Thank you so much for this video. Even though this is only 6 minutes It has provided me with so much clarity on the physics of sound as well as subwoofer application!
Yes, I am one of those people that do not use the like button here on UA-cam. However, you earned the thumbs up maneuver! Great presentation on this subject!
hmmm, I can't remember. But, choose a frequency near the center of the operating range. Consider the relationship with the mains. If you plan to haystack, you may be extending the operating range considerably.
Superb! Thank you for this video. One question to clear out some thoughts: In this example the choice of frequency and distance was purely based on working with round numbers, right? What's important is to consider the distance between subs. If we try to achieve the same effect with stacked subs one will be faced forward and the second one faced backward with polarity inverted and with a delay equivalent to the depth to the cabinet, is this correct?
Hey Diogo. You're welcome! :) > the choice of frequency and distance was purely based on working with round numbers, right? Yes it worked out well for the demo to have round number, but 1m spacing is not an uncommon design and is an easy goto template of mine when I am in a hurry. > If we try to achieve the same effect with stacked subs one will be faced forward and the second one faced backward with polarity inverted Yes, an in-line gradient and its stacked brother use the same processing and calibration procedure. In a 2-element array, the one you reverse will control where you want the null to go. In front of a stage, for example, you might want the null to go towards the downstage vocal mic, if there is such a thing. In a 3-element array, which one you reverse can have some minor affects, which @MerlijnvanVeen has written about here: www.merlijnvanveen.nl/en/nl/studiezaal/135-which-sub-to-reverse > a delay equivalent to the depth to the cabinet That will be the case in MAPPXT, but in the real world it's a little longer because sound has to travel around the box, which is why you need to calibrate it in the rear in the field.
What if we want a 3 line array instead of 2 ? Should we just reverse the 3rd one with no delay ? Making it kinda of a mix between cardioid and endfire ? Or you have a better idea ?
Hey Paulo, the only way I know how to do gradient with an odd number of subs is the inverted gradient stack. If you want three in a line, I think you're better off with an end-fire design. But, please experiment and let us know what you find out! :)
Hey Gordon, no, you would not get the same result. Everything in front would 90º out of phase and everything in the back would be a 180º out of phase, I think. Give it a shot in MAPP or in the real world and let me know your result.s
I don't have MAPP, but my RayEnd shows it working without reverse polarity. Several articles present it that way as well. One article in Sound Forums has yet another twist on fine tuning the array by stair stepping and tweaking the levels. Thanks a lot. I am going to try it this summer. soundforums.net/articles/stair-cardioid-subwoofer-array-192534/
@@nathanlively The technique Gordon refers to provides broadband summation in the front and a phase-based (frequency-dependent) cancellation in the rear. The technique you discuss in your video provides broadband (polarity-based) cancellation in the rear and frequency-dependent summation in the front. I showed the effects of both in this article: www.prosoundweb.com/tech-topic-friends-in-low-places/
You can try! :) Using two matching subs is already a challenge because no manufacturer produces subwoofers that within 1dB of each other through the operating range. This is why 30dB of rejection is impossible in the field. So if you have two different subs, you'll have to spend some time making them match. Maybe try an FIR filter with auto magnitude?
Hello Nathan, how are you? I understand the concept and thank you very much for your videos, but what's throwing me off, I know low-frequencies are omni-directional, but aren't the front and rear speaker's waves performing a phase shit when they diffract backwards?
Hey Lorenzo, I'm doing alright. Thanks. >rear speaker's waves performing a phase shit when they diffract backwards? I don't think so, but I could be wrong. This might be true if you had something like an open back guitar cabinet, but then you would end up with a figure-8 polar pattern. Have you tried measuring a sub in the rear and found that it had a polarity inversion? What makes you think that sound waves going around an object would change their polarity?
Hello Nathan, thank you for your reply. I just noticed I wrote phase shit instead of phase shift. This was just a thought I had since I was wondering if low-frequencies performed a slight phase shift to diffract to the back of their cabinet. I may be wrong. Unfortunately I don’t have the equipment to test this. Thank you again for your videos and for your reply. Have a good one :)
Simple í coupling 2 inductor coils like when design electronic always twin coils with control in middle so we had choice of up, down or neutral if connect two so induction automatic neutral inbound frequency should immediately react with negative outward
@@nathanlively I want to use Subwoofer as End Fire at 35Hz Frequency. If there is an equation to find out how much to delay the subwoofer on the back, can I explain it? Thank you.
Sorry im a bit slow... so one meter at 80hz would be 1/4 wave lengths righ. im trying to do a inverted cardio set up with out using DSP. just an old xover.
couldn't we even get more a perfect-like cardioid pattern over a large range of frequencies if we'd process the low-end frequencies by a multiband filter and apply different delays to each of them?
Hey Paul, the delay is not really the limitation to perfect cancellation. It's manufacturing tolerances. No manufacturer builds speakers that match to within 1dB. So if you want a custom performance increase, you might try micro surgery EQ until you can get your subwoofer response to match through the operating range to within 1dB.
Nathan Thanks a lot for the nice explaination. But you considered just 86 Hz. With this we get full summation due to 360 Degrees Shift. But what about the other frequencies? Suppose we drive 26Hz to 100 Hz with the Subs. So how can i make sure complete Subwoofer spectrum has summation? is this 1 Meter sub differenze and 1 Sec Delay a constant? so that we can use in any ocation? Or do we have to calculate it targeting a certain Frequency? Fo instance u targeted 86 Hz here. The same do i have to do differently if i target 60 Hz?
> But you considered just 86 Hz. Yes, this seemed like a good way to simply the discussion. > But what about the other frequencies? A gradient does have limitations on its usable bandwidth and you need to be away of these when it is deployed in consideration of the filter placement and level relationship. Please read this -> www.merlijnvanveen.nl/en/study-hall/167-lack-of-impact-revisited > So how can i make sure complete Subwoofer spectrum has summation? You can't, but you can plan for the bandwidth limitations using your design modeling software, the Subwoofer Array Designer, or with careful measurements in the field. > is this 1 Meter sub differenze and 1 Sec Delay a constant? Is that what I said? I thought I said 1m and 2.9ms of delay. 1 second would be about 345 meters. Either way, the answer is no. It will depend on the speed of sound. >so that we can use in any ocation Yes, I often use 1m and 3ms as a template in the field and then optimize from there. > Or do we have to calculate it targeting a certain Frequency? Yes, a design considering the maximum forward summation frequency is recommended, but if you want to work from a template, then just don't exceed 1m. You are asking a lot of good questions. My challenge to you is to open up Subwoofer Array Designer, do some experiments for yourself, and let me know what you discover. Good luck! :)
I'm not convinced this is correct. I'm not a trained acoustical engineer, just a musician who runs sound from stage. But in my admittedly limited understanding, the only physical thing that happens to a speaker when the polarity is changed is that the speaker moves in the exact opposite direction of how it moved with correct polarity. In correct polarity, the initial speaker movement the moment sound is first produced is toward the FRONT of the speaker. When polarity is changed, that initial speaker movement is instead directed toward the rear of the speaker. That is the only physical, mechanical, thing that happens when polarity is reversed. If I, instead of changing the initial speaker movement of a second enclosure in relation to the first one by changing it's polarity, simply rotate the second enclosure 180° in relation to the first enclosure, thereby causing the speaker in the second enclosure to move in the exact opposite direction of the first, which is the same physical motion caused by changing it's polarity, haven't I accomplished the exact same outcome? And with proper time allignment via delay, haven't I already accomplished the same goal of cancellation toward the stage and summation toward the audience? Help me understand where I'm wrong.
Hi tlb2732! Thanks for your interest in the video. I have another video that talks about this, but it's possible that you will only be convinced by trying it yourself. Which is easy enough to do, even without an audio analyzer. With an audio analyzer, take a measurement of a sub (ideally outdoors), then rotate it 180º and measure it again. What do you observe? Without an audio analyzer, you could take two subs side by side with the same signal. Turn one one, turn the other on. They should sum. Invert the polarity of one, they should cancel. Now restore normal polarity and rotate one of them 180º and try the same test. ua-cam.com/video/G0l6yniHslo/v-deo.html
A job well done but not all system react the same so there must be a standard or must have calibration engineer to calibrate each system before use and that is a good enough job for AT or ET
I believe Underfilip meant that when you were explaining how it works behind, you moved the wave to the wrong direction. Delay doesn’t make it sound “faster”…
That was the most understandable explanation I've ever heard. Thanks for dumbing it down for me!
This was EXACTLY the explanation I was hoping I would find. I knew generally THAT this worked but hadnt yet see visually HOW. Thanks much.
You're welcome!
Nice explanation
Thank you so much for this video. Even though this is only 6 minutes It has provided me with so much clarity on the physics of sound as well as subwoofer application!
You're very welcome!
Yes, I am one of those people that do not use the like button here on UA-cam. However, you earned the thumbs up maneuver! Great presentation on this subject!
I appreciate that!
Great video Nathan making Carioid much easier to understand. Thank you sir!
Glad you liked it!
Exactly what I was trying to find. Very nice
Is it possible to have a cardioid system with 2 stack subwoofers?
Thank you for your easy to understand technical explanation 👍
Hi Nathan, why did you choose 86Hz? Is it because it’s the filter point with front speakers ?
hmmm, I can't remember. But, choose a frequency near the center of the operating range. Consider the relationship with the mains. If you plan to haystack, you may be extending the operating range considerably.
Superb! Thank you for this video. One question to clear out some thoughts:
In this example the choice of frequency and distance was purely based on working with round numbers, right? What's important is to consider the distance between subs.
If we try to achieve the same effect with stacked subs one will be faced forward and the second one faced backward with polarity inverted and with a delay equivalent to the depth to the cabinet, is this correct?
Hey Diogo. You're welcome! :)
> the choice of frequency and distance was purely based on working with round numbers, right?
Yes it worked out well for the demo to have round number, but 1m spacing is not an uncommon design and is an easy goto template of mine when I am in a hurry.
> If we try to achieve the same effect with stacked subs one will be faced forward and the second one faced backward with polarity inverted
Yes, an in-line gradient and its stacked brother use the same processing and calibration procedure. In a 2-element array, the one you reverse will control where you want the null to go. In front of a stage, for example, you might want the null to go towards the downstage vocal mic, if there is such a thing.
In a 3-element array, which one you reverse can have some minor affects, which @MerlijnvanVeen has written about here: www.merlijnvanveen.nl/en/nl/studiezaal/135-which-sub-to-reverse
> a delay equivalent to the depth to the cabinet
That will be the case in MAPPXT, but in the real world it's a little longer because sound has to travel around the box, which is why you need to calibrate it in the rear in the field.
@@nathanlively Thank you very much!!!
Can I delay the front speaker and get the same result?
Hmmm...I think you would get the opposite result.
Excellent video! great info and well presented.
How about not flipping polarity on rear sub and delay only front sub?
Hey Piotr, that sounds like an end-fire array to me. Have you tried it?
@@nathanlively yup. Work flawelessly. I don’t quite understand Your method vs mine. Any toughts?
What if we want a 3 line array instead of 2 ?
Should we just reverse the 3rd one with no delay ? Making it kinda of a mix between cardioid and endfire ?
Or you have a better idea ?
Hey Paulo, the only way I know how to do gradient with an odd number of subs is the inverted gradient stack. If you want three in a line, I think you're better off with an end-fire design. But, please experiment and let us know what you find out! :)
Do we get the same result by delaying only the front sub, (by 1/4 wavelength,) and neither is reverse polarity? Perhaps simpler.
Hey Gordon, no, you would not get the same result. Everything in front would 90º out of phase and everything in the back would be a 180º out of phase, I think. Give it a shot in MAPP or in the real world and let me know your result.s
I don't have MAPP, but my RayEnd shows it working without reverse polarity. Several articles present it that way as well. One article in Sound Forums has yet another twist on fine tuning the array by stair stepping and tweaking the levels. Thanks a lot. I am going to try it this summer. soundforums.net/articles/stair-cardioid-subwoofer-array-192534/
@@nathanlively The technique Gordon refers to provides broadband summation in the front and a phase-based (frequency-dependent) cancellation in the rear. The technique you discuss in your video provides broadband (polarity-based) cancellation in the rear and frequency-dependent summation in the front. I showed the effects of both in this article: www.prosoundweb.com/tech-topic-friends-in-low-places/
Can I use another sub up to front sub?
You can try! :)
Using two matching subs is already a challenge because no manufacturer produces subwoofers that within 1dB of each other through the operating range. This is why 30dB of rejection is impossible in the field. So if you have two different subs, you'll have to spend some time making them match. Maybe try an FIR filter with auto magnitude?
What software is this?
MAPP XT
Always on point, thank u man👌🏻
Hello Nathan, how are you? I understand the concept and thank you very much for your videos, but what's throwing me off, I know low-frequencies are omni-directional, but aren't the front and rear speaker's waves performing a phase shit when they diffract backwards?
Hey Lorenzo, I'm doing alright. Thanks.
>rear speaker's waves performing a phase shit when they diffract backwards?
I don't think so, but I could be wrong. This might be true if you had something like an open back guitar cabinet, but then you would end up with a figure-8 polar pattern. Have you tried measuring a sub in the rear and found that it had a polarity inversion? What makes you think that sound waves going around an object would change their polarity?
PS. you should really see if you can register phaseshit.com :)
Hello Nathan, thank you for your reply. I just noticed I wrote phase shit instead of phase shift. This was just a thought I had since I was wondering if low-frequencies performed a slight phase shift to diffract to the back of their cabinet. I may be wrong. Unfortunately I don’t have the equipment to test this. Thank you again for your videos and for your reply. Have a good one :)
why can you delay the first one by one metre isntead of reversing rear one and delaying? Thnx
Hi Annzu. You can, but that would give you a different result. Try it and let me know what you come up with.
Simple í coupling 2 inductor coils like when design electronic always twin coils with control in middle so we had choice of up, down or neutral if connect two so induction automatic neutral inbound frequency should immediately react with negative outward
Hello I'm Sri Lankan. I want 30 or 37Hz Subwoofer how to delay time Rear Subwoofer? I'm use End Fire Cardioid Sub Technique. Please help me.
Hey Vishen, I'm not really clear on your question. Maybe you could draw a diagram of what you are talking about?
@@nathanlively I want to use Subwoofer as End Fire at 35Hz Frequency. If there is an equation to find out how much to delay the subwoofer on the back, can I explain it? Thank you.
Sorry im a bit slow... so one meter at 80hz would be 1/4 wave lengths righ. im trying to do a inverted cardio set up with out using DSP. just an old xover.
couldn't we even get more a perfect-like cardioid pattern over a large range of frequencies if we'd process the low-end frequencies by a multiband filter and apply different delays to each of them?
Hey Paul, the delay is not really the limitation to perfect cancellation. It's manufacturing tolerances. No manufacturer builds speakers that match to within 1dB. So if you want a custom performance increase, you might try micro surgery EQ until you can get your subwoofer response to match through the operating range to within 1dB.
Thank you. Explained it really really well. 🙏
You're welcome 😊
thanhkyou all
Awesome thank you that. was what I was looking for!
Nathan Thanks a lot for the nice explaination. But you considered just 86 Hz. With this we get full summation due to 360 Degrees Shift. But what about the other frequencies? Suppose we drive 26Hz to 100 Hz with the Subs. So how can i make sure complete Subwoofer spectrum has summation? is this 1 Meter sub differenze and 1 Sec Delay a constant? so that we can use in any ocation? Or do we have to calculate it targeting a certain Frequency? Fo instance u targeted 86 Hz here. The same do i have to do differently if i target 60 Hz?
> But you considered just 86 Hz.
Yes, this seemed like a good way to simply the discussion.
> But what about the other frequencies?
A gradient does have limitations on its usable bandwidth and you need to be away of these when it is deployed in consideration of the filter placement and level relationship. Please read this -> www.merlijnvanveen.nl/en/study-hall/167-lack-of-impact-revisited
> So how can i make sure complete Subwoofer spectrum has summation?
You can't, but you can plan for the bandwidth limitations using your design modeling software, the Subwoofer Array Designer, or with careful measurements in the field.
> is this 1 Meter sub differenze and 1 Sec Delay a constant?
Is that what I said? I thought I said 1m and 2.9ms of delay. 1 second would be about 345 meters. Either way, the answer is no. It will depend on the speed of sound.
>so that we can use in any ocation
Yes, I often use 1m and 3ms as a template in the field and then optimize from there.
> Or do we have to calculate it targeting a certain Frequency?
Yes, a design considering the maximum forward summation frequency is recommended, but if you want to work from a template, then just don't exceed 1m.
You are asking a lot of good questions. My challenge to you is to open up Subwoofer Array Designer, do some experiments for yourself, and let me know what you discover. Good luck! :)
Really helpful bro 👍
Thanks For Lovely Information 👍👌
Welcome
I'm not convinced this is correct. I'm not a trained acoustical engineer, just a musician who runs sound from stage.
But in my admittedly limited understanding, the only physical thing that happens to a speaker when the polarity is changed is that the speaker moves in the exact opposite direction of how it moved with correct polarity. In correct polarity, the initial speaker movement the moment sound is first produced is toward the FRONT of the speaker. When polarity is changed, that initial speaker movement is instead directed toward the rear of the speaker. That is the only physical, mechanical, thing that happens when polarity is reversed.
If I, instead of changing the initial speaker movement of a second enclosure in relation to the first one by changing it's polarity, simply rotate the second enclosure 180° in relation to the first enclosure, thereby causing the speaker in the second enclosure to move in the exact opposite direction of the first, which is the same physical motion caused by changing it's polarity, haven't I accomplished the exact same outcome? And with proper time allignment via delay, haven't I already accomplished the same goal of cancellation toward the stage and summation toward the audience?
Help me understand where I'm wrong.
Hi tlb2732! Thanks for your interest in the video. I have another video that talks about this, but it's possible that you will only be convinced by trying it yourself. Which is easy enough to do, even without an audio analyzer. With an audio analyzer, take a measurement of a sub (ideally outdoors), then rotate it 180º and measure it again. What do you observe? Without an audio analyzer, you could take two subs side by side with the same signal. Turn one one, turn the other on. They should sum. Invert the polarity of one, they should cancel. Now restore normal polarity and rotate one of them 180º and try the same test.
ua-cam.com/video/G0l6yniHslo/v-deo.html
A job well done but not all system react the same so there must be a standard or must have calibration engineer to calibrate each system before use and that is a good enough job for AT or ET
The last bit was wrong surely? Delay went wrong way
Hey Underflip, thanks for checking on that. Could you give me the time stamp?
I believe Underfilip meant that when you were explaining how it works behind, you moved the wave to the wrong direction. Delay doesn’t make it sound “faster”…
🙌🙌🙌