I quote you again because there is something interesting, an observation about time stuff.
I'll quickly reiterate that reflection on a boundary is specular, meaning that angle of incidence is same as reflection angle. Answer your question more accurately is that the angle is same to all frequencies, for example ~30deg towards floor and ~45deg towards ceiling and this depends only on geometry of the listening setup in a room, from transducer to ear through a bounce. This is because sound radiates spherically to all directions no matter what the directivity, which does not change how sound radiates as in physics but only towards which direction gets more and which less and at which frequency.
Now, if its a multiway speaker drivers stacked on a baffle then the reflection angles are bit different for all the transducers because they are at different height. This would mean that a woofer that is closer to floor than a tweeter and would hit floor sooner and closer to speaker. Sound from tweeter would hit floor wee bit closer to listener and arrive second as it has little bit longer path length to travel. In this sense one could think that wider directivity woofer hits the boundary first but its not due to directivity but due to geometry. The quote is true, wider incident angle hit room boundaries first, but its not related to coverage angle, directivity. High directivity speaker would just emit less sound towards higher angle.
And at ceiling bounce the setup is reversed, tweeter is now closer to it hitting it first at lower incident angle, arriving first to ear. Imagine mirror on ceiling, the speaker would appear upside down.
And finally here is the observation that occured to me: there is group delay, bass hits the bounce spots always later than highs, even though they always have wider directivity 😀
edit. attached illustrations of imaginary tweet, mid and bass woofer height and their first vertical specular reflection angles.
Few observations from the attached example, take the tweeter and mid: first, there is ~0.2ms difference on path length between them, the other leads through ceiling and the other through floor. In addition to this time "smear" or difference, temporal stuff, the sound could have different frequency response in comparison to direct sound and to each other as the angles are different and due to crossover. Second, floor reflection has a lot shorter path length so its louder than ceiling bounce and comes before, precedence. Million things to take into loudspeaker design work 😉
Last edited:
Agreed. Reflection angles are simple geometry.
Which direction gets more and which less and at which frequency, is what matters in my mind..... all that matters really.
The fact that reflection angles don't change is relatively immaterial.
The ratio of direct to reflected sound varies by frequency at a listening position, largely as a function of directivity .
That ratio determines strength of combing per frequency ........or so it all seems to me.
So I continue to think wider radiation angles have to hit room boundaries first...from a 'what matters' perspective.
group delay is not automatically necessary 😉
Yep, those angles offsets have to have some effect from the vertical arrangement of the drivers...but honestly, seems mainly like analyzing what's wrong with speaker design ....
For me, I say why not just work on moving more towards either controlled directivity point sources, or lines.....
Yes, directivity matters, just wanted to straighten out the reflection points aka bounce does not move further into room with narrower directivity as weltersys let us believe 🙂
Yeah, point source, FIR for group delay and many more things to affect vertical first reflections.
ps. Here is a nugget, latest thing to experiment with from the few past days on this thread and from similar misconception (moving path length, aka moving bounce) also few days ago by Patrick Bateman on another thread, there is Gradient Helsinki mentioned with tilted mid and perhaps the waveguide as well. Got me thinking what the tilting achieves.
Tilting highs towards ceiling could have many implications but it also seems to balance image height, perhaps. More specific, try and keep difference (floor minus ceiling bounce frequency response) constant, avoid this going negative. Its easy to have crossover region so that there is less sound towards ceiling than towards floor at some particular frequency, this will pull the frequency apparently lower down in image making tge sound wee bit confusing. Perhaps even better try get more ceiling bounce and less floor bounce for high frequencies (rug woukd do something like this too), and more low frewuencies towards floor than to ceiling. All this seems to happen quite naturally just by tilting.
Example, look the ER ceiling and ER floor from the two attachments. First one is system simulation before paying attention to these. Wee bit adjusting delay and small tilt (for waveguide tweeter, this a three way system) makes ERDI and DI a little "worse" but balances out the ceiling and floor reflections, more lows to floor and more highs to ceiling and never a zigzag between the two, ~1kHz being the "middle point" when image should be at speaker height. Elongated soundstage height instead of confusing one. Did only short listening test and might be biased but it seems to work, at least in theory and perhaps in action 😉 I suspect this happens with many speakers, symmetric LR filters (usually) have symmetric nulls (same angles both below and above listening axis) but the angles towards ceiling and floor reflection are different, hence possibility for zigzag balance on these around xo. Don't mind the listening window and power responses, they could be equalized more alike/taste, important thing is with the ER ceiling and ER floor lines.
Yeah, point source, FIR for group delay and many more things to affect vertical first reflections.
ps. Here is a nugget, latest thing to experiment with from the few past days on this thread and from similar misconception (moving path length, aka moving bounce) also few days ago by Patrick Bateman on another thread, there is Gradient Helsinki mentioned with tilted mid and perhaps the waveguide as well. Got me thinking what the tilting achieves.
Tilting highs towards ceiling could have many implications but it also seems to balance image height, perhaps. More specific, try and keep difference (floor minus ceiling bounce frequency response) constant, avoid this going negative. Its easy to have crossover region so that there is less sound towards ceiling than towards floor at some particular frequency, this will pull the frequency apparently lower down in image making tge sound wee bit confusing. Perhaps even better try get more ceiling bounce and less floor bounce for high frequencies (rug woukd do something like this too), and more low frewuencies towards floor than to ceiling. All this seems to happen quite naturally just by tilting.
Example, look the ER ceiling and ER floor from the two attachments. First one is system simulation before paying attention to these. Wee bit adjusting delay and small tilt (for waveguide tweeter, this a three way system) makes ERDI and DI a little "worse" but balances out the ceiling and floor reflections, more lows to floor and more highs to ceiling and never a zigzag between the two, ~1kHz being the "middle point" when image should be at speaker height. Elongated soundstage height instead of confusing one. Did only short listening test and might be biased but it seems to work, at least in theory and perhaps in action 😉 I suspect this happens with many speakers, symmetric LR filters (usually) have symmetric nulls (same angles both below and above listening axis) but the angles towards ceiling and floor reflection are different, hence possibility for zigzag balance on these around xo. Don't mind the listening window and power responses, they could be equalized more alike/taste, important thing is with the ER ceiling and ER floor lines.
Attachments
Last edited:
I suspect this has been a mere communication breakdown. I agree with @weltersys that the effective reflection region and hence the nominal reflection point does move further back along the ceiling as DI narrows, and this manifests in a different level heard at the listening position up to the point where it could be said that it has effectively gone over your head.
Duh, I can't get my head around this, why would it be the case? Does the reflection point move again if you tilt/rotate the speaker? How about rotating the reflections out of window.
Amount of reflection changes as directivity increases because there is now less sound towards a reflection point, as well as "bulk" of sound is pointed more towards listening spot when directivity is high, and not towards walls if the speaker is rotated and tilted to do so, but the reflection points stay the same as long as speaker and listener and boundaries stay put.
So, just be carefull about the wording not to lead people to wrong conclusions. All of you have longer track record than I so I agree its just communication breakdown.
Amount of reflection changes as directivity increases because there is now less sound towards a reflection point, as well as "bulk" of sound is pointed more towards listening spot when directivity is high, and not towards walls if the speaker is rotated and tilted to do so, but the reflection points stay the same as long as speaker and listener and boundaries stay put.
So, just be carefull about the wording not to lead people to wrong conclusions. All of you have longer track record than I so I agree its just communication breakdown.
Last edited:
You were the one that assumed we were talking about the listening position reflection point. (Maybe this is getting lost in translation ?) I am not saying reflection point, I am saying effective reflection point and implying I am looking at the speaker and ceiling but the listener is a separate thing.
I think you guys are thinking a transducer coverage angle like beam of flashlight in cartoons, - 6db point defines an edge, a cone of sound? This is nice way to think the coverage angle, but I suggest add red dot to the boundary at specular reflection point in the imagination to get the reflection point in account. It stays put (moves with listener and source) and one lights it more or less with the sound flashlight depending on directivity at frequency in question.
Narrower flashlight would point further back in room than wider one and less to the red dot, but the red dot stays still no matter how wide flashlight beam.
Just keep in mind this is only cartoon and the sound flashlight doesnt have that sharp focus, or edge, like in cartoons 🙂
Narrower flashlight would point further back in room than wider one and less to the red dot, but the red dot stays still no matter how wide flashlight beam.
Just keep in mind this is only cartoon and the sound flashlight doesnt have that sharp focus, or edge, like in cartoons 🙂
Yes, this is what "effective" means in this case. Nothing to do with a red dot.
Waveguides also have a coverage angle, we use that number even though it is not a sharp cutoff.
Eh no, it just does not correlate anyway with reflection location, only changes amount of the reflection at the location, - 6db, - 10db, what ever. This is getting silly.
ps.
At -20db reflection (compared to direct sound) the effect (summed with direct sound) is only +/-1db and around -40db there is practically no sign of it, at observation point. -6db has effect but not that much. Aim past 10db attenuation towards the red dot for audible difference. Any attenuation helps some. I think there is now enough discourse on this for people to learn the stuff on their own so I'm out on the subject.
ps.
At -20db reflection (compared to direct sound) the effect (summed with direct sound) is only +/-1db and around -40db there is practically no sign of it, at observation point. -6db has effect but not that much. Aim past 10db attenuation towards the red dot for audible difference. Any attenuation helps some. I think there is now enough discourse on this for people to learn the stuff on their own so I'm out on the subject.
Last edited:
So the "dispersion " of the direct reflections off the ceiling and the floor will vary in loudness in a manner that will follow the directivity change
affected by the interference in any/most vertical driver arrangement . Can def see the gradient moving with the shape of the directivity in the previous graphs i tested .
Probably many more things that can affect this in different ways.
As for what Allen pointed out, never though that the angle could be enough that the reflections ( the ones that have enough energy to matter )
would just go right over our head in a typical seated and or high ceiling room .
So i'd guess that all loudspeakers with controlled vertical directivity ( controlled or behaved at least ) can achieve that.
Waveguides/horns with shorter vertical dimensions/angle probably achieve that also ?
The floor must behave somewhat differently though .
What happens to waves that hit our legs, the sofa or the floor right in front of our feet ?
Do they diffract and bounce on us ? If so the loudness is probably somewhat attenuated by the diffusion of it all?
lots of questions .sorry folks i'd have 20000 more but i'll try to show some restraint .lol
affected by the interference in any/most vertical driver arrangement . Can def see the gradient moving with the shape of the directivity in the previous graphs i tested .
Probably many more things that can affect this in different ways.
As for what Allen pointed out, never though that the angle could be enough that the reflections ( the ones that have enough energy to matter )
would just go right over our head in a typical seated and or high ceiling room .
So i'd guess that all loudspeakers with controlled vertical directivity ( controlled or behaved at least ) can achieve that.
Waveguides/horns with shorter vertical dimensions/angle probably achieve that also ?
The floor must behave somewhat differently though .
What happens to waves that hit our legs, the sofa or the floor right in front of our feet ?
Do they diffract and bounce on us ? If so the loudness is probably somewhat attenuated by the diffusion of it all?
lots of questions .sorry folks i'd have 20000 more but i'll try to show some restraint .lol
Your 'takeaway' from this discussion IME should be that ideally there will be no reflections between your ears and the sound wall and that this 'collection' will be suitably damped down enough to not reflect back forward, such as with a very randomized/diverse book/knick-knack case that occupies > ~70% of this back wall and for those that need it, has an excellent SAF/whatever rating. 😉
Hehe yeah 🙂
Little bit back to AllenB and JinMTVM posts recent post that yeah sound goes everywhere, over the head, to a first specular reflection point, to ankles, to back wall, all directions. Its the first order reflections and some second order reflections that make into ear before the so called reverberant field starts where there is so much sound coming from all directions they are indistinguishable, the other points in room don't simply matter, they are meaningless in this sense. Well, perhaps reducing reverberation time and what not, but thats a whole lump and not too much location dependent stuff.
First order reflections are the most significant as they have only little path length difference to direct sound so they arrive with very little attenuation and only short delay. Stuff that goes above head or to ankles takes so many reflections before hitting ear they just don't matter much at all, at least not in comparison to the first order reflections.
This is the reason I don't and did not understand the thinking and posts before: -6db point on boundary is simply not relevant unless its the first order reflection point which then hits ear, but it could be -5.9db or - 2db, or -19db, this is what matters, what hits the ear. The nominal dispersion (-6db angle) is very important for the designer and user to know though, to be able to position the speaker so that first order reflections get attenuated, or what ever is the main motivation to select the coverage angle. Its also very useful to realize that directions between first reflection points and ears make mostly to late reverberation so the poo can be directed there.
Well, everyone got their ways of thinking.
Here is tool to get insight, draw a rectangle as side (or top) view of a room, adjust how many orders of reflections to show. https://amcoustics.com/tools/amray
As real time tool very very good for experiments and to get insight. It doesn't show the angles and attenuation in numbers though.
Little bit back to AllenB and JinMTVM posts recent post that yeah sound goes everywhere, over the head, to a first specular reflection point, to ankles, to back wall, all directions. Its the first order reflections and some second order reflections that make into ear before the so called reverberant field starts where there is so much sound coming from all directions they are indistinguishable, the other points in room don't simply matter, they are meaningless in this sense. Well, perhaps reducing reverberation time and what not, but thats a whole lump and not too much location dependent stuff.
First order reflections are the most significant as they have only little path length difference to direct sound so they arrive with very little attenuation and only short delay. Stuff that goes above head or to ankles takes so many reflections before hitting ear they just don't matter much at all, at least not in comparison to the first order reflections.
This is the reason I don't and did not understand the thinking and posts before: -6db point on boundary is simply not relevant unless its the first order reflection point which then hits ear, but it could be -5.9db or - 2db, or -19db, this is what matters, what hits the ear. The nominal dispersion (-6db angle) is very important for the designer and user to know though, to be able to position the speaker so that first order reflections get attenuated, or what ever is the main motivation to select the coverage angle. Its also very useful to realize that directions between first reflection points and ears make mostly to late reverberation so the poo can be directed there.
Well, everyone got their ways of thinking.
Here is tool to get insight, draw a rectangle as side (or top) view of a room, adjust how many orders of reflections to show. https://amcoustics.com/tools/amray
As real time tool very very good for experiments and to get insight. It doesn't show the angles and attenuation in numbers though.
Last edited:
Yeah, seen so many rooms damped like a Sanatorium 'rubber' room at the sound/screen wall when all that was needed was the carpet and a bit of damping at the sides/top 1st reflection before (1st set of) ears/ear height.
The 'review' that tells me all I need to know though is one with any obstructions between the sound wall, ears unless damped with at least a super thick moving van's blanket draped over it.
The 'review' that tells me all I need to know though is one with any obstructions between the sound wall, ears unless damped with at least a super thick moving van's blanket draped over it.
Hehe yeah, or the egg crates on band room walls 🙂 Damping for good measure might work fine or even make things worse, kill nice reverberation (envelopment) and leave bass problems for example. Certainly it wouldn't be very comfortable environment to be long times in, if very anechoic.
There is lot to room acoustics I don't want to go into, but preventing flutter echo and reverberation time for speech intelligibility and all kinds of stuff are very much a good thing when its needed and might demand more stuff on the walls than the first specular reflection points. Handling room modes is another one. Many problems that are somewhat independent but still related as acoustics. There is even some research with mixed results, at least it feels like so, that some prefer some amount of first reflections, others do not and so on. Its a lot of work for anyone trying to tackle problems at home, to figure out what should one do. For typical living room with normal furniture, some flimsy wall(s) and then controlled directivity speakers seems to work fine without much additional acoustic treatment. Not much flutter echo or reverberation problems, only bass modes and perhaps some first reflection stuff going on.
There is lot to room acoustics I don't want to go into, but preventing flutter echo and reverberation time for speech intelligibility and all kinds of stuff are very much a good thing when its needed and might demand more stuff on the walls than the first specular reflection points. Handling room modes is another one. Many problems that are somewhat independent but still related as acoustics. There is even some research with mixed results, at least it feels like so, that some prefer some amount of first reflections, others do not and so on. Its a lot of work for anyone trying to tackle problems at home, to figure out what should one do. For typical living room with normal furniture, some flimsy wall(s) and then controlled directivity speakers seems to work fine without much additional acoustic treatment. Not much flutter echo or reverberation problems, only bass modes and perhaps some first reflection stuff going on.
I don't know which first reflections I prefer, or other people prefer, or how long my room reverberation time is or any of it, how audible the reflections are if they all land on different time, or all at the same time or what? What is important and what is not, how do they affect the sound? How does it sound?
Many interesting and important questions on the subject. All I know is this imagined very good sound with detail and tangible feel to it, this is whats the goal to me no matter the means, and my gut feel is reverberation and first reflections do not help with this, at least when they are a mess. Need some kind of order to it all, or at least enough experience / information to keep on going to right direction achieving the sound. Its a slow process and dead ends and wrong assumptions are certain. Its so mucho fun though and can enjoy music 247 so, would mind never getting there 😀 peace
ps. Hardest part is to connect information to sound. Visuals, text, thoughts all help but don't have a sound. Educated prototyping and experimenting with sound, in the application, takes forward.
Many interesting and important questions on the subject. All I know is this imagined very good sound with detail and tangible feel to it, this is whats the goal to me no matter the means, and my gut feel is reverberation and first reflections do not help with this, at least when they are a mess. Need some kind of order to it all, or at least enough experience / information to keep on going to right direction achieving the sound. Its a slow process and dead ends and wrong assumptions are certain. Its so mucho fun though and can enjoy music 247 so, would mind never getting there 😀 peace
ps. Hardest part is to connect information to sound. Visuals, text, thoughts all help but don't have a sound. Educated prototyping and experimenting with sound, in the application, takes forward.
Last edited:
And you did a very good job straightening it out!
The directivity of a symmetrical exponential horn increases with frequency, increasing the ratio of direct to reflected sound in a room at higher frequencies.
@weltersys are there horn/guides that lessen this effect ? probably a tradeoff somewhere else lol
the emitted energy is always the same globally, it is only distributed differently right ?
this guy popped up on my tube list :
He might be in here right ? really lots of DIY stuff on his tube
so i recall watching the video when he tested and played with the " sound panels " using the " exciter " motors from parts express
I automatically thought about the " pistonic " motion of drivers and when the cones start to breakup with resonances etc.
So those very large panels, the exciter " jiggles " them back and forth, and starts a wave that propagates through the panel at a fixed speed,
then probably resonate around inside the panel to some degree , and the next waves mixes in and it probably gets very messy very fast right ??
I looked into Electro Static Panels when i was younger, like mid 90's iirc ..
The whole panel is driven by the stator so it moves all together ( to some degree at least )
Those panels are not. So This demo setup he did, just feels like everything wrong one can possibly do 🤔
Might be just my noobite assumptions but damn .. can't see this " box" of resonating walls doing anything
more than just tons of distortion and interferences .
What do you guys think ?
the emitted energy is always the same globally, it is only distributed differently right ?
this guy popped up on my tube list :
so i recall watching the video when he tested and played with the " sound panels " using the " exciter " motors from parts express
I automatically thought about the " pistonic " motion of drivers and when the cones start to breakup with resonances etc.
So those very large panels, the exciter " jiggles " them back and forth, and starts a wave that propagates through the panel at a fixed speed,
then probably resonate around inside the panel to some degree , and the next waves mixes in and it probably gets very messy very fast right ??
I looked into Electro Static Panels when i was younger, like mid 90's iirc ..
The whole panel is driven by the stator so it moves all together ( to some degree at least )
Those panels are not. So This demo setup he did, just feels like everything wrong one can possibly do 🤔
Might be just my noobite assumptions but damn .. can't see this " box" of resonating walls doing anything
more than just tons of distortion and interferences .
What do you guys think ?
I'm not understanding. From what I am visualizing, the effective or real reflection point will always be the same, only the ratio of direct sound to sound reflected from that point will change based upon DI. How could it be otherwise?
It seems you are saying that the 2 combined, produce an "effective" reflection point, but I am not sure our brain interprets it that way, the information arriving at slightly different times.
You guys may be talking about a specific reflection, but I've read about our brain responding quite differently to floor, wall, and ceiling reflections. That would only muddy the waters of any perceived "effective" reflection point.
Anyways I'm not sure I'm actually grasping this.
Edit: I see, I should have read further before replying. You guys are not talking about reflections off a floor, from speaker to listener.