Hello
I'm experimenting cardioid implementation.
The basic principle seems self-explanatory.
You place the same woofer in the front and back of an enclosure, and feed an inverted signal delayed by that distance to the woofer behind it.
A practical problem arises.
Due to the radiation characteristics of the enclosure, the frequency response of the front woofer and back woofer is not the same.
It can be fixed by applying PEQ to the rear woofer to match the frequency response.
But the position of the microphone becomes an issue.
Should it be right in front of the rear woofer?
Or should 1 meter to the rear of the enclosure?(I'm measuring directivity at 1m)
My intuition is that it doesn't matter either way, as long as the effects of reflections are properly eliminated.
However, if I use the rear woofer as a reference, it doesn't cancel properly at the 1m position,
If I use the 1m position as a reference, it doesn't cancel properly directly behind the enclosure.
It was easy in the sub-bass band, where the enclosure has little effect, but seems not so easy in the midrange.
I would appreciate any help.
I'm experimenting cardioid implementation.
The basic principle seems self-explanatory.
You place the same woofer in the front and back of an enclosure, and feed an inverted signal delayed by that distance to the woofer behind it.
A practical problem arises.
Due to the radiation characteristics of the enclosure, the frequency response of the front woofer and back woofer is not the same.
It can be fixed by applying PEQ to the rear woofer to match the frequency response.
But the position of the microphone becomes an issue.
Should it be right in front of the rear woofer?
Or should 1 meter to the rear of the enclosure?(I'm measuring directivity at 1m)
My intuition is that it doesn't matter either way, as long as the effects of reflections are properly eliminated.
However, if I use the rear woofer as a reference, it doesn't cancel properly at the 1m position,
If I use the 1m position as a reference, it doesn't cancel properly directly behind the enclosure.
It was easy in the sub-bass band, where the enclosure has little effect, but seems not so easy in the midrange.
I would appreciate any help.
The frequency that you can get cardioid radiation to work through interference depends on the distance between the drivers. Large drivers facing in opposite directions will work better at low frequencies. To get a cardioid pattern higher up the cancelling drivers need to be on the side of the enclosure so the distancs between the sources and time delay is less.
This can be simulated to a reasonable degree in Vituixcad just based on the driver sizes and distances.
This can be simulated to a reasonable degree in Vituixcad just based on the driver sizes and distances.
@fluid
Of course it is.
My experimental enclosure is only 150mm deep, which should provide enough cardioid to the midrange frequency.
In practice, the influence of the enclosure causes a frequency response mismatch between the front and rear woofer.
The gist of my question is, where should be the reference position(where the microphone should be placed) to compensate that mismatch?
Right behind the enclosure (right above the rear woofer)? Or 1 meter behind the enclosure, as in a typical directivity measurement?
Of course it is.
My experimental enclosure is only 150mm deep, which should provide enough cardioid to the midrange frequency.
In practice, the influence of the enclosure causes a frequency response mismatch between the front and rear woofer.
The gist of my question is, where should be the reference position(where the microphone should be placed) to compensate that mismatch?
Right behind the enclosure (right above the rear woofer)? Or 1 meter behind the enclosure, as in a typical directivity measurement?
Unless I'm not understanding what your saying you should
Be measuring the speaker in 180 degrees.
Be measuring the speaker in 180 degrees.
Yes this is another reason why having cancellation drivers on the side is the preferred way to do it for the midrange.
If you are measuring for Vituix you should measure the rear driver as if it was forward facing, then rotate it and change the Z distance to put it in the right place in the simulation. If the drivers are the same front and back you could just measure one and use that as the basis for both drivers in the simulation.
Could it be my English?
1. In order to cancel the front woofer at the rear of the enclosure, the frequency response of the front and rear woofers must match.
2. Even though I use the same woofers, the frequency response is different due by the influence of the enclosure.
3. I want to apply EQ to the rear woofer to make the two frequency responses equal.
4. But where should be the microphone's position to equalize the two frequency responses?
5. 180 degrees of course, but how far?
I'm talking about practical implementation, not simulation.
I did simulation thru vituixcad a lot already..
1. In order to cancel the front woofer at the rear of the enclosure, the frequency response of the front and rear woofers must match.
2. Even though I use the same woofers, the frequency response is different due by the influence of the enclosure.
3. I want to apply EQ to the rear woofer to make the two frequency responses equal.
4. But where should be the microphone's position to equalize the two frequency responses?
5. 180 degrees of course, but how far?
I'm talking about practical implementation, not simulation.
I did simulation thru vituixcad a lot already..
I don't think so
No they don't, but the more different they are to start with the trickier it is
You need to low pass the rear woofer and set the delay to match the sound flying time between them. The group delay of an IIR filter can be enough delay on it's own if you choose the type and frequency correctly
The observation point should be where you intend to listen to them from.
I think he means making full polar measurements, if the enclosure is horizontally symmetric 0 to 180 degree measurements.
The best way to get a practical implementation is to measure the drivers you have and use Vituix to simulate the crossover/delay etc. So you can see what you need to get what you want.
Here’s something I prepared earlier.
A lower woofer is measured in the nearfield and farfield, and merged as per VituixCAD2 manual instructions, including the simulation of baffle diffraction simulation, where one woofer represents two. Check the manual Measurement with CLIO or ARTA or REW or SoundEasy for details
Here’s the response after filtering.
I "moved" the lower woofer to the back panel by putting (Z=400) (depth of the cabinet) and reversed the phase on that woofer, then performed a quick and dirty copy of all of the DSP setting that kind @vineethkumar01 shared in his post
With a bit of fine tuning I could sculpt it a bit more, but this is what it looks now. As you can see there is an additional octave of directivity control.
A lower woofer is measured in the nearfield and farfield, and merged as per VituixCAD2 manual instructions, including the simulation of baffle diffraction simulation, where one woofer represents two. Check the manual Measurement with CLIO or ARTA or REW or SoundEasy for details
Here’s the response after filtering.
I "moved" the lower woofer to the back panel by putting (Z=400) (depth of the cabinet) and reversed the phase on that woofer, then performed a quick and dirty copy of all of the DSP setting that kind @vineethkumar01 shared in his post
With a bit of fine tuning I could sculpt it a bit more, but this is what it looks now. As you can see there is an additional octave of directivity control.
Last edited:
In a pa there are several types, In a line array type, an enclosure reversed in the stacked array, either physically or just put out of phase. In a horizontal array, they are behind or beside each other.
And cancellation happens at wave length intervals, subtracted by the standing waves propagated by the in phase woofer.
Yes this is definitely not new.
In fact many years ago @john k... presented a dipole subwoofer by putting two subwoofers in the room; at opposite ends of the room,
One facing the room, the other on the opposite wall, this one facing the wall…
In fact many years ago @john k... presented a dipole subwoofer by putting two subwoofers in the room; at opposite ends of the room,
One facing the room, the other on the opposite wall, this one facing the wall…
I understand the question. Let's go though the situation step by step.
Lets say we have a cabinet with a depth of 0.25m with a front and back woofer.
We measure the speaker at a distance of 1m . Let's consider the woofer facing the microphone the 'rear' woofer and the one facing away from the mic the 'front'. So we are measuring at the back of the speaker where we desire cancellation.
To achive optimal cancellation at the mic position the 'rear' woofer needs to be attenuated by 2dB because it is closer to the mic by 0.25m.
However, when we consider an 'infinite' far-field condition the cancellation is now worse because that 0.25m difference has no impact on the relative SPL.
What if we want the speaker to cancel optimaly closer to the cabinet than 1m? Say we plan to place it 0.25m from a wall and want the cancellation maximal at that distance. Now we need to attenuate the rear woofer by 6dB.
If we carry on like this we end up with no rear woofer output at all!
I'm not sure what the corrwct answer is. I'd probably stick to 1m or infinite.
In the case above with the speaker 0.25m from a wall and say we want to minimise transmission to a neighbour. Would the infinite distance tuning simply have the cancellation beyond the wall, providing best attenuation for your neighbour or should we aim for best attenuation at the wall boundary?
If we simply consider sound power in the room, we get best reduction in the low frequencies with an 'infinite' tuning. I think that makes sense - least sound energy in the room for the same SPL at the listening position (provided you have the dynamic range to EQ flat).
Lets say we have a cabinet with a depth of 0.25m with a front and back woofer.
We measure the speaker at a distance of 1m . Let's consider the woofer facing the microphone the 'rear' woofer and the one facing away from the mic the 'front'. So we are measuring at the back of the speaker where we desire cancellation.
To achive optimal cancellation at the mic position the 'rear' woofer needs to be attenuated by 2dB because it is closer to the mic by 0.25m.
However, when we consider an 'infinite' far-field condition the cancellation is now worse because that 0.25m difference has no impact on the relative SPL.
What if we want the speaker to cancel optimaly closer to the cabinet than 1m? Say we plan to place it 0.25m from a wall and want the cancellation maximal at that distance. Now we need to attenuate the rear woofer by 6dB.
If we carry on like this we end up with no rear woofer output at all!
I'm not sure what the corrwct answer is. I'd probably stick to 1m or infinite.
In the case above with the speaker 0.25m from a wall and say we want to minimise transmission to a neighbour. Would the infinite distance tuning simply have the cancellation beyond the wall, providing best attenuation for your neighbour or should we aim for best attenuation at the wall boundary?
If we simply consider sound power in the room, we get best reduction in the low frequencies with an 'infinite' tuning. I think that makes sense - least sound energy in the room for the same SPL at the listening position (provided you have the dynamic range to EQ flat).
Last edited:
I hope it is OK to share an idea in this thread? 🙂
I am designing a two way active speaker who will be assisted by separate subwoofers.
The horn is 58 cm wide and will be highpassed around 650 Hz.
The midwoofer will work fine in closed cabinet around 44 liter. My goal is ca 80 Hz to ca 650 Hz.
I use DSP (Audiolense).
Any comments and ideas of the best way to implement cardioide slots?
https://www.precision-devices.com/products/all-products/pd-153er-2/
I am designing a two way active speaker who will be assisted by separate subwoofers.
The horn is 58 cm wide and will be highpassed around 650 Hz.
The midwoofer will work fine in closed cabinet around 44 liter. My goal is ca 80 Hz to ca 650 Hz.
I use DSP (Audiolense).
Any comments and ideas of the best way to implement cardioide slots?
https://www.precision-devices.com/products/all-products/pd-153er-2/
Last edited:
If you add slots to get cardioid polar pattern the speaker will loose A LOT of bass. It's one of the reasons cardioid is not common.