The Objectives of a Loudspeaker in a Small Room

[MBK]

Paul,

Just for the record, my apologies for taking the liberty of carrying your torch into unchartered territory here, I wasn't sure if you're listening into this thread and I thought the subject was appropriate ...

[Brett]

Originally posted by Paul W
[B]Here is the "quad" midrange MBK referenced. The concept is to achieve the directivity of a 15+" with the energy storage characteristics of a smaller driver. This isn't new as I've seen similar designs from Europe...the only significant difference here is the use of a heavy wool felt half-cylinder to damp the rear-wave for a smoother transition to a forward firing waveguide (felt is not shown in these photos). This midrange is used with a BMS4552ND in either a DDS Eng-1 or PHL/MSC/PE 12" waveguide.

I drew up something very similar to this design this morning, completely forgetting that I'd seen yours before.

Is the waveguide at the centre of the mids? It's not clear from the photo's and I'm having trouble guessing from the dimensions. Cheers

[Paul W]

my apologies

...no secrets here

Is the waveguide at the centre of the mids?

The waveguide is in a separate module (not shown) that sits on top of the mid/bass baffle. I've seen a quad arrangement with a WG in the center but, in this case, didn't care for the spacing with a larger WG.

Paul

[jlo]

from gedlee : if you look at the radiation pattern of the ESP line you will see a gradual reduction in level as one moves off axis. This makes the "sweet spot" much wider as I think that you can understand. If one move back slightly as one moves to the side the effect is even more pronouced because the time of travel is also being compensated.

I'd like to confirm the time/amplitude compensation of the precedence effect (aka Haas effect), that means, when we move nearer to one speaker, can we compensate the time offset (the sound of the nearer speaker will come sooner) by a higher amplitude from the far speaker through increased directivity and toe-in.
To check this, I just did a small and free software here so I can try and find best curves, and verify how directivity should be dependant or not of frequency.
While listening on headphones, just move a dot on the time axis and also move it on the amplitude axis, so that the signal stays centered.
As you have 7 dots, you can trace a curve level vs time.
If somebody wants to try, would be nice if he gives us his results...

[LineArray]

Originally Posted by gedlee

VER, to me, are less than about 5 ms. In the < 2ms. region we found that the VER would actually sound like nonlinear distortion, and are highly dependent on absolute SPL, but also muck up the image. At greater than about 2 ms. the effects transition more to coloration than image. At about 10 ms. the negative effects of reflections begins to go away and by 20 ms. the reflections become positive being perceived as "spaciousness", reverb, etc.
...

There is something happening between 10ms and 20ms.

Some unsorted thoughts on that.

1) I worked in developing speech recognizing systems in the early 90's.
Around 20ms has been the processing time frame that was suitable
for feature extraction in the system we worked with.

Overlapping frames of approx 20ms are often used in that area.

2) In 10ms sound travels around 3.4m
1.5m is the approximate minimum distance of an open baffle dipole
speaker to be placed away from a rear wall, to achieve that reflections
are not spoiling sonic impression.



Timeframes in that range seem to form an "acoustical context" in
perception. If early reflections are present in that context, the image
of original sounds is blurred from

distortion to
coloration to
change in the perception of the "sound event space" (Schallerignisraum)

in order of the time delay of the reflections.


The VER Range <5ms is related to the design of the
loudspeakers structure (enclosure) itself:

- Reflecting surface structures
- Sharp edges causing diffraction,
- Internal cabinet reflections which find their way outside
and so on.

Interestingly to a certain extent a loudspeaker can be judged
by its look, if such flaws are to be expected.

Then comes the time range where the loudspeaker has
to deal with the next boundaries in the room.

E.g. a flat speaker directly in front of wall may simulate
in wall mounting.

There are only 2 possibilities to keep reflections out of the
10 ms range:
1. Directivity (as constant as possible) of the speaker or
2. Keeping distance to the walls.

Strategy 1 leads to a speaker which is less room dependent.
A less room dependent speaker is the better speaker, especially
if a minimum distance cannot be kept.

An open baffle dipole is a combination of both strategies,
while demanding distance to the rear wall it is relatively
benign to the sides.

The diffuse radiation of DML may add a new or supporting strategy,
in making early reflections less pronounced, since it is the
transient character in reflections which seem to blurr imaging,
especially in small rooms or at low distances to the surrounding walls.

Interference in more stationary sound seem to cause coloration which
can also be minimized by directivity and diffusivity. They are no mutually
exclusive strategies.


When comparing the concert hall sound field with that of the small home
listening room we have a major problem:

We want an image of the (high quality) sound field of the concert hall
(or virtual sound event space) to be placed in the acoustical small room.

If the job where the other way round, the task would be much easier ...

The quality of the speaker-room system has to pimped in terms of

- suppression of early reflection
- modal density
- diffusivity

Since the job is impossible enough, one cannot afford to leave out
one of these dimensions IMO.

Kind Regards

[LineArray]

...
instead of "modal density" i should have written
"balanced modal excitation below schroeder frequency"
because the modal density of a given small room cannot be changed.

Kind Regards

[markus76]

Originally Posted by LineArray

"sound event space" (Schallerignisraum)

That's probably the wrong term.
Schallereignis = sound event = physical properties of the sound field
Hoerereignis = auditory event = subjective sound perception of sound

I guess you wanted to talk about spatial attributes of auditory events?

Originally Posted by LineArray

There are only 2 possibilities to keep reflections out of the
10 ms range:
1. Directivity (as constant as possible) of the speaker or
2. Keeping distance to the walls.

3. Flush-mount speaker into a wall.

[LineArray]

Originally Posted by markus76

That's probably the wrong term.
Schallereignis = sound event = physical properties of the sound field
Hoerereignis = auditory event = subjective sound perception of sound

I guess you wanted to talk about spatial attributes of auditory events?

Right, it is the auditory event in the first place which changes by
listening room reflections.

The image formed by auditive perception contains an estimation
of the properties of the "sound event space" (Schallereignisraum)
e.g. a concert hall where a live recordig had taken place, with
its unique reverberation characteristics.

That perceptive image of the "sound event space" may also be changed
by listening room reflections, not the "sound event space" itself.

Kind Regards

[graaf]

Originally Posted by gedlee

Later lateral reflections are all good, very early ones are not. Floyd does not make this distinction and to that I do disagree. There are many other consultants that I have talked to who agree with me on this point.

well, this is quite fundamental disagreement, with immense impact on practical speaker design principles, isn't it?

[Pano]

Old threads never die?