"Better" speakers to begin with, or more "fixes" everyone knows?

[AJ Soundfield]

Several threads here and everywhere on how to deal with "room" acoustics. The reality is, what's really being discussed is speaker/room interaction. There is no separating the two.
Multi-subs, EQ, room treatments. You've seen that innumerable times.
There is always a common factor. Boxes. In rare cases there may be a large panel speaker (dipolar) or two, but in 99% of the cases, its all monopolar box speakers...and how they interact with typical rooms.
Hopefully we can discuss some alternatives...

cheers

AJ

 

[Purité Audio]

Sounds good, personally my ideal system would be five way,front loaded horns, front loaded bass horns each the size of a grand piano but still only extending to around 30-35 Hz.
Upper bass horns are 1.5 metres diameter , midrange ,upper mids and tweeter all front loaded.
Downsides are mostly space related, you need a very large room, and unless you make them yourself a fairly deep pocket.
Keith.

 

[AJ Soundfield]

Borrowing from Amir, Harman/Revels (not sure which model, Salon should go deeper unless high passed?):

Stereophile, but with averaging over wider area around room:

Clearly one can see room interaction effects well up into a few hundred Hz, not just in the "sub" 10-100hz region.
Keep in mind, EQ cannot(should not) be applied to nulls. Attempting 7-8db boost at 480Hz in Amirs case to fill the hole, would cause large errors in the off axis response and create coloration in the reverberant field of the room.
The Salon has a narrow baffle and uses (3) 8" woofers. A typical 8" woofer has a radiating piston diameter closer to 7" and will become directional at around 700Hz or so. This means below 700Hz, the Salon will be radiating omni-directionally. i.e, there will be as much sound being radiated in front towards the listener, as there is towards the sides, floor, ceiling and rear behind the speaker, reflecting off the surfaces at full strength, then combining with the front/"direct" output, with delays, to create the peaks and nulls observed in the FR at the listening position.
It should hardly be surprising that this is exactly where one starts to see these effects 700hz +/-.
More to come...

cheers,

AJ

 

[Purité Audio]

With you so far!
Is there a quick and dirty way to determine at what sort of frequency a cone will start to become omni directional?
Keith

 

[RayDunzl]

... in rare cases there may be a large panel speaker (dipolar) or two...

That's me!

My room interaction, per impulse response, and recording of an impulse in the room (slowed way down to hear the echoes) reveals:

Some reflections from the top of the couch or the 7" of rockwool immediately behind the couch in the 1-2 ms area
The reflection off the wall behind the speakers at 7ms
The reflection 3x room length at 27ms - direct going over the couch, bouncing that wall, back to the wall behind the speakers, and
back to the mic at the listening position.

Nothing distinct for sidewall or floor or ceiling.

I have a 50Hz black hole at the listening position, that refuses to bother me. I didn't know it was there before measuring, and still
don't become annoyed by it.

Measures:
FR and Phase (AcourateDRC applied), crossover panel to woofer at 180Hz:

Step with and without DSP:

Impulse:

 

[amirm]

With you so far!
Is there a quick and dirty way to determine at what sort of frequency a cone will start to become omni directional?
Keith

I am glad you asked for quick and dirty as the exact computation is quite complex. The simple concept goes like this:

Here we have a circular transducer. If it were 100% beaming then there would be no spreading and the waveform goes out with the same diameter of the transducer as shown with the dashed horizontal lines ("BEAM AXIS"). The degree the beam spreads is said to be the angle/2 or alpha/2. Computation of that is as follows, assuming we stop when the beam strength is at -20 db

S (alpha/2) = .870 * c / fD

c = speed of sound in the medium
f = frequency of the sound
D = Diameter of the transducer

It is super important to stay consistent with units of f, c, and D. Don't mix inches, meters and millimeters . I have a spreadsheet for this although my application was computing this for transducers underwater (fish finders) so uses different constant for "c".

Without going through calculation, we see that as frequency gets higher, S becomes smaller. When S becomes smaller, then the beam axis is getting narrower and closer to the transducer width.

When diameter gets larger, the same thing happens. So all else being equal, a larger woofer will be more directional than a smaller one at the same frequency.

A typical bookshelf speaker with an 8 inch woofer and 1 inch tweeter falls victim to this. The crossover frequency will be too high forcing the 8 inch woofer to play higher frequencies. Higher frequencies means our beam angle gets smaller and hence the woofer will become directional. A directional woofer then means there is less sound coming out the sides. Translated, we have a dip in mid frequencies off-axis. That dipped response combines with the direct sound and screws up the overall response.

A better speaker would use a 5.25 inch woofer which would become less directional per above formula. But it won't sell as well because it won't play as low and look cheaper.

 

[Purité Audio]

Amir Hi, thanks for that,
If I understand correctly we wouldn't want a loudspeakers whose drivers produced sound only within their respective 'beam-axis' .
Because there would be no off axis response ?
So how do we ensure that we have smooth and even off axis response?
Also earlier AJ mentioned large papaer cones breaking up within their pass band a lá JBL, so why use them?
Are there any inherent advantages to using a large diameter paper cone other than historical?
Keith.

 

[amirm]

I am going to wait to see where AJ wants to take this thread. For now, here is an ominus warning from Dr. Toole from his book (repeated countless times in his other work and lectures)

"In some ways, our problems with rooms, especially small rooms, began when
we started to make measurements. Our eyes were offended by things seen in
the measurements, but our ears and brain heard nothing wrong with the audible
reality."

The only way to not fall victim to this is to use measurements below transition frequencies of 200 to 300 Hz. In other words, low frequency optimization is safe. What the measurements show is more or less what you hear. A 10 db peak at 60 Hz means exactly that -- you are amplifying notes in that frequency by that much. What happens at 480 Hz is a completely different animal. Remember this: you have two ears. You cannot capture what they hear with a single microphone.

 

[NorthSky]

Mom assisted to Le Sacre du Printemps directed by Ken Nagano couple weeks ago (March 10) in that acoustically treated concert hall room in Montreal; La Salle Symphonique de Montreal. There were 94 musicians.

She said that I cannot reproduce even 10% of it @ home. I would need a new home, new room, new speakers, and a lot of other stuff... I believe her.
The concert lasted three hours, with an encore and other varieties.

Ken Nagano is on tour in the USA, and the critics in New York (tough crowd) raved all over. New York wonder why they don't have Ken as their master orchestral music director. But Montreal does.

The hall is all wood, and the organ is truly something else; the Casavant organ.

 

[Blumlein 88]

So why not use a binaural dummy head for measurements?

 

[AJ Soundfield]

Non-box speaker system in 2 different rooms, one JAs as shown in the Salon measurements above:

http://www.stereophile.com/content/...-john-atkinson-march-1997#69P1t0B1H7KxozF2.97

http://www.regonaudio.com/Gradient Revolution Loudspeaker.html

Unfortunately, not the same measuring methods/resolution, but consistent with research:
http://www.aes.org/e-lib/browse.cfm?elib=12401

Showing LF directional control can result in more consistent room response, just like high/mid frequency directional control as advocated by Harman. Of course, there is more profit with selling room treatments and 12 subwoofers, so YMMV.

cheers,

AJ

 

[Purité Audio]

Interesting , so cardioid mid range and open baffle bass?
OB reduces box colouration and output , but won't affect room modes?
Keith.

 

[amirm]

Non-box speaker system in 2 different rooms, one JAs as shown in the Salon measurements above:

You need to include the label for the graph AJ. Once there, a different picture emerges:

1/3 octave filtering means the response is over smoothed for anything below thousands of hertz. Remove that filtering and the low frequencies will go up and down like there is no tomorrow.

We can see that in my article on time/frequency resolution trade off: http://www.audiosciencereview.com/f...urements-understanding-time-and-frequency.25/

This is with 25 Hz resolution. Pay attention to the back of the graph

Now let's increase the frequency resolution to 3.3 Hz:

Now look at the back of the graph. Now we have deep dips that were simply hidden when we had lower frequency resolution. Notice that these are from the exact same measurement. All we have done is decided how much to smooth the frequency response.

1/3 octave is only useful in acoustic measurements to see the overall timbre of the system. It hides the ups and downs so that you can see the overall response. It is NOT at all proper for measurement of bass. There you want to go at 1/12 octave resolution or better.

 

[amirm]

Showing LF directional control can result in more consistent room response, just like high/mid frequency directional control as advocated by Harman.

It is not the same thing. Mid to high frequencies are subject to psychoacoustics due to smaller wavelengths relative to our body dimensions. Bass frequencies have far longer wavelengths and as such, are much simpler to analyze. What we hear there is not at all the same as what we hear at higher frequencies. You would need to show independent research for "LF directional control" mattering.

 

[AJ Soundfield]

Interesting , so cardioid mid range and open baffle bass?

Yes, directivity control and more consistent sound power full spectrum. You can read the perceptual comments of each reviewer regarding the result, though non-blind.

OB reduces box colouration and output , but won't affect room modes?

Gradients (dipoles/cardioids, not the brand) reduce the amount of power radiated into the room significantly. Depending on polar shape, one could, for example, reduce the power directed at the wall directly behind the speaker, by 20db. So instead of adding padding, aka "bass traps" too absorb that power you've generated, you avoid generating it in the first place.
The modes will be excited differently than a monopole, which excites all maximally due to the omni-directional radiation. The gradients will excite some modes with far less strength. In the case of the Revolution, the bottom cabinet can be rotated, allowing it to further couple with/excite modes differently, in the same exact spot, simply by rotation. Something impossible with a box.
The "coloration" aspect would be related to modal decay...the audible variety. The gradients would be the equivalent of the boxes with some absorption to reduce decay strength. IOW, you're "fixing" the source to begin with, rather than after the fact. One can then apply EQ as needed.

cheers,

AJ

 

[AJ Soundfield]

Unfortunately, not the same measuring methods/resolution

You need to include the label for the graph AJ. Once there, a different picture emerges

Yep.

So the Backman paper is wrong because you said so? Or do you have a counter study?

 

[AJ Soundfield]

Showing LF directional control can result in more consistent room response, just like high/mid frequency directional control as advocated by Harman.

It is not the same thing. Mid to high frequencies are subject to psychoacoustics due to smaller wavelengths relative to our body dimensions. Bass frequencies have far longer wavelengths and as such, are much simpler to analyze. What we hear there is not at all the same as what we hear at higher frequencies. You would need to show independent research for "LF directional control" mattering.

When did your strawman claim low/mid/high perception is the same???

 

[Purité Audio]

Sounds all good, cardioid bass is what Bruno is attempting with the Kii THREE of course are there any disadvantages?
Keith

 

[AJ Soundfield]

You would need to show independent research for "LF directional control" mattering.

Sure: http://www.aes.org/e-lib/browse.cfm?elib=12663

 

[AJ Soundfield]

Sounds all good, cardioid bass is what Bruno is attempting with the Kii THREE of course are there any disadvantages?
Keith

Yes, Kii is a cardioid. Disadvantages? Cost, complexity, overcoming audiophile beliefs and love affairs with boxes, the only thing they know.
The Pro world of course is moving forward, many gradients starting to appear for large and small venues, like studios.