Stair Cardioid Subwoofer Array

Re: Article: Stair Cardioid

Sebastian Rivas Godoy said:
So the turned off acts like a resonator, or pasive device. In this case the obstacle box was not conected, and not shorted. So new measurements are required.

On the measurement you showed, where was the mic located ? Is that a local effect near to the obstacle sub, or a general change in the farfield ?
Click to expand...
Sebastian,

The measurement mic was on axis on the ground 2 meters from the mouth of one sub. The mic position remained in the same position when the "dummy" (unpowered, then shorted) sub was put in place along side.

The effect of the second sub is apparent in both near and far field measurements, though on my property there are some building reflections that make the 2 meter measurements more indicative of the actual response.

The resonant effects of the non terminated sub correspond with the impedance curve, being a tapped horn there is a secondary impedance peak evident around 100 Hz that causes a corresponding frequency response dip similar to the one around 45 Hz. A bass reflex enclosure would likely only have one dip in response in the pass band.

I will look forward to your new measurements, as another data point it would be interesting to compare the results with the obstacle cabinet shorted compared to connected to a turned on amplifier, no doubt the three different measurements (open, shorted, connected to turned on power amp) will be recognizably different.

Art
 

Attachments

  • Impedance, Frequency response .jpg
    Impedance, Frequency response .jpg
    65.7 KB · Views: 3
Re: Article: Stair Cardioid

Tim Padrick said:
Why measure a cabinet at a distance that's closer than the nearest likely listener? Distance changes things.
Click to expand...
Tim,

As I mentioned, on my property there are some building reflections that make the 2 meter measurements more indicative of the actual response than a greater distance. Two meters actually is in the far field of the sub measurement, little change occurs at further distance, other than the introduction of building reflections.

Measuring high frequency response requires a greater distance to be in the far field.
 
Re: Article: Stair Cardioid

Hey Guys,
you should measure the polar response of one cabinet, then add a second inactive cabinet (with its terminals shorted) and see the change in the polar :)

cheers
ferrit
 
Re: Article: Stair Cardioid

Ivan Beaver said:
That is the reason when I take a system into a venue that has a house system-I ask the house guys to turn their system on and mute the DSP.

The amplifiers act as "brake" on the cone movement. Not as good as shorting out the cabinets-but better than leaving them open-as they would be if the amps were off.
Click to expand...

Audiophilia doesn't get much love in prosound circles (rightly so in many cases), but removing unused speakers or at least shorting them was de riguer in high-end listening salons in the late 70s/early 80s. Unwanted sympathetic compliances do muck up the results.
 
Re: Article: Stair Cardioid

Hey Guys,
with the physical size of Pro subwoofers, even with the large wavelengths involved, there's a physical disturbance (boundary, mutual-impedance etc.) that can have surprising effects. :)

cheers
ferrit
 
Re: Article: Stair Cardioid

Martyn ferrit Rowe said:
Hey Guys,
with the physical size of Pro subwoofers, even with the large wavelengths involved, there's a physical disturbance (boundary, mutual-impedance etc.) that can have surprising effects. :)
Click to expand...
Ferrit,

Why don't you measure the polar response of one cabinet, then add a second inactive cabinet (with its terminals shorted) in front, and post up the change in the polar response for us ;^) ?

Cheers,
Art
 
Re: Stair Cardioid Subwoofer Array

In an end-fire configuration, for example the two tapped horn boxes Art used to demonstrate the difference a passive radiator can have on measurements, if that arrangement was used with both boxes powered (and the forward box appropriately delayed) does the actual box-type of the front box have an effect that can be either measured or heard?

What I was thinking is that in the case of horn loaded sub boxes arranged and delayed to make an end-fire array, whilst the driver in the front box is not passive (as in Art's example) - surely the sound propagated forwards from the rear box through the front one will travel a second time through the horn loading already used to create the original sound, but the second time it also combines with the output of the front box, achieving a degree of directivity but also potentially having an unwanted effect on the sound originating from the rear box.

My query is whether in a case where the box loading is a significant contributing factor to the sound produced (such as a horn path of some type), is there an effect that would influence the resulting sound of both boxes due to the sound from the rear box having to travel through the horn path of the forwards box? If there is such an effect, how would it change the sound heard and any measurements taken? Finally would this only be an issue in more complex box designs (eg horn loaded subs, 4th order bandpass) and not a concern when using front-loaded bass reflex designs, such as those used in Sebastian's experiment?
 
Re: Stair Cardioid Subwoofer Array

Guy Graham said:
1.My query is whether in a case where the box loading is a significant contributing factor to the sound produced (such as a horn path of some type), is there an effect that would influence the resulting sound of both boxes due to the sound from the rear box having to travel through the horn path of the forwards box?
2. If there is such an effect, how would it change the sound heard and any measurements taken?
3. Finally would this only be an issue in more complex box designs (eg horn loaded subs, 4th order bandpass) and not a concern when using front-loaded bass reflex designs, such as those used in Sebastian's experiment?
Click to expand...
1. When speaker enclosures or obstructions of any type are in the sound path between the rear point of origin and the listening position, they will have an effect on the sound. When an obstructing speaker enclosure is powered (or shorted) it will have similar effect on the sound regardless of it's design, if left passive, the effects are dependent on the impedance curve of the particular enclosure. In an actual array, the only reason a cabinet would be left passive would be due to blown transducers or improper termination.
2. The effects are peaks and dips in the response as can be observed in Sebastian's "obstruction" tests with an non shorted cabinet, and my tests with a single cabinet, then with a flanking cabinet shorted and not shorted.
3. There are measurably different effects that result from arrays using different enclosure types, the resultant behavior needs to be equalized both for the array, enclosures, and room. Langston Holland's tests using EAW horn loaded enclosures in a "stair" configuration nearly a decade ago showed good results, but certainly would require different EQ and delay settings than a bass reflex enclosure. Every different array requires different settings, and will have differing polar response, even if the different arrays measure the same in one on axis position.
 
Re: Stair Cardioid Subwoofer Array

Thanks Art for your swift and thorough response. Despite much searching, I was not able to find the thread relating to Langston's tests. Hence I have some further questions:

You mention good results achieved using horn loaded subs in an end-fire configuration. But you also describe this as a "stair" configuration, which I assume is arranged similarly to that used by Sebastian at the start of this thread. Also it is not quite the same as an end-fire array where both/all boxes are on the same level plain.

Thus my question is - would the same good results happen, if horn loaded subs are arranged and appropriately delayed (not taking into account the delay needed to align them with the mains) when both/all cabinets are on the same level?

Is there not an effect that might influence results, and needs to be accounted for, when a horn loaded box which derives some of its efficiency (and perhaps extension) from its own horn path, box loading etc - then directs a significant amount of energy right through another identical horn path of and via the identical box in front.

This could mean when both boxes sum and propagate forwards - the part of the signal from the rear box, which has subsequently also passed through an additional horn path, causing the whole of the sound propagated forwards by the rear box to be split into multiple sources, potentially lacking temporal coherence - as the simple sum of all sound from the rear box, once it has reached the front of the forward box, is compromised in the time domain purely due to the different paths that different parts were forced to take moving forwards?

Furthermore would this not then compromise the phase coherence of the sum of both the rear and front boxes, when the totality of their output sums at any point in front of the forward box ... and finally can these potential issues be corrected via DSP tools?
 
Re: Stair Cardioid Subwoofer Array

Hey Guys,
Okay First up is a polar measurement made of a 2 x 18" horn loaded subwoofer in landscape mode, measured at 5 degree increments at a distance of 10m. Looking at 1/3rd octave intervals from 44.5Hz to 200Hz.
It clearly demonstrates that its is not a point source, but has some directivity.

Next we place a second enclosure alongside, with its terminal shorted and measure again, this polar is shown as a blue trace overlayed on the single reference trace (black).
Left.png
In the next post I'll show an enclosure on each side and then spaced enclosures.
cheers
ferrit
 

Attachments

  • single.png
    single.png
    270.1 KB · Views: 3
Last edited:
Re: Stair Cardioid Subwoofer Array

Okay Here's a single enclosure with "shorted" enclosures each side, again with the blue trace overlayed on the reference trace (black).

And finally three enclosures spaced 50cm, with only the center enclosure driven.

spaced.png

:twisted:
ferrit
 

Attachments

  • middle.png
    middle.png
    277.1 KB · Views: 3
Re: Stair Cardioid Subwoofer Array

Hey Arthur, that was the first set of polars, the one with no cabinet showing, and then its shown in black as a reference in the other polars :)
With the blue polar overlayed.
cheers
ferrit
 
Re: Stair Cardioid Subwoofer Array

Guy Graham said:
1.You mention good results achieved using horn loaded subs in an end-fire configuration. But you also describe this as a "stair" configuration, which I assume is arranged similarly to that used by Sebastian at the start of this thread. Also it is not quite the same as an end-fire array where both/all boxes are on the same level plain. Thus my question is - would the same good results happen, if horn loaded subs are arranged and appropriately delayed (not taking into account the delay needed to align them with the mains) when both/all cabinets are on the same level?

2. Is there not an effect that might influence results, and needs to be accounted for, when a horn loaded box which derives some of its efficiency (and perhaps extension) from its own horn path, box loading etc - then directs a significant amount of energy right through another identical horn path of and via the identical box in front.

3.This could mean when both boxes sum and propagate forwards - the part of the signal from the rear box, which has subsequently also passed through an additional horn path, causing the whole of the sound propagated forwards by the rear box to be split into multiple sources, potentially lacking temporal coherence - as the simple sum of all sound from the rear box, once it has reached the front of the forward box, is compromised in the time domain purely due to the different paths that different parts were forced to take moving forwards?

4.Furthermore would this not then compromise the phase coherence of the sum of both the rear and front boxes, when the totality of their output sums at any point in front of the forward box ..

5. and finally can these potential issues be corrected via DSP tools?
Click to expand...
1. Langston's 2005 tests were done with the rear EAW BH760 lifted above the front sub, what Sebastian refers to as a "stair" configuration. The result of any different configuration will be different.
As I recall, his tests were done shortly after this thread hit:
subs and deploying them 'in the round'
In this post:
subs and deploying them 'in the round'
He mentions:
"Bonus: (bottom 2 BH760's only, notice the 6dB front to back peak output differential and the 10dB spread around 75Hz. This is not a cardioid arrangement, just an advantage of horn subs.)"

Four years ago I did some testing to determine the difference in directivity between three different subs types:
http://www.diyaudio.com/forums/subwoofers/184982-tapped-horn-directivity.htm
The proximity to local surrounding objects had a rather large influence on the test results, but applied to all the cabinets tested.
The bass reflex tested was basically omnidirectional within about 1 dB to 60 Hz, then drops about 3 dB at 125 rearward, and around 7 dB at 160 Hz. The front loaded horn was basically omnidirectional within about 3 dB to 60 Hz, then drops about 7 dB at 125, and above. The tapped horn was basically omnidirectional within about 2 dB to 60 Hz, then drops about 12 dB at 125, and around 9 dB at 160 Hz.
The tapped horn with a mouth extender showed the advantage of a large frontal area, it has some directivity all the way down to below 30 Hz. At 60 Hz, it has around 7 dB difference, and 125 Hz, a bit over 14 dB. A huge directionality increase from a sheet of plywood.

2. Horn subs do have more directivity in the upper range than bass reflex subs, so will show more difference in response when that upper directivity reflects off another box placed in front of it. The lower frequencies, being more omni, only "see" the front box obstruction as a small percentage of the hemisphere (if on the ground plane) they radiate in to.

3. Most of the low frequency sound waves from the rearward box do not actually "pass through an additional horn path", but yes, upper bass frequencies are compromised in the time domain due to the different path lengths around a box placed in front.

4. Yes.

5. More or less, depending on frequency and location ;^).

Art
 
Last edited:
Re: Stair Cardioid Subwoofer Array

Martyn ferrit Rowe said:
Hey Guys,
Okay First up is a polar measurement made of a 2 x 18" horn loaded subwoofer in landscape mode, measured at 5 degree increments at a distance of 10m. Looking at 1/3rd octave intervals from 44.5Hz to 200Hz.
It clearly demonstrates that its is not a point source, but has some directivity.

Next we place a second enclosure alongside, with its terminal shorted and measure again, this polar is shown as a blue trace overlayed on the single reference trace (black).
View attachment 9887
In the next post I'll show an enclosure on each side and then spaced enclosures.
cheers
ferrit
Click to expand...
Ferrit,

1. What does "in landscape mode" mean?
2. What subwoofer was measured?
3. Are the measurements actual, or simulated? If actual, could you describe the measurement conditions, that is indoors or outside, distance to any large objects, mic height, etc.

Thanks,
Art
 
Re: Stair Cardioid Subwoofer Array

Hey Art,
1. "Landscape mode" refers to the orientation of he cabinet, in this case "lying down". Measurements were also made in portrait mode, i.e. "standing up" or on end.
2. This enclosure is the Martin-Audio MLX, a self-powered, dual 18", horn-loaded "Hybrid" enclosure.(Hybrid refers to Horn-loaded plus reflex-ported design).
3. Measurements are Actual, using MLSSA, performed at 5 degree intervals at 10m rotation, centered on the front of the enclosure, ground plane, indoors with the nearest boundary (ceiling) at @35m, they were then post processed to produce the 1/3rd octave data for ease of viewing.


cheers
Ferrit
 
Re: Stair Cardioid Subwoofer Array

Martyn ferrit Rowe said:
3. Measurements are Actual, using MLSSA, performed at 5 degree intervals at 10m rotation, centered on the front of the enclosure, ground plane, indoors with the nearest boundary (ceiling) at @35m, they were then post processed to produce the 1/3rd octave data for ease of viewing.
Click to expand...
Ferrit,

35 meter ceiling, the riggers there must have some strong arms!

Interesting that the tight pack flanking cabinets result in a tighter pattern at 80 than 100 Hz.
Any major differences noted in portrait compared to landscape orientation?
Have you measured polar response of the Martin-Audio MLX in cardioid or end fire arrays?
 

Attachments

  • MLX 80,100Hz.png
    MLX 80,100Hz.png
    74 KB · Views: 2
Top Bottom