There are few true solutions to this problem.
Here's one: keep resonances outisde of the passband of the driver in the given enclosure. That may be tricky when it comes to the air cavity resonances and backwave reflections...
Good question. I think it is more than a couple of degrees. It also depends on the arrival direction and whether a horizontal or vertical seperation.
Some studies have looked at minimum angular seperation for distinguishing between two sources. Here is one study that found a minimum of 20 degrees to seperate two sounds.
Hollander - Thesis Chapter 8.
Here is one that found a just noticeable seperation of a few degrees.
http://przyrbwn.icm.edu.pl/APP/PDF/121/a121z1ap02.pdf
The real question is at what angle does coloration stop and only spatial properties are perceived. I know that spaciousness doesn't become significant until sounds start to arrive from 30 degrees off the centerline and the effect maximises for 60 to 90 degrees (these relative to frontal sound). Does that mean that for angles of low spaciousness there will necessarily be high coloration?
David S
I had a single strong reflection in my listening room arriving some 20 ms after the direct sound. This was a reflection from the Right channel speaker only, and it arrived from the rear left corner of the room. Even though the delay of the reflection was fairly large, and the direction it came from almost 180 degrees different, it still sounded like it came from the right channel speaker. My stereo image had a tendency to "hang" over to the right side. When the reflection was removed, the stereo image got better centered.
Are single reflections are much easier to hear than a swarm of reflections? If so, that may be the reason why diffusing the rear wall sounds so good?
Are single reflections are much easier to hear than a swarm of reflections? If so, that may be the reason why diffusing the rear wall sounds so good?
Was just watching television with around 10 degrees between speakers, an empty ceramic vase 1 degree to the right of the right speaker, resonant, and reflecting sound around. I could hear this as a mid harshness and a tearing of the image, with the pipe resonance more of an ambient annoyance. If I move the vase another degree (10cm) the image goes back.
With this separation the image is obviously less easy to corrupt and harder to separate but whenever one speaker produces its own distortion that points to itself, the direction to that speaker is clear.
With this separation the image is obviously less easy to corrupt and harder to separate but whenever one speaker produces its own distortion that points to itself, the direction to that speaker is clear.
It's all about precedence and richness of the reflexions!
http://www.audiophilerecordingstrust.org.uk/articles/pa_msu_franssen.pdf
regarding the localization and angle precision, it's ranging from 2 to 4 degres. Have a look at point G!
http://www.audiophilerecordingstrust.org.uk/articles/pa_msu_rooms1.pdf
http://www.audiophilerecordingstrust.org.uk/articles/pa_msu_franssen.pdf
regarding the localization and angle precision, it's ranging from 2 to 4 degres. Have a look at point G!
http://www.audiophilerecordingstrust.org.uk/articles/pa_msu_rooms1.pdf
Thanks for those papers. they seem very good. I have heard the Franssen effect demonstration and it was very convincing! (Sine wave is feed to two speakers, one gets the turn on transient but then the sound is quickly panned to the second speaker. You remain convinced that the sound always comes from the first.)
I'm still not sure that minimum audible angle equally defines the point where colorations convert to separate sources. I did some searching through Toole and, in spite of all the tests he describes, nothing was exactly on point.
Regards,
David
conclusions of the 2nd paper, point 3:
"early reflections which come from the same direction as the direct sound reinforce the sens of localization of the source"
how does this work with a LEDE type room?
point 8:
"Localization performance for steady sounds improves with increasing spectral density of the source"
I am sure we can find opposite conclusions on other papers, especially about the role of early reflections from ceiling/sides, but Hartmann seems to be the most reliable source in that field.
but we are now really getting side tracked..shame it's a great subject! 🙂
"early reflections which come from the same direction as the direct sound reinforce the sens of localization of the source"
how does this work with a LEDE type room?
point 8:
"Localization performance for steady sounds improves with increasing spectral density of the source"
I am sure we can find opposite conclusions on other papers, especially about the role of early reflections from ceiling/sides, but Hartmann seems to be the most reliable source in that field.
but we are now really getting side tracked..shame it's a great subject! 🙂
Last edited:
No, because we got adaptation. Sit a few minutes in a room and coloration does not matter anymore, the perception of coloration simple vanishes over time.
The real problem of stereo is listener head movements and turning which are too fast for the perception to adapt and neglect. In the direct sound field of a stereo triangle the coloration due the cross talk is the worst with two heavily correlated sound sources.
- Elias
Check also out series III of Rakerd & Hartmann paper which deals with onsets and durations:
http://www.audiophilerecordingstrust.org.uk/articles/pa_msu_rooms3.pdf
They got really interesting indicatations of integration times of precedence effect being much larger, up to ten fold, than previously known 1ms.
This indicates that room reflections indeed can improve localisation 😀
- Elias
I strongly disagree with you. Every time I try to listen in a regular room I get annoyed by the coloration, and for me it just gets worse over time as I "listen in" to how the room sounds.
My preference is a room that does not have its own sound. Impossible it might be, but even so ...
Exact same problem with my room. I have to sit right back near the rear wall and the right hand rear corner is pretty much bare, the reflection from this corner (I don't know whether it originates from the left speaker or right) causes the stereo image to be pulled across to the right significantly even though it still localises as coming from the front...it also becomes more diffuse and indistinct.
Adding a bit of damping in the right rear corner such as a bean bag balanced on the corner of the sofa makes quite a dramatic difference - the stereo image returns to the centre, and imaging and definition improves quite dramatically.
If you're forced to sit quite close to the rear of the room, it seems that a certain amount of absorption and/or diffusion at the rear of the room especially the corners is necessary for best imaging...
But this, in itself, is a form of adaption, or perhaps, an example of resistance to it. Early on, I learned to ignore the room and listen to the speakers (job requirement). But neither of us are "average listeners", so our data points don't affect the truism Elias presented. 🙂
Since I have a new room being built, how about having the woofers mounted in the wall cavity open the the outside?
Starting my dipole project I noticed that increased SPL made reflections pretty messy. My room is not too bad and the walls are damping waves to a reasonable level. System is fairly directional. Push it and everything begins to fall apart. I got proactive at the rear of the baffle. For now that is FG against the baffle and a lambs wool sheet ( known as a "wooly" around here. Tufts are around 5/8" thick bonded to a mesh screen) Big positive difference without losing all of the dipole openness. I didn't want to fight with the room ... EQing everything that wasn't nailed down. I do EQ but it's just a little here and there right now. Attenuating the rear wave at the source helped with that. Gets plenty loud and it'll take you wherever you want to go 😛
That is impossible. Our ear and brain can not separate the direct sound from early reflections, so what we will hear is the sum of direct sound and early reflections. The reflections will sound as they are part of the direct sound.
Haas effect - Wikipedia, the free encyclopedia
I suspect Keriwena is probably referring to acclimatisation, not Haas effect, in other words you can get used to a certain frequency response error after a while of exposure both subconsciously and consciously. (we tune out some of the room effects in our mental audio scene) This particularly happens with bass where we can get used to a response with +/- 10dB or more variations within an octave - we're still aware that it's not flat, but we can largely listen through it, and it sounds better after a period of acclimatisation than upon first listening to it. (probably part of the so called "break-in" effect)
Also, Haas effect is not quite as black and white in operation as it appears. Have a look at the papers linked by lolo and Elias, where they found significant imperfections and limitations in the Haas effect.
Last edited: