I don't think constant directivity sounds better per se, for one listening position and one particular set-up, but I think it can give a greater degree of freedom to tune the room response with toe in that can't be easily achieved with some other types.
An ideal constant directivity speaker (approached reasonably closely by a large CD Waveguide above 1Khz) would have response that varies very little within it's CD window, (far less than a cone+dome) which means the direct signal will stay very uniform as the toe in is adjusted within a certain range, as well as the direct response heard by different side by side listeners in different listening spots being more uniform. More consistent direct response over the expected range of listener angles I think is a benefit.
On the other hand the early ipsilateral wall reflection will be either just within or just outside the CD angle depending on toe in, so by moderate adjustment of toe in significant changes to the amplitude of this early reflection can be effected with little or no change in spectral balance of the reflection, or the direct path signal.
I think possibly the biggest advantage of a constant directivity approach though, (and also applicable to just "directional" systems in general, they don't have to be CD) is not so much what it does in the room, but what it does on the baffle.
Constraining the natural directivity of the drivers to less than 180 degrees over as much as their range as possible dramatically reduces the effects of cabinet edge diffraction, probably more so than rounded edges. I think this is of real benefit to imaging and sound quality in general.
CD allows you to do this in a way where the directivity is wide enough to give good coverage of the expected listening area, but not so wide that the signal is diffracting off the cabinet edges, and to maintain this balanced situation over a wide range of frequencies. Perhaps this is what is valuable.
Asymmetry is one of the most detrimental factors on imaging to me. The little board I've installed does tremendously improve localization within the left side of the sound stage.
There´s one more thing that does not make sence: If we can hear the direction of the reflections, we would not be able to pin point a soundsource. Because the reflections come from everywhere, the sound must also be comming from everywhere (if we can determine the location of reflections). But I clearly can determine the location of a sound, even in the most reflective rooms.
David,
That's a great paper indeed! I take another look at it every once in a while too. However, although this test is very valuable, it is not definitive. There apparently was no CD design in Toole's test. Who's to say how it would have performed if it had the same direct sound as the highest ranked speaker in this test and it had no dip in the DI, and/or maybe a slightly higher overall directivity?
Most of us would agree a flat or at least smooth direct sound is a very important factor. Probably the next most important factor is directivity. The speakers in Toole's test were all quite conventional speakers with relatively similar directivity. I'm sure Harman has by now done those same tests with many different kinds of speakers (they have all the facilities, like for example their fancy speaker shuffler). I just wish they'd disclose the results of those tests.
That's a great paper indeed! I take another look at it every once in a while too. However, although this test is very valuable, it is not definitive. There apparently was no CD design in Toole's test. Who's to say how it would have performed if it had the same direct sound as the highest ranked speaker in this test and it had no dip in the DI, and/or maybe a slightly higher overall directivity?
Most of us would agree a flat or at least smooth direct sound is a very important factor. Probably the next most important factor is directivity. The speakers in Toole's test were all quite conventional speakers with relatively similar directivity. I'm sure Harman has by now done those same tests with many different kinds of speakers (they have all the facilities, like for example their fancy speaker shuffler). I just wish they'd disclose the results of those tests.
This is all covered in section 8a of the article I linked (which was linked by someone else earlier in the thread):
"The maximum possible interaural time difference that can occur due to a sound source's move around the head (assuming an average head diameter of ~0.22m and speed of sound = 345m/s) is ~0.65ms or ~0.00065s.
For a sine signal to take advantage of IPD cues without resulting in ambiguity, its period must be at least twice this value (i.e. 2 x 0.00065s = 0.0013s), corresponding to a wavelength of at least twice the head's diameter (i.e. 2 x 0.22m = 0.44m/cycle).
Therefore, the absolute highest frequency for which IPDs provide useful cues is 1/0.0013 = ~770Hz.
For higher frequencies, single-cycle IPDs are not useful cues because they cannot be interpreted reliably, as they depend not only on sound source location on the horizontal & medial planes but also on frequency and, most importantly distance (widely different sound source locations can result in the same IPDs or the same angular location in the localization coordinate system may result in different IPDs)."
So you're right in your thinking, but wrong in your calculation
Sorry, I should have read those "modules" you recommended before commenting. Those modules are some of the best research I've seen yet on this subject. They seem to have done quite a bit of measuring and graphing. There is one apparent inconsistency though in their last graph, at the bottom of module 2.
I may be confusing myself here, but this graph seems to imply that timing differences may be significant to above 1kHZ, since 1kHZ has a period of about 1 mS. I wonder if Haas effect is not related to perception of timing cues (?).
Reflections don't come from "everywhere". Look at the ETCs that have been posted. Those reflections come from very different locations.
Member
Joined 2009
I ran some tests last night to evaluate different degrees of toe in. The speakers are described in this link - very wide dispersion, flat directivity, -3dB@60deg, -9dB@90deg from 500Hz-8kHz. I setup my speakers for similar early reflections even if that results in slight asymmetry in the physical setup. I find that to be very beneficial to imaging.
On the graphs red = left and green = right. I'm not sure what is the official way to describe toe-in/out so I made up my own. I assume 0 degrees is when the speakers are facing one another. At 90 degrees they will be facing perpendicular to the side walls and at 180 they will be facing the close side wall. I am not entirely sure which peak on the graphs corresponds to what reflection but I've indicated my best guesses based on distances in my room. The speakers are setup close to the narrow walls of the room and there is a wall behind them.
At 30 degrees, the extreme toe in, there were a lot of mirror images, it was hard to pinpoint the location of instruments. The center image was well pronounced and the overall clarity of the sound was OK. The overall sound detail was good and clear but the image had collapsed except for the center.
At 60 degrees things were better and this is my usual listening setup. The side image opened up and the clarity was good. The only thing that I didn't like was that a stereo recording of a piano failed to properly place the low notes on the left and high on the right.
At 90 degrees the image widened and the piano keys were correctly laid out from left to right. Clarity however was not as good as with 60 degrees. I did not like the overall sound quality but I can't quite describe what was wrong.
At 120 degrees the extreme toe out, the image was enormous. It grew the size of the room. It was more defined at the edges and somewhat blurry in the center. Sound clarity was OK but not optimal. Instrument separation was the best in this configuration. It is enjoyable for the short term but I probably won't keep it permanently.
------------
When it comes to personal preference I like 60 and 120 the best but they are different. 30 is bad and 90 is OK but I don't like it.
The image widens when the speakers are toed out towards the close wall and it narrows when they are toed in towards the far wall. Widening and narrowing might not be a correct description because the effect is different than the one when the listener moves closer or further to the speakers. When toeing out speakers the sides of the image become more clear while the center blurs and when toeing in the center improves but the side image deteriorates.
On the graphs red = left and green = right. I'm not sure what is the official way to describe toe-in/out so I made up my own. I assume 0 degrees is when the speakers are facing one another. At 90 degrees they will be facing perpendicular to the side walls and at 180 they will be facing the close side wall. I am not entirely sure which peak on the graphs corresponds to what reflection but I've indicated my best guesses based on distances in my room. The speakers are setup close to the narrow walls of the room and there is a wall behind them.
At 30 degrees, the extreme toe in, there were a lot of mirror images, it was hard to pinpoint the location of instruments. The center image was well pronounced and the overall clarity of the sound was OK. The overall sound detail was good and clear but the image had collapsed except for the center.
At 60 degrees things were better and this is my usual listening setup. The side image opened up and the clarity was good. The only thing that I didn't like was that a stereo recording of a piano failed to properly place the low notes on the left and high on the right.
At 90 degrees the image widened and the piano keys were correctly laid out from left to right. Clarity however was not as good as with 60 degrees. I did not like the overall sound quality but I can't quite describe what was wrong.
At 120 degrees the extreme toe out, the image was enormous. It grew the size of the room. It was more defined at the edges and somewhat blurry in the center. Sound clarity was OK but not optimal. Instrument separation was the best in this configuration. It is enjoyable for the short term but I probably won't keep it permanently.
------------
When it comes to personal preference I like 60 and 120 the best but they are different. 30 is bad and 90 is OK but I don't like it.
The image widens when the speakers are toed out towards the close wall and it narrows when they are toed in towards the far wall. Widening and narrowing might not be a correct description because the effect is different than the one when the listener moves closer or further to the speakers. When toeing out speakers the sides of the image become more clear while the center blurs and when toeing in the center improves but the side image deteriorates.
Isn't this still assuming that the spectral response of the reflection is critical? If power respones can have dips (not necessarily peaks) then we have to assume that at least some reflections can have dips as well.
I believe that reduction of edge effects and cleaning up the direct sound accordingly, are very good (audible) goals, but if your approach to doing this requires higher directivity, aren't you trading response smoothness for spaciousness? You can have wide dispersion and reflection free edges with careful cabinet design. (Design the baffle as a 180 degree CD horn with Keele type end flaring, i.e.radius the edges.)
David S.
That just makes my point:
The the first reflections of the direct sound are all comming from different locations. But there are much much more reflections all comming from different locations. But my experience is that I´m perfectly able to locate the direct sound quite accurately.
If I realy can hear the location of the reflections, I must not be able to pinpoint the direct sound. But I can, and so can my grandma.
Member
Joined 2009
some thoughts
I find that:
1) Long distance to the close sidewall is desirable.
2) Stronger close side wall reflections are preferable to far wall ones for correct image.
3) Reflections intensity should be minimized or eliminated for anything other than side walls when possible.
4) Side wall reflections should maybe come at different times than ceiling, front wall and be not be clumped together on the time scale.
Comments please?
I find that:
1) Long distance to the close sidewall is desirable.
2) Stronger close side wall reflections are preferable to far wall ones for correct image.
3) Reflections intensity should be minimized or eliminated for anything other than side walls when possible.
4) Side wall reflections should maybe come at different times than ceiling, front wall and be not be clumped together on the time scale.
Comments please?
Last edited:
Hi Guys
Science in the service of selling;
I have been following this thread and posted a few times as the stereo image is a keen interest of mine personally. A reoccurring theme here is citing work done by people who definitely have a “horse in the race” and to a degree that would include me as well.
For a moment, then, pretend you know nothing about “how it’s supposed to be” in the home, you never read a hifi magazine and that you have never read anything about desirable room reflections or what produces a satisfying effect.
The problem begins with the recording, while you hear with two ears, two microphones do not “hear” in anywhere the same way, they do not have a brain which processes the input into ONE image. As a result, the various ways of capturing a live stereo image will never fool you into thinking you were actually there; at best they remind you strongly of that experience. When you play any of these recordings through speakers, the NEVER have the spatial details one has when you listen with headphones and that in addition to the “up and over” your head sound field of headphones vs “between the two speakers” of the speaker sound field.
Partly as a result of not having a way to capture live events accurately, the vast majority of the music we play are entirely contrived sound stages, made from many individual microphone inputs (tracks) and then panned into the desired positions. The recording to CD process involves a number of stages where the recording is optimized (or ruined) according to the producer and mix engineers desire and skill. In other words, when you listen at home, you have no idea what it is supposed to sound like and if it a normal multi-track recording meant to be played through speakers, it’s sound field is entirely artificial.
The subject of studio monitors has come up too. I don’t know how many of you have done any recoding at the studio level but it can be a lot of fun if you like the music. It also requires a great deal of art and experience, for example part of it is recognizing the “voice” of each microphone and then choosing the right one and it’s placement for the instrument or task. ALL microphones except instrumentation mics have a “pattern” like a loudspeaker and so how they sound depends on angle etc.
Anyway, so you have your say 24 tracks, each one a mic input in the orchestra and now you begin the composition.
You have two options now, you can use the large , usually partly horn loaded monitors in the soffit in front of you or you can use the small two ways speakers a 3 feet away on your meter bridge. The wall mounted speakers are flush mounted so they DO NOT radiate any sound as a result of edge diffraction etc and the entire wall in front is highly absorptive with special foam on it. A great deal of effort went into NOT producing any close reflections.
Alternately, the near field monitors are right in front of you only 2 or 4 feet away. These have low directivity BUT notice they are over the mixing board and there is a large path between them and any side or rear reflections. In other words, even though they have a low Q (wide dispersion), there is nothing to reflect off of for a large distance. The rear of the room is not absorptive though, it has diffusion which breaks up the direct sound so instead of a reflection, it is scattered. Google RPG diffusers as a sample of what is used in the rear.
In a much earlier post or two I urged people to take speakers outside and set up your stereo with NO reflections and then judge if they help or not.
This is the condition the near field monitors are used in for studio use.
I suppose if you had a large room it would work indoors too if the speakers and you were very close like 3 or 4 feet but outdoors you can grill a burger or two.
The object of only hearing the speakers is that when the pan pot for a channel is placed at the center, an identical and equal signal goes to each speaker and with a coherent speaker and minimal other noise (reflected sound) , the brain hears the instrument or voice floating in directly in front of you. With a coherent speaker that doesn’t broadcast it’s own location, one can make a VERY believable voice float dead ahead “like there has to be a speaker hidden there” as a friend said.
In the studio, this makes the placement of each instrument across the left to right panorama an enjoyable task and all that effort in the “live end / dead end” room and such was to make this possible and the Art of doing all this was what the great recordings of the past had. Now days, we have a recording medium which many would have given an arm for 30 or 40 years ago being used to record music compressed into 10 dB or less dynamic range. Maybe a little like a paper company that has gradually developed the best paper ever created while the artists have given up oil paints and have switched to working with feces. I don’t mean to condemn the recording industry but at least the “loudness wars” remind me of the paper situation.
My point is that it appears the approach used in the home is rather different, where it is the pleasing nature of what is heard that is the measure of being faithful to the recording where as the studio approach one wants to hear exactly what is on the recording. Personally, I want to hear exactly what is in the recording at home. If I want to be “surrounded” with a normal recording, I can listen in 5.1 or 7.1 where one can adjust “how much” of this you get..
Fwiw, all this is just how I see it and like I said I am interested in the process. While most of what I do is on the reproducer side of the equation, I am fiddling with a “capture” device for capturing a live stereo image. I believe I have found a new wrinkle in the process.
If you have headphones or better yet good headphones and are interested in the stereo image, try a couple of these recordings and let me know how they sound to you. What I have posted are 2 channel recordings, these are the front two channels of a 5ch 360 degree capture and it should correspond more or less to ones field of vision. I must apologize for the “environmental” nature of the subjects but the mic array is cumbersome to move and real ugly. Also though natural sounds in addition to being available in ones yard, tend not to be harmonic (unlike music) so both even or odd harmonics are more audible. Anyway, try the Harley and trains at the bottom of the page here, be sure to use headphones as most speakers can’t reproduce the image or the dynamic range of the fireworks..
Danley | Technical Downloads
Best,
Tom Danley
Science in the service of selling;
I have been following this thread and posted a few times as the stereo image is a keen interest of mine personally. A reoccurring theme here is citing work done by people who definitely have a “horse in the race” and to a degree that would include me as well.
For a moment, then, pretend you know nothing about “how it’s supposed to be” in the home, you never read a hifi magazine and that you have never read anything about desirable room reflections or what produces a satisfying effect.
The problem begins with the recording, while you hear with two ears, two microphones do not “hear” in anywhere the same way, they do not have a brain which processes the input into ONE image. As a result, the various ways of capturing a live stereo image will never fool you into thinking you were actually there; at best they remind you strongly of that experience. When you play any of these recordings through speakers, the NEVER have the spatial details one has when you listen with headphones and that in addition to the “up and over” your head sound field of headphones vs “between the two speakers” of the speaker sound field.
Partly as a result of not having a way to capture live events accurately, the vast majority of the music we play are entirely contrived sound stages, made from many individual microphone inputs (tracks) and then panned into the desired positions. The recording to CD process involves a number of stages where the recording is optimized (or ruined) according to the producer and mix engineers desire and skill. In other words, when you listen at home, you have no idea what it is supposed to sound like and if it a normal multi-track recording meant to be played through speakers, it’s sound field is entirely artificial.
The subject of studio monitors has come up too. I don’t know how many of you have done any recoding at the studio level but it can be a lot of fun if you like the music. It also requires a great deal of art and experience, for example part of it is recognizing the “voice” of each microphone and then choosing the right one and it’s placement for the instrument or task. ALL microphones except instrumentation mics have a “pattern” like a loudspeaker and so how they sound depends on angle etc.
Anyway, so you have your say 24 tracks, each one a mic input in the orchestra and now you begin the composition.
You have two options now, you can use the large , usually partly horn loaded monitors in the soffit in front of you or you can use the small two ways speakers a 3 feet away on your meter bridge. The wall mounted speakers are flush mounted so they DO NOT radiate any sound as a result of edge diffraction etc and the entire wall in front is highly absorptive with special foam on it. A great deal of effort went into NOT producing any close reflections.
Alternately, the near field monitors are right in front of you only 2 or 4 feet away. These have low directivity BUT notice they are over the mixing board and there is a large path between them and any side or rear reflections. In other words, even though they have a low Q (wide dispersion), there is nothing to reflect off of for a large distance. The rear of the room is not absorptive though, it has diffusion which breaks up the direct sound so instead of a reflection, it is scattered. Google RPG diffusers as a sample of what is used in the rear.
In a much earlier post or two I urged people to take speakers outside and set up your stereo with NO reflections and then judge if they help or not.
This is the condition the near field monitors are used in for studio use.
I suppose if you had a large room it would work indoors too if the speakers and you were very close like 3 or 4 feet but outdoors you can grill a burger or two.
The object of only hearing the speakers is that when the pan pot for a channel is placed at the center, an identical and equal signal goes to each speaker and with a coherent speaker and minimal other noise (reflected sound) , the brain hears the instrument or voice floating in directly in front of you. With a coherent speaker that doesn’t broadcast it’s own location, one can make a VERY believable voice float dead ahead “like there has to be a speaker hidden there” as a friend said.
In the studio, this makes the placement of each instrument across the left to right panorama an enjoyable task and all that effort in the “live end / dead end” room and such was to make this possible and the Art of doing all this was what the great recordings of the past had. Now days, we have a recording medium which many would have given an arm for 30 or 40 years ago being used to record music compressed into 10 dB or less dynamic range. Maybe a little like a paper company that has gradually developed the best paper ever created while the artists have given up oil paints and have switched to working with feces. I don’t mean to condemn the recording industry but at least the “loudness wars” remind me of the paper situation.
My point is that it appears the approach used in the home is rather different, where it is the pleasing nature of what is heard that is the measure of being faithful to the recording where as the studio approach one wants to hear exactly what is on the recording. Personally, I want to hear exactly what is in the recording at home. If I want to be “surrounded” with a normal recording, I can listen in 5.1 or 7.1 where one can adjust “how much” of this you get..
Fwiw, all this is just how I see it and like I said I am interested in the process. While most of what I do is on the reproducer side of the equation, I am fiddling with a “capture” device for capturing a live stereo image. I believe I have found a new wrinkle in the process.
If you have headphones or better yet good headphones and are interested in the stereo image, try a couple of these recordings and let me know how they sound to you. What I have posted are 2 channel recordings, these are the front two channels of a 5ch 360 degree capture and it should correspond more or less to ones field of vision. I must apologize for the “environmental” nature of the subjects but the mic array is cumbersome to move and real ugly. Also though natural sounds in addition to being available in ones yard, tend not to be harmonic (unlike music) so both even or odd harmonics are more audible. Anyway, try the Harley and trains at the bottom of the page here, be sure to use headphones as most speakers can’t reproduce the image or the dynamic range of the fireworks..
Danley | Technical Downloads
Best,
Tom Danley
That's because of the precedence effect. You can read about it here: Precedence effect - Wikipedia, the free encyclopedia
But just because the precedence effect "filters out" the direct sound, this doesn't mean that reflections have no effect on intelligebility, tonality, spaciousness, etc.
It's call the precedence effect or Haas effect
Check out one of the many references to it that have been linked to through this thread.
What's different about the different reflections ? The direct sound arrives first and all the rest are copies (more or less) of the original. That is how your brain knows which one is the real sound source direction.
That doesn't mean you don't hear the reflections though, just that you don't necessarily perceive their direction. If there is no direct path and the only source of sound is a reflection (such as listening to noise outdoors through a 45 degree open window) then the reflection becomes the apparent source of the sound, as it is now the first arrival, even though it's physically a reflection.
Try this little test to see how time delay of duplicated sounds is localized - play some mono music or speech on a normal Stereo speaker set-up, but instead of sitting at the normal equidistant point, sit far to the right so that you're in line with the right speaker. (Same distance from the right wall as the right speaker, roughly)
The sound appears to only come from the right speaker, correct ? You can't hear anything from the left speaker at all ? Now while listening, ask someone to turn the balance control fully to the right to silence the left speaker. You should immediately notice a big difference in the width and character of the sound, even though the localization of the sound hasn't shifted and still appears to be coming from the right speaker.
Last edited:
I've just measured ETC curves of my speakers and room for a comparison. Quite an interesting exercise.
Great to see the ETC curves Markus and Simon. Don't forget that they are "white noise weighted" so the floor bounce that gets well blocked by the rug might actually be fairly strong in the frequency range of interest (the few hundred Hz region). It might be interesting to filter the input at 500 or so and ETC it again.
David S.
Small dips in power response haven't been proven to be bad, I agree, but equally, I'm not sure that lack of power response dips has been proven bad either...(I don't think we "rely" on power response dips at certain key frequencies, as LineArray suggested earlier)
The problem with strong toeing in of a conventional 2/3 way cone/dome type design in an attempt to reduce the early side-wall reflection is that you only succeed in making the dips in the spectral response of the reflection at low power response frequencies even deeper, but without achieving much broadband attenuation of the reflection.
With a CD design you can get a significant broadband reduction of the side wall reflection from the toe in, the fact that its nearly spectrally balanced (in the ideal case) is not really the important factor, the important thing is you're getting a lot of attenuation over a wide frequency range, but with minimal impact on the direct signal to the listener.
I agree, if diffraction was the only concern and you purposely wanted very wide dispersion you could use a 180 degree horn / wave-guide which terminates smoothly at the cabinet edge as you describe, however I think there are benefits to be had in deliberately restricting the directivity to much less than 180 degrees in the vertical plane, if nothing else. (Nobody has yet to show anything good about floor and ceiling reflections)
Although you can get diffraction down to a low level with cabinet design, this would usually mean a narrow and very sculpted cabinet (something like the KEF blade) which is a design approach a bit out of most DIY'ers reach. On the other hand a wave-guide in an ordinary ugly looking shoebox shaped cabinet can have just as low if not lower diffraction with minimal effort, by simply not illuminating the baffle face. If you're willing to use those type of drivers, its far more DIY friendly.
Last edited:
If your speakers are toed-in so as to cross over their axis in front of the listening position, and have the right type of polar response then the image can stay in the middle despite moving the listening position around like this. It won't be as focused an image but will definitely come from the middle. This is because the furthest speaker is now on-axis and sounds louder than the off-axis nearer speaker and this tends to compensate for the earlier time arrival of the nearer speaker sound.
Last edited:
I did that. Ran sweeps from 1-500Hz and 500-20000Hz. The difference is insignificant. That's probably because my front wall is very well damped (20cm batting + 20cm air gap).
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.