distinct psychoacoustic effect from the precedence effect (a.k.a., Haas effect):
Dual source time delays less than 8-10 inches apart from the listener (in distance or time) produce a perceived localization of sound that moves to between the two sources of sound, with the latter sound source delayed by less than 0.7 ms. This is what Tom is talking about in terms of in-ear cancellations that combine to give hearing cues to source distance/delay. Only one sound is perceived (just like the precedence effect) but the perceived source of the sound is moved to a position between the two sources, very unlike the precedence effect (longer delays between sources greater than 0.7 ms) in which the sound source is localized to the position of the first source arrivals.
So there are two psychoacoustic effects in play that affect our perception of distance, one at less than 8 inches of (time-based) separation, the other at greater than 8-10 inches of (time-based) separation.
In multiway loudspeakers using multiple horn apertures or drivers separated vertically and horizontally on a flat baffle, there is another effect on stereo imaging that is related to a phase difference (time-based separation) of 1.0 to 1.5 wavelengths:
For typical tweeter-midrange crossover frequencies that are used in loudspeakers --crossing typically at 4-6 kHz, the physical separation (time misalignments) of 0.25 ms (0..00025 seconds) to 0.16 ms (0.00016 seconds) are audible. For even higher frequencies approaching the limit of human hearing, the phase-based time delays are even shorter.
So for perceived distance perception, the human hearing system seems to be less sensitive to estimating distance to the sources than in the case two-source phase separation due to crossed multiway loudspeakers, as much as 10x less sensitive.
In my own experience, once the loudspeaker drivers are within 1/4 wavelength at the crossover frequency interference band in x,y,z space (i.e., including depth) as they are in a time aligned MEH (such as Danley produces), they will appear as a single source to the human hearing system as close as putting your head inside the aperture of the horn mouth.
When you think about the physical size of multiway loudspeaker drivers mounted on a flat baffle (i.e., multiple drivers physically separated--not coaxial), there is no way to achieve time alignment off-axis of perpendicular to the plane of a flat loudspeaker baffle. The more off-axis you go, the greater the time misalignments.
Chris