You know what, I'm starting to think you made the mistake of thinking in terms of amplification not directivity. Maybe that's why you're confused.
Most of the difference between nude direct radiator and a horn or tall line array comes from the room around them. The ratio of direct vs. reflected sound at certain distance in the room. Narrow dispersion sort of moves you closer to the speaker.
Critical distance Acoustics: Critical Distance
Critical distance Acoustics: Critical Distance
Last edited by a moderator:
Three people have pointed out your are wrong and you are saying they are confused? Inverse square law trumps everything else.
You know bradleypnw, like we're all saying, it ain't so.
Audio flashlights and floodlights both fall 6dB per distance doubling.
Anechoic and outdoors.
And in a room, i think you have it backwards.
Total level in a room is the sum of direct and reflected sound.
A narrow pattern speaker will have relatively less reflected than a wide pattern.
I think reflected sound levels likely vary less with distance than direct sound.
So i think narrow is likely to fall off faster than wide, to our ears and on an RTA which sums both speaker and room.
There's a degree of speculation in my in-room analysis, and I'm sure a lot will depend on the room. But for small rooms, where even the rear wall contributes heavily, it seems to make intuitive sense.
If anybody knows the nuts on in-room, direct to reflected vs distance, pls post.
Audio flashlights and floodlights both fall 6dB per distance doubling.
Anechoic and outdoors.
And in a room, i think you have it backwards.
Total level in a room is the sum of direct and reflected sound.
A narrow pattern speaker will have relatively less reflected than a wide pattern.
I think reflected sound levels likely vary less with distance than direct sound.
So i think narrow is likely to fall off faster than wide, to our ears and on an RTA which sums both speaker and room.
There's a degree of speculation in my in-room analysis, and I'm sure a lot will depend on the room. But for small rooms, where even the rear wall contributes heavily, it seems to make intuitive sense.
If anybody knows the nuts on in-room, direct to reflected vs distance, pls post.
In all of my measurements, here was the one that blew my mind:
I was measuring a speaker indoors. The measurement looked mostly accurate, but the treble was rolled off badly.
I fiddled with the tweeter a bit, but it *sounded* like it was working just fine.
Then I realized what had happened:
I'd set the gate wrong on my measurement, and I was recording everything AFTER the first bounce off the wall.
In other words, the REFLECTION off of the nearest wall in my room was almost a perfect replica of the original sound. Except the treble was rolled off, and it was delayed in time.
That blew my mind, because instinctively I'd assumed that the reflection off the wall would be a small fraction of the original sound.
But when you think about it for a minute, why would it? We basically have a hard surface that's acting like an acoustic mirror. So if the loudspeaker is generating 90dB of output, that reflection could be as much as 87-89dB, depending on how close it is to the speaker.
That event really informed a lot of my thinking, on speaker placement, waveguides, etc. For instance, I mostly push my speakers all the way into the corner of the room, because the larger the delay between the incident wave and the first reflection, the more that the sound is 'muddied' imho. Of course, there's another school of thought where you move the speakers as far from the walls as possible...
Corner and in-wall placement minimize many problems in-room mostly SBIR ie. early reflections in low freq. This is practical in HT setups for FL/R
Just few weeks ago I set up speakers in our bedroom - and they are in corners! Normal low directivity 3-ways with sealed bass and I'm listening in the other end of the room lying on bed almost 5m away... Bass boost was easy to compensate with dsp and it took me only two weeks to learn to hear something like stereo image... too long distance so almost everything I hear is reflected!
Just few weeks ago I set up speakers in our bedroom - and they are in corners! Normal low directivity 3-ways with sealed bass and I'm listening in the other end of the room lying on bed almost 5m away... Bass boost was easy to compensate with dsp and it took me only two weeks to learn to hear something like stereo image... too long distance so almost everything I hear is reflected!
Last edited:
You're welcome Camplo, glad it's help.
I am not a good physicist.... and don't want to compare to a laser, but my understanding is that a laser obeys inverse square law just like everything else, once it's measured in the far field... something that takes a heck of alot of distance to get to.
I do know that audio, once in the far field, falls at 6dB per distance doubling regardless of pattern. (and ignoring air temp, humidity... effects on high frequencies etc,)
Far field is a bit important. Go back to Juhazi's chart in #24.
The line between near field and far field is marked by when SPL starts obeying inverse square law.
Inside that near field distance, closer to the speaker, drivers are acting a bit independently, and SPL may fall slower than 6dB per octave.
An old rule of thumb is that multi-way measurements need to be taken no closer than 3X the longest baffle dimension. IOW, nearfield multi-way measurements are a no-no.
Funny how we end up learning stuff like that !
Interesting how high a SPL the reflections measured.
I swear, i learn more from my mistakes than i do from what goes right.
Well, that also may just be due to my mistake-to-right quantity ratio
I've mentioned it before in the past, but worth mentioning again. When I had the Yorkville U15's I had them in the corners in a room that was really just insanely small for that large a speaker. With 1 absorber at the first reflection point on each side and one behind, the result was fabulous. The narrow/controlled directivity of the Unity really worked well in that arrangement.
Required EQ, to be sure, but the U15 needed EQ anyway. I used DRC-FIR and it was almost point-and-shoot. Just a great sounding system. Ugly and impractical though, which is why I had to reluctantly move on.
Do these two short articles help?
True Audio Speaker Topics: Spatial Loading
TA Speaker Topics - Points, Line, and Plane Sources
This site's ridiculously talented math and physics minds can sometimes grow tired of explaining the same fundamentals over and over. I understand their fatigue. But, repetition is exceptionally helpful when trying to learn, and even the Tom Danley's and Nelson Pass's are always trying to learn.
I am no expert in any of this, but I'm better than average with words, so I know when to refer to other people's.
I am not a ridiculously talented math and physics mind, certainly unworthy of those who grow tired of explaining. And thank you for the links that said...
Now, I do think it's a little weird why that strange magic would apply only to line source control patterns but not horn control patterns. Nevertheless, I'd never expect the great ones deign to explain it.
Now, I do think it's a little weird why that strange magic would apply only to line source control patterns but not horn control patterns. Nevertheless, I'd never expect the great ones deign to explain it.
It's not strange magic, the important information is that bolded in your quote above. An ideal line source produces a cylindrical wavefront and a horn a spherical one.
Most line sources are not ideal and therefore the -3dB with doubling of distance does not hold over all frequencies.
An example of that here Line Array Attenuation - Ideal vs Actual
Most line sources are not ideal and therefore the -3dB with doubling of distance does not hold over all frequencies.
An example of that here Line Array Attenuation - Ideal vs Actual
Was that in 04? Holy smokes - how could it have been that long ago? Had an engaging chat with Earl Geddes at that show, right?
Moved a couple times since then, most recently last fall. No shop for a while, and tons of stuff in storage. I've had to probably let go of the Unity/Synergy idea, and possibly speaker building in general at least for a while. But I am running Kef R3s, so trying to hold to the point source w/ somewhat limited dispersion idea.
Saw that design - have you posted more details? Drivers etc? Guess i should check the thread....
You are right when inside a horn but when it is finished after a few inches there is the same 4pi space as of direct radiators. The mouth behaves like a big diaphragm and has the same behaviour -6db per double distance in the far-field.
Line sources need to be long, at least a wavelength to function as a line pattern and also only in the same or less height further away you have only -3dB. It is mainly used in Public Address systems to reach further and compensate for the damping of the air of high frequencies.
Last edited:
Actually infinite line sources don't exist. Nice theory, but high frequencies beam like h*** vertically because of interferences. CBT works better, so distance doesn't change tonality, but spl drop is normal.
Don Keele's CBT (Constant Beamwidth Transducer) Page
Don Keele's CBT (Constant Beamwidth Transducer) Page
Attachments
Hi bradleypnw,
I'm only repeating and building on what's already been said in attempt to help clarify 6dB falloff vs 3dB falloff...
The relevant text from the TA article you quoted, was at its end:
"A finite line source will behave more as an infinite line when the observation point is very close to the line compared to its length. At greater distances the source looks more like a point radiator and the SPL will fall at 6 dB per doubling of distance.
A finite plane source likewise will exhibit constant SPL with distance only when the observation point is very close to the plane compared to the dimensions of the plane. At larger distances the finite plane source behaves more like a point radiator and the SPL will fall at 6 dB per doubling of distance."
This again gets to near field (very close observation point) vs far field.
Very close means just that!
So I doubt anyone is going to listen to a Summa or Hyperion close enough to be in the near field.
That "line arrays only fall at 3dB" is one of the biggest marketing myths floating. A unicorn imagined from a theoretical infinitely long straight line.