Before the stereo age there have been no multiple and physically separated primary sound sources, which were correlated nevertheless. It has always been one primary source and any number of correlated reflections.
Evolution has even enabled us to locate a sound source from its reflections only, if the primary signal is subdued in any way. Only if our brain can't associate a stereo mix with an evolutionary learned source-reflection relationship, it will locate a sound directly at a loudspeaker.
Rudolf
indeed. I introduced them here before the topic subject change... back then they seemed pretty well on-topic.That's not exactly the type of speaker discussed in this thread
idea, design and measurements are not mine: you'd better post your interesting comments and questions where they would be more on-topic and the author will read them. Actually, I'd love to see that. It's always a good thing when very competent people meet and discuss, and new ideas may emerge...
About the response, I don't know exactly how the measurements have been done, etc. You should really ask the author about that. I can only tell you that the author briefly explained that he have put a lot of care in the choice of the components (speakers) as well as cross-over point and design used for this project exactly to achieve this kind of goal. In particular the small and "almost naked" tweeter have been chosen because of its very wide and uniform dispersion.
Last edited:
Before the stereo age there have been no multiple and physically separated primary sound sources, which were correlated nevertheless. It has always been one primary source and any number of correlated reflections.
Evolution has even enabled us to locate a sound source from its reflections only, if the primary signal is subdued in any way. Only if our brain can't associate a stereo mix with an evolutionary learned source-reflection relationship, it will locate a sound directly at a loudspeaker.
Rudolf
C'mon Rudolf - just try a transaural setup and then You will know
Dome tweeters and wide dispersion? This one's more than 10dB down at 10kHz:
Demokrit-T
Lead by Linkwitz, the new mantra seems to be "on-axis and off-axis frequency responses need to be identical" and yet the associated speakers don't have these properties.
well, that can not be done...
laid on a side they will no longer work as intended (no more omnidirectional emission, no more "flooding", "wrong" frequency response, etc). Perhaps, a more reasonable experiment could be done with a conventional (e.g. a small bookshelf) speaker laid on its back... 
Anyway, I've got your point. Having no direct experience I can't tell whether you're right or not.
Though I suspect that dispersion and freq. response of the used speakers (as well as their position in the room) may make a lot of difference.
For instance, the old, first version of the periakusma (see here) were smaller and shorter than the current one (and also had a less well-balanced freq. response). With them, we found out that their performance was better if we listened being seated on the floor, thus with ears at a level closer to that of the speakers. But! the problem was not (it have never been) about "wrong" imaging or speakers becoming localizable. Even when standing up near one of them! There was only a problem of frequency response (too much highs when the speakers were "too low" WRT the listener ears).
As said, the most peculiar characteristic of these speakers is that there is no small "sweet spot" where you need to be to enjoy a correct presentation: you can move freely and even "walk around" the room while listening... (hence their name).
If properly placed in the room, the image always remains the same (well, perhaps about the same... I didn't measure it with a meter!
), in the same position, with same apparent width and height. Even if standing near one of the speakers, thus with direct sound coming "from the floor" on your side.Given that experience, I don't see it impossible for something like what Graaf propose to behave similarly.
The Dayton ND20 is significantly better in this regard
:
Measured without additional baffle to speak of.
I don't see a significant difference? In the end "on-axis and off-axis frequency responses need to be identical" is probably an unnecessary exaggeration which should really read "on-axis frequency response and frequency response of significant early reflections need to be identical".
Last edited:
No, I'd side with Sigfried on this one. The critical part of getting this right is all in the balance between direct and reflecting. The speaker I mocked up was in a sense an SL Pluto. To get this right both must balance. 6.5" Peerless SDS-160F25PR01-08 excels at this when crossed LR2 @2200. Cone distortion from this driver is quite low, but it's an old school motor with all the odd order harmonics that come with them. Nice smooth non fatiging midrange. BTW I do use a miniDSP for this, but even a passive wouldn't be difficult to do the same on this driver.
I'll add that diffraction play a big role in getting the best out of this too
Last edited:
Certainly the amount of reflection and how they are dispersed into the room is important but it's a typical given that most ceilings are ~8' and are free of obstructions and at least here in the US have ceilings made of drywall and are normally small popcorn (cellulose) textured. From this and the typical stereo triangle staging should yield very similar results.
Also I don't suggest using a driver inverted like this. Far too much diffraction from the frame and motor assembly. Flip the driver and suspend the tweeter over it using fishing line, spacing it forward on posts if necessary and time correcting with DSP. Dampening material on the tweeters backside would also help.
Personally believe that the best drivers to use have little or no peaking breakup and must naturally roll off nice and smooth. This is more important than distortion specs any day of the week. Pistonic action above physical driver dimensions is a waste for anything but Prosound.
Been thinking about my current project and wondered if there was a means of successfully adding this to my design. MTWWW or MTMWWW or the latter but the upper M would be facing up somewhere between flat and 30 degrees, while maintaining CTC 1/4 wavelength @2200. < Oy Vie, there's a tuff nut to crack. Perhaps due to the rising response of my AMT if placed 15-30 degrees off axis upward... just thinking out loud The WWW are the 160's in a floor loaded MLTL
The wifus accepts these things about me thank heavens
BTW I don't like crossing in the 3-4k range, my ear is just too sensitive to it. That typically rules out 4-5" drivers, my favorite sizes for near field, go figure
Also I don't suggest using a driver inverted like this. Far too much diffraction from the frame and motor assembly. Flip the driver and suspend the tweeter over it using fishing line, spacing it forward on posts if necessary and time correcting with DSP. Dampening material on the tweeters backside would also help.
Personally believe that the best drivers to use have little or no peaking breakup and must naturally roll off nice and smooth. This is more important than distortion specs any day of the week. Pistonic action above physical driver dimensions is a waste for anything but Prosound.
Been thinking about my current project and wondered if there was a means of successfully adding this to my design. MTWWW or MTMWWW or the latter but the upper M would be facing up somewhere between flat and 30 degrees, while maintaining CTC 1/4 wavelength @2200. < Oy Vie, there's a tuff nut to crack. Perhaps due to the rising response of my AMT if placed 15-30 degrees off axis upward... just thinking out loud
The WWW are the 160's in a floor loaded MLTLThe wifus accepts these things about me thank heavens
BTW I don't like crossing in the 3-4k range, my ear is just too sensitive to it. That typically rules out 4-5" drivers, my favorite sizes for near field, go figure
granted.
Though it could very possibly (and likely) be related to some perceptual "effect", in this case I was thinking to some physical effect, too.
Obviously, when two (or more) correlated physical sound sources are emitting in a common space you don't get two independent sound fields, but just one which is the result of their combination. Now, what if the physical interactions of the two correlated sound sources in a confined space, when the proper conditions (distances, relative amplitudes, phase, ...) are met, creates e.g. something similar to what the ambiophonics "RACE" processing does (that is XTC), and perhaps not just for one small spot?
I've cited ambio because I've tried it (using RACE and its recommended speaker setup) and have got results which are quite similar to what I've got with my conventional setup with only very careful "sweet spot" speaker positioning.
(...except that with RACE the "effect" is limited to one really small listening sweet spot, just moving or turning the head a little bit screws up everything - the infamous "head in a vice" effect which I can't stand - while with my conventional system I can turn my head, move freely on my couch and even stand up and move around in a quite large fraction of the room without perceiving abrupt changes in sound and "image").
BTW: I've just tried to listen to 3D3A Lab "PureStereo" (which basically is yet another form of XTC) demo video on my setup. Curiously
(but maybe not: it could even be sort of an indirect proof of my hypothesis), while it works pretty well e.g. on a laptop, with my setup it is completely screwed up... making some measurements to try to figure out what's goin' on is something that I too was very curious about. Unfortunately I do not have the required equipment (let alone the expertize) to do so. Should get someone who does to do the measurements, someday...
Last edited:
You are right. A closer look made me aware of that too.
Better "control" of reflections had always two components in Linkwitz' work. There is the "identical tonality" aspect, but there is the increased direct/reflected ratio too. In fact you don't get the first in a dipole without the second. Only exception in SLs work is the Watson omni, where he has always emphasized the need of near field listening to improve the direct/reflected ratio.
I don't see where SL has ever separated both aspects and looked at them independently. So how can he state any priority of the identical off-axis response over the increased d/r ratio?
Does anybody know of any research which has compared the influence of reflections with different "quality" - meaning more or less truth to the original sound?
interesting. This is the same thing that Filippo "audiofanatic" (the author of the periakusmas as well as many other speakers design) was telling me a few years ago...
Last edited:
He doesn't seem to prioritizes off-axis response over D/R but claims that a couple of conditions need to be met: Sound_field_control_for_stereo
I don't want to sound disrespectful but unfortunately he presents nothing more than an educated guess. Real research is needed.
Nothing conclusive with regards to what we are interested in.
g
Obviously, when two (or more) correlated physical sound sources are emitting in a common space you don't get two independent sound fields, but just one which is the result of their combination. Now, what if the physical interactions of the two correlated sound sources in a confined space, when the proper conditions (distances, relative amplitudes, phase, ...) are met, creates e.g. something similar to what the ambiophonics "RACE" processing does (that is XTC), and perhaps not just for one small spot?
I've cited ambio because I've tried it (using RACE and its recommended speaker setup) and have got results which are quite similar to what I've got with my conventional setup with only very careful "sweet spot" speaker positioning.
no surprise because - I believe - it's all about lowering the interaural cross correlation (IACC) by means either of some CTC or a certain pattern of the reflected sound
IMHO the requirement is that the shape of the two transients/wavefronts - direct and reflected - have to be essentially similar - the critical frequency range is around 400<4000 Hz
Hi Markus, Hi all ...
this kind of discussion seems endless, nice
I had some closer listening to the "Denon anechoic" recordings some weeks
ago in my listening room, current reverberation characteristics can be found here:
http://www.dipol-audio.de/model2-messungen-dateien/model2_listeningroom_t30.JPG
Interesting that an orchestra recorded in an anechoic chamber, still sounds
"like that" when reproduced in my rather "wet" listening room:
It sounds very dry.
The reverberation time in a "common" concert hall, is much longer than in a usual
"living room like" listening room, which is what i try to achieve here.
While my listening room being rather "wet", i usually sit rather close to the speakers,
often 2,2 - 2,4 meters.
Listening to such a strange recording, i would have expected the listening room would
become more audible, say at least "auditively assessable in dimensions", but in my case
it doesn't.
It sounds like sitting in front of an anechoic chamber ... i could even imagine
sitting "behind" the orchestra very close during a usual concert event while
the direct sounds of the instruments are dominating any reverb coming back from
the concert hall.
But there is still no "they are here", there is "i am there" and that "there" is
definitely something different than my listening room.
Some kind of "reverb" comes from the - especially huge bodied - string instruments:
You can hear the decay of the contrabass' body e.g. under these conditions and
it produces some -faint- kind of "aural substitute" for the reverb we are accustomed to
in a hall. Other instruments like transverse flute do not produce decay subjectively:
The sounds stop almost immediately.
Some instruments are only changed slightly in tonality:
You can hear the noise of the bow on the string instruments in a very detailed
manner of course and one seems to hear more from the "string" itself and somewhat
less from the instrument's body than usual, but the basic character of the
string instruments remains.
Brass is very bright and crisp, but still instruments are identifyable easily.
Kettle drums are changed very much: No bass subjectively. It takes some time to
adapt to hear it is still there, but the dominating impression is just a "peng"
without the "boom". But as a reward you can follow the decay of the drumskin / kettle
system easily and observe the change in timbre while the sound decays.
____________________
Is the fact, that i do not hear my listening room with that kind of recording
a kind of "proof" that my speaker/room system works rather well ?
I think it may be an indicator at least.
Do i obey the rules listed by SL mentioned above ? No.
"The loudspeakers must have a polar response, such that on-axis and off-axis frequency responses are
identical and may only differ in level. Omni, dipole, cardioid or other frequency independent radiation
patterns are required."
Some thoughts on that:
- For direct and reflected sound to have "same timbre" as also stated by SL, CD is not sufficient anyway:
In addition also the absorbtion over frequencies would have to be flat in the listening room.
Typically this is not the case.
- "Directivity" and polar response is something taking place above the Schroeder Frequency of the
listening room. Below the Schroeder Frequency there is no "direction" or free propagation of
soundwaves, there are only room modes. Surely the LF polar pattern of the speakers used will
a f f e c t the way room modes are excited. But inside the room there is no directivity pattern
present at LF which could be measured or experienced.
> Bass below Schroeder Frequency has it's own rules, in which way a listening room may become
"psychoacoustically transparent" due to say a stereophonic reproduction, if this is the goal
to be formulated.
- Without going into details there is strong evidence IMO, that spatial hearing changes
the "preferred strategy mix" around/above say 2Khz (add the word "continouusly" if you like ...)
- Especially lateralization, which is an important aspect of "imageing in natural hearing",
seems to rely more on interaural phase difference at low to mid frequencies but more on
other cues like
- interaural level differences
- interaural time differences (due to envelope detected ...)
- Head Related Transfer Function
- ... ?
at higher frequencies.
Natural Hearing: Distance between ears does not allow estimation of lateral angle of sound source
by comparison of interaural phase solely above 2Khz: Result gets ambigue.
Stereo System: Cannot deliver proper phase information for lateralization anyhow at high frequencies,
due to interference between left and right speakers. Fixation of listener exactly in the median plane
is not realistic for that purpose, especially for "home listening settings".
Intensity cues are the ones to rely on and even those get hard to transmit via stereo robustly at HF.
Without low to mid frequencies it can be doubted, whether stereo works at all.
IMO you are better off, making a speaker in such a way that it does not confuse lateralization
cues detected on a recording by the listener relying mostly on low to mid freqauencies < 2Khz.
There is no evidence to me, that a uniform directivity pattern is necessary over the entire
audible bandwith. In contrast to that, there is strong evidence to me, that regarding
- room acoustics
- human spatial hearing, especially lateralization and
- properties of the stereophonic system
there are at least 3 frequency bands to be identified, where different aspects apply in order to
make a listening room "psychoacoustically transparent" in case of reproduction using stereo speakers:
- Modal range: Lowest audible frequency ... to Schroeder Frequency,
which for many living rooms may be within 20 ... 120 Hz
- "Interaural phase sensitive" frequency range due to lateralization
which may be within 120 ... 2Khz
- "HRTF" and "interaural evelope" sensitive frequency range which may be within
2Khz ... 20Khz (for the youngest among us).
There is no need for assuming a "one pattern fits all conditions" kind of strategy being right.
For each frequency band a different strategy may be chosen, regarding
- directivity pattern, number and placement of sources
- room treatment (absorbtion, diffusion: slight, heavy, none)
- listening position and distance from the sources
Surely in one certain installation we can ask:
- Are there valid strategies applied for each frequency range under question ?
- Do the specific strategies blend ?
Last edited:
"Lateralization" is a reserved word in psychoacoustics. It's solely used in the context of headphone playback.
I'm currently rereading Theile's thesis. 33 years have passed and it seems we still don't know much more:
http://www.hauptmikrofon.de/theile/ON_THE_LOCALISATION_english.pdf
http://www.hauptmikrofon.de/theile/UEBER_DIE_LOKALISATION_deutsch.pdf (German version)
I'm currently rereading Theile's thesis. 33 years have passed and it seems we still don't know much more:
http://www.hauptmikrofon.de/theile/ON_THE_LOCALISATION_english.pdf
http://www.hauptmikrofon.de/theile/UEBER_DIE_LOKALISATION_deutsch.pdf (German version)