What is the ideal directivity pattern for stereo speakers?
I think most industry professionals would agree about many criteria that constitute a good or maybe 'perfect' loudspeaker. On another forum speaker dave (who is also a diyaudio member) posted the following list (for the complete list, click on the link):
The Goals for an "Ideal Loudspeaker" - The Classic Speaker Pages Discussion Forums - Page 8
An Ideal speaker has:
1) Very flat on-axis response
2) Very smooth on axis response
3) Very flat and smooth response through any likely listening window
4) Smooth and resonance free power response, but of no particular curve
5) Holes in the power response are acceptable but peaks are not
6) Generally rising directivity (non flat d.i.)
7) Generally wide dispersion
8) Wide bandwidth with a -10dB cutoff below 35 Hz
9) Interfaces well with the room, gives a smooth in-room curve below 200Hz
10) Adequately low distortion. Low AM distortion with high woofer excursion
Most criteria will probably be on the list of most other pro's as well, but I think not everybody would agree about point 7. Opinions seem to vary quite a lot when it comes to what is the ideal directivity.
What would the dispersion pattern of a theoretically ideal speaker look like?
Please, psycho-acoustics first! The practical implementation can be saved for later.
I think most industry professionals would agree about many criteria that constitute a good or maybe 'perfect' loudspeaker. On another forum speaker dave (who is also a diyaudio member) posted the following list (for the complete list, click on the link):
The Goals for an "Ideal Loudspeaker" - The Classic Speaker Pages Discussion Forums - Page 8
An Ideal speaker has:
1) Very flat on-axis response
2) Very smooth on axis response
3) Very flat and smooth response through any likely listening window
4) Smooth and resonance free power response, but of no particular curve
5) Holes in the power response are acceptable but peaks are not
6) Generally rising directivity (non flat d.i.)
7) Generally wide dispersion
8) Wide bandwidth with a -10dB cutoff below 35 Hz
9) Interfaces well with the room, gives a smooth in-room curve below 200Hz
10) Adequately low distortion. Low AM distortion with high woofer excursion
Most criteria will probably be on the list of most other pro's as well, but I think not everybody would agree about point 7. Opinions seem to vary quite a lot when it comes to what is the ideal directivity.
What would the dispersion pattern of a theoretically ideal speaker look like?
Please, psycho-acoustics first! The practical implementation can be saved for later.
In principle there are 2 ideal directivity patterns: mono-directional and omni-directional. All other states are involves a acceptable level of compromise.
Mono-directional is not achievable by loudspeakers as such but are with headphones where your ears will only recieve the direct radiation of the loudspeakers.
Omni-directional speakers do exist but most are poor implementations that doesn't deliver an acceptable level of sound quality.
Bi-polar and to a lesser extent di-polar speakers delivers an directivity pattern in the far field that closely resemble a true omni-directional speaker, and that's the reason behind the DIY popularity of di-polar speakers. Strangely, bi-polar speakers aren't nearly as popular although it has some very significant advantages over di-polar speakers but that is probably due to lack of knowledge to some extent, and that it incurs twice the cost.
Mono-directional is not achievable by loudspeakers as such but are with headphones where your ears will only recieve the direct radiation of the loudspeakers.
Omni-directional speakers do exist but most are poor implementations that doesn't deliver an acceptable level of sound quality.
Bi-polar and to a lesser extent di-polar speakers delivers an directivity pattern in the far field that closely resemble a true omni-directional speaker, and that's the reason behind the DIY popularity of di-polar speakers. Strangely, bi-polar speakers aren't nearly as popular although it has some very significant advantages over di-polar speakers but that is probably due to lack of knowledge to some extent, and that it incurs twice the cost.
Last edited:
I think especially for 4) and 7) it is a bit difficult to choose independently
from the listening room.
Having a "good" listening room with high diffusivity, relatively low
decay time, no spectral anomalies in decay time (say smoothly falling
with frequency or constant above say 300Hz) a speaker designed
according to that criteria will surely perform very well, if listening
distance is not too far.
When i was visiting an audio fair some years ago using empty
hotel rooms for presentation the "Amphion" models from Finland
presented in that year had a clear advantage over the "wide"
but also often "discontinouus directivity" conventional multiway
speakers presented under similar conditions.
Amphion uses a resistance box for the midrange in some of their Models,
achieving a cardioid like radiaton pattern, and waveguides for the tweeter
to match that pattern to some extent at the crossover frequency.
But that pattern could still be considered as "fairly" wide, as there seems
to be no big change for small off axis angles ... as long as you have that and
tonality is consistent even for somewhat larger angles even such
a pattern may work well, and surely better under poor conditions.
I was referring to horizontal dispersion, if the vertical dispersion
can be made fairly consistent it may be considerably narrower than
the horizontal dispersion without any disadvantage IMO.
Kind Regards
from the listening room.
Having a "good" listening room with high diffusivity, relatively low
decay time, no spectral anomalies in decay time (say smoothly falling
with frequency or constant above say 300Hz) a speaker designed
according to that criteria will surely perform very well, if listening
distance is not too far.
When i was visiting an audio fair some years ago using empty
hotel rooms for presentation the "Amphion" models from Finland
presented in that year had a clear advantage over the "wide"
but also often "discontinouus directivity" conventional multiway
speakers presented under similar conditions.
Amphion uses a resistance box for the midrange in some of their Models,
achieving a cardioid like radiaton pattern, and waveguides for the tweeter
to match that pattern to some extent at the crossover frequency.
But that pattern could still be considered as "fairly" wide, as there seems
to be no big change for small off axis angles ... as long as you have that and
tonality is consistent even for somewhat larger angles even such
a pattern may work well, and surely better under poor conditions.
I was referring to horizontal dispersion, if the vertical dispersion
can be made fairly consistent it may be considerably narrower than
the horizontal dispersion without any disadvantage IMO.
Kind Regards
Last edited:
Member
Joined 2009
Talking about the dispersion pattern might still be too implementation specific. We might discover that "one CD pattern fits all" doesn't necessarily exist.
I think that first and foremost we need to map out the arrival of the ideal signal to the listener. I've heard that early reflections are good and that late reflections are bad but the exact time range, intensity and frequency are subject of speculations. I'm not even convinced that we need to know what is ideal, just what is not desired.
I would emphasize that once we have a solid understanding of the requirements, the technical details of the delivery method will be more or less trivial.
I think that first and foremost we need to map out the arrival of the ideal signal to the listener. I've heard that early reflections are good and that late reflections are bad but the exact time range, intensity and frequency are subject of speculations. I'm not even convinced that we need to know what is ideal, just what is not desired.
I would emphasize that once we have a solid understanding of the requirements, the technical details of the delivery method will be more or less trivial.
In a home environment the goal is to have delayed reflections as replicas of direct sound. Hence these would qualify:
- Omni-directional
- Dipole
- Cardioid
Considering that the lower the frequency, the less sensitive the perception is as to direct vs. reverberant sound, a speaker with narrow directivity (aka waveguides) may qualify too.
Each would have their own limitations and challenges, ie. there is no perfect dipole speakers or perfect Omni speakers.
- Omni-directional
- Dipole
- Cardioid
Considering that the lower the frequency, the less sensitive the perception is as to direct vs. reverberant sound, a speaker with narrow directivity (aka waveguides) may qualify too.
Each would have their own limitations and challenges, ie. there is no perfect dipole speakers or perfect Omni speakers.
Linearray, then how would you explain positive reviews of all kinds of speakers, from omni to horn radiation patterns. It appears to be a matter of taste, not something that has correct way of doing. How many people can tell a difference between omni and waveguide speakers in a blind test?
remember this long thread:
http://www.diyaudio.com/forums/multi-way/177403-linkwitz-orions-beaten-behringer-what.html
speakers had different radiation patterns, none of the speakers had decisive victory.
p.s. I understand importance of directivity in PA applications( night clubs, stadiums etc.)
remember this long thread:
http://www.diyaudio.com/forums/multi-way/177403-linkwitz-orions-beaten-behringer-what.html
speakers had different radiation patterns, none of the speakers had decisive victory.
p.s. I understand importance of directivity in PA applications( night clubs, stadiums etc.)
My recipe these days is for mid and hf horns with tightly controlled directivity oriented relatively on axis at my listening position. (60 degrees or less of dispersion in the horizontal plane, and 40 - 60 degrees in the vertical plane.. This works well in my room - good image and good uniform measured frequency response at my listening position. Probably not optimum for larger rooms or where a very wide listening area is desired, but it works for me.. 
Everyone needs to find their own path of course, for a long time I was quite enthralled with the MG1.6QRs I owned - these days and in this room, unlikely...
Good subject..
Everyone needs to find their own path of course, for a long time I was quite enthralled with the MG1.6QRs I owned - these days and in this room, unlikely...
Good subject..
No early reflections, i.e. none off any boundary until behind the listening position, ergo would be different in virtually every multi-use room.
GM
I know that to me there is a big difference between e.g. omni and
cardioid especially under poor conditions (long decay).
If you want something to work under any (home listening) conditions, IMO
you first have to make it work under poor conditions.
There are not so many rooms, which allow an omni source to be used
IMO, except for nearfield listening maybe.
In well behaved rooms and not too far listening distance, i have no
problems with omni sources either.
Sitting at one seat i cannot identify the type of radiation pattern
as such, i think. How could i ?
But under given (especially poor) conditions the narrower patterns
sound less "reverberant".
Last edited:
How to explain positive reviews of any kind of radiation pattern ?
1) Different types of listeners. Listeners are not the same.
I would distinguish at least two archetypical types of listeners,
without value judgement.
"Critically motivated listeners" tend to judge in terms of
"similarity" or "correctness" of a reproduced sound event
either from direct comparison (if possible) or from memorizing
similar events. They compare in terms of credibility and
similarity.
Of course this type of listener is somewhat artificial or at least
"culturally new", since he did not exist in the times before
''recording" was invented.
"Aestetically motivated listeners" basically do not
distinguish between "production" and "reproduction".
Musicians playing acoustic Instruments e.g. are quite often
of the latter type.
Of course intermediate states between those archetypes
exist and conditions play a role too. There are also persons
able to "switch modes" if needed.
2) Different conditions.
We all know there are "dry" and "wet" rooms, and most
rooms have special "oddities".
3) Different kinds of programme material.
An exceptionally "dry" recording of strings reproduced in a
compatible room ... may be "OK" from an aestethical point of view.
A spacious (and credible) recording reproduced in a "wet" room, which is
too small ... may be horrible from both mentioned points of view.
That last point is the reason for me to avoid speakers with DI too low,
especially under reverberant conditions.
1) Different types of listeners. Listeners are not the same.
I would distinguish at least two archetypical types of listeners,
without value judgement.
"Critically motivated listeners" tend to judge in terms of
"similarity" or "correctness" of a reproduced sound event
either from direct comparison (if possible) or from memorizing
similar events. They compare in terms of credibility and
similarity.
Of course this type of listener is somewhat artificial or at least
"culturally new", since he did not exist in the times before
''recording" was invented.
"Aestetically motivated listeners" basically do not
distinguish between "production" and "reproduction".
Musicians playing acoustic Instruments e.g. are quite often
of the latter type.
Of course intermediate states between those archetypes
exist and conditions play a role too. There are also persons
able to "switch modes" if needed.
2) Different conditions.
We all know there are "dry" and "wet" rooms, and most
rooms have special "oddities".
3) Different kinds of programme material.
An exceptionally "dry" recording of strings reproduced in a
compatible room ... may be "OK" from an aestethical point of view.
A spacious (and credible) recording reproduced in a "wet" room, which is
too small ... may be horrible from both mentioned points of view.
That last point is the reason for me to avoid speakers with DI too low,
especially under reverberant conditions.
Last edited:
It a blind test it would be difficult if not impossible to identify the dispersion pattern but there would be a huge difference between them. You would just have problem identifying which is which without knowing the sound of each type in that room beforehand.
Cardiod dispersion speakers, that means regular normal speakers with speakers on only one side, no matter what type of speakers are used and whatever waveguide might be used will only produce complete reverberant sound at the listening position of the portion of the frequency range which is omni-directional, ie. the lowest frequencies.
Omni-directional speakers, be it true omni-directional or figure-8 directional like di- or bi-polar speakers, will make complete reverberant sound of all frequencies. Note however that unlike bi-polar speakers, di-polar speakers necessitates extremely careful positioning and listening room conditioning because the output from the back side is in opposite phase. Bi-polar speakers does not have such problems and can basically be placed almost anywhere in the room and will produce the same sound.
You missed the more fundamental divide between biased and unbiased listeners
therefore You forgot about audiophile-type listener, as Linkwitz put it:
because they are both
OTOH there are various kinds of audiophile-type listeners:
You missed the more fundamental divide between typical (normal) living-room and abnormal audiophile listening-room
normal living-rooms are pretty much standardized (architectural standards plus decor trends), not especially wet nor dry, RT60 0.2-0.5
but yes audiophile listening-rooms have their special "oddities" being home-made attemtps at creating studio room or anaechoic chamber
yes, I agree, a typical bathroom is useless as a listening room
Last edited:
Which is quite a range, 0.2 is in the recommended control room range
and 0.5 is already pretty "wet" for a listening room.
"Oddities" like
- modal density in the bass not continouusly rising with frequency, but with "gaps", due to
the room's proportions
- longer decay than in neighboured bands occuring e.g. somewhere in the midrange
are not restricted to "audiophile" environment. It is what's happening in
usual living rooms.
...
yes, I agree, a typical bathroom is useless as a listening room
...
At least here seems to be some agreement ...
Last edited:
to be more precise 0.2 is just a value I have found on one of audiophile pages
on the other hand on another page I have found that 0.5 is the "optimum midfrequency RT60" for a "broadcast studio"
http://www.vividaudio.com/downloads/Acoustics_an_introduction.pdf
0.5 is also typical/recommended RT60 for small rooms according to KTH handbook:
TMH KTH :: Music - Acoustics for violin and guitar makers
which relies in that regard on "textbook knowledge" and results of a KTH research project "Speech and Music in Rooms"
for bigger rooms like 6x8 classroom size room it is 0.8
yes, sure, but has nothing to do with all this "early relfections destroy imaging" science, has it?
yeah!
I would say early reflections tend to blow up phontom image sizes
(smear them spatially) and they may affect tonality.
Especially the performance of speakers having "steps" in frequency
dependent directivity suffers from standing too close to reflecting side
walls e.g.
Speakers with more continouus directivity (and higher directivity in general)
tend to perform more uniformly in tonality, when placed in differerent environments.
Last edited:
Which in a round robin way brings us back to the original statement I made in the very first post, that there is in principle 2 ideal directivity patterns, omni-directional and mono directional. Either all room resonances must be excited or none of them (or the total absence of room resonances). Everything in between is some level of acceptable compromise.
Remember the original poster asked for the theoretical ideal directivity pattern, not for a practical implementation.
using "continouus" directivity i meant "only smoothly variing
with frequency", maybe that was misunderstandable.
Whether the polar patterns should be falling off sharply ?
In the likely listening windows * there should be
little change, and the pattern should fall off monotonic
without lobing.
I think a cardioid would be my base pattern in the
horizontal plane.
You can morph it into an omni or a dipole gradually depending
on the environment.
___________
* deliberately borrowed from David S.
with frequency", maybe that was misunderstandable.
Whether the polar patterns should be falling off sharply ?
In the likely listening windows * there should be
little change, and the pattern should fall off monotonic
without lobing.
I think a cardioid would be my base pattern in the
horizontal plane.
You can morph it into an omni or a dipole gradually depending
on the environment.
___________
* deliberately borrowed from David S.
Last edited:
...tend to bring spaciousness - "images that fill a space (ASW)" (as opposed to pin points) which is a thing many consider necessary for realism of sound reproduction, for example Markus76 here, on diyaudio forum?
yes, perhaps some early reflections may affect tonality but do You know any detailed study of this question - of which reflections (with regard to their delay, angle, frequency response etc.), how and to what degree?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.