Hopefully there is a simple answer to this, but I suspect it's going to be very complex. 🤐
I've never worried about delaying any boxes in my PA, but I know that lots of guys delay there subs a bit, especially if they are "in line" with the tops.
I have the ability to do this is ms increments with my current electronics, so I want to experiment. - But I assume there's a better way to do it than just by ear, yes?
So what's the common procedure for this?
I have Winspeakerz, TrueRTA, REW, and WinISD, but I have not had time to learn them well, past BR box design and room tuning. I assume one of them will do what I need, but I'll buy something else if you suggest it.
My PA is currently two point source speakers, typically about 8 -10 feet apart and 6 -7 feet high, with subs on the floor and clustered in the center. For smaller rooms I use four sealed 18", for larger rooms I use two (soon to be three) dual 12" BR cabs.
Crowd size is the typical "local gig and weddings" thing. anywhere from 30 to maybe 100 people.
-----------------------------------------
Also, how do you deal with the fact that there is a different physical distance between the tops & subs, for a close audience vs the far audience? Do you just pick the mid-point and calculate for that? - Or aim for the dance floor?
Thanks as always for your time & expertise !
I've never worried about delaying any boxes in my PA, but I know that lots of guys delay there subs a bit, especially if they are "in line" with the tops.
I have the ability to do this is ms increments with my current electronics, so I want to experiment. - But I assume there's a better way to do it than just by ear, yes?
So what's the common procedure for this?
I have Winspeakerz, TrueRTA, REW, and WinISD, but I have not had time to learn them well, past BR box design and room tuning. I assume one of them will do what I need, but I'll buy something else if you suggest it.
My PA is currently two point source speakers, typically about 8 -10 feet apart and 6 -7 feet high, with subs on the floor and clustered in the center. For smaller rooms I use four sealed 18", for larger rooms I use two (soon to be three) dual 12" BR cabs.
Crowd size is the typical "local gig and weddings" thing. anywhere from 30 to maybe 100 people.
-----------------------------------------
Also, how do you deal with the fact that there is a different physical distance between the tops & subs, for a close audience vs the far audience? Do you just pick the mid-point and calculate for that? - Or aim for the dance floor?
Thanks as always for your time & expertise !
Last edited:
I have no idea why you would want to delay the subs if everything is in "in line". AFAIK the only reason you would want to delay the subs is if they are in front of the other cabs towards the audience. This is less and less important as the crossover frequency falls below 100Hz. But for PA work I assume the crossover point is higher.
The purpose of delaying one driver or band compared to the others is to better align the acoustic wavefront. Sound travels at about 343 m/sec (1100 feet/sec), so you just do the math to figure out what delay is necessary based on the separation distance of the acoustic sources. Roughly speaker it is about 1 millisecond per foot. Which ever source is closer to the listener should be delayed by this amount. But... keep in mind that the relative delay remains the same regardless of the listening distance when subs are below the tops. But that is not how you are setting up. With subs in the center the relative distance from subs to tops depends on where in the room you are located and how far away. Honestly, with that sort of setup and if you are doing "gigs and weddings", I would not even bother to do any time alignment unless you re-locate the subs very near the tops.
The purpose of delaying one driver or band compared to the others is to better align the acoustic wavefront. Sound travels at about 343 m/sec (1100 feet/sec), so you just do the math to figure out what delay is necessary based on the separation distance of the acoustic sources. Roughly speaker it is about 1 millisecond per foot. Which ever source is closer to the listener should be delayed by this amount. But... keep in mind that the relative delay remains the same regardless of the listening distance when subs are below the tops. But that is not how you are setting up. With subs in the center the relative distance from subs to tops depends on where in the room you are located and how far away. Honestly, with that sort of setup and if you are doing "gigs and weddings", I would not even bother to do any time alignment unless you re-locate the subs very near the tops.
Hey, Charlie.
FWIW, my x-over (subs to tops) is at 80 Hz.
I thought the purpose was because LF takes longer to develop, so it's acoustically a bit behind the tops, even if they are all touching each other.
No?
-------------
"keep in mind that the relative delay remains the same regardless of the listening distance when subs are below the tops."
Not really. If someone is standing 10' feet away, the tops are much closer to their ears than the subs. If someone is standing 50' away, the tops and subs are pretty much equi-distant.
Granted, this difference may not be significant with a 7' long wavelength, (one octave above my xover point) but maybe it is. Hence my question.
FWIW, my x-over (subs to tops) is at 80 Hz.
I thought the purpose was because LF takes longer to develop, so it's acoustically a bit behind the tops, even if they are all touching each other.
No?
-------------
"keep in mind that the relative delay remains the same regardless of the listening distance when subs are below the tops."
Not really. If someone is standing 10' feet away, the tops are much closer to their ears than the subs. If someone is standing 50' away, the tops and subs are pretty much equi-distant.
Granted, this difference may not be significant with a 7' long wavelength, (one octave above my xover point) but maybe it is. Hence my question.
Last edited:
Even if:
If the LF takes longer to develop it is because that is built into the music signal, plus the way that a bandlimited audio device behaves when the passband does not extend very high in frequency. In short, because of the 80 Hz crossover the subs are "slow". That's perfectly normal. Delaying the subs would just make things worse.
I said:
"keep in mind that the relative delay remains the same regardless of the listening distance when subs are below the tops."
You counter with:
But getting back to what you said:
Does this make sense to you?
why you want to delay the subs more, making them even more "behind"? Where is the logic in that?
If the LF takes longer to develop it is because that is built into the music signal, plus the way that a bandlimited audio device behaves when the passband does not extend very high in frequency. In short, because of the 80 Hz crossover the subs are "slow". That's perfectly normal. Delaying the subs would just make things worse.
I said:
"keep in mind that the relative delay remains the same regardless of the listening distance when subs are below the tops."
You counter with:
I see your point, but that's not the "relative" I was thinking about. By relative I mean the difference between the time sound travels between the source and the listener will be exactly the same as the listener moves farther away from the sources when they are vertically aligned and compact. If at a distance of 10 feet the sub sounds arrive 1 millisecond after the tops, at 100 feet the sub sounds are still arriving 1 millisecond after the sound from the tops. You might be confusing this with the relative SPL levels as a function of distance. These grow more and more similar as you move farther away because of how SPL scales with distance. But delay DOESN'T scale with distance, at all, because the velocity of sound is constant and not dependent on frequency. Unless your tops are really high above the crowd and subs at the ground, what you are concerned wtih is not all that important.
But getting back to what you said:
In this case (I assume that the subs are below the tops) the subs would be further away from the ears, by geometry. Farther away means more delay, so delaying them more makes the situation worse. And let's say you correct the delay so that the tops and subs are "aligned" for people at 10 feet distance, well then they will just be "misaligned" for people at 100 feet because now the relative distance to subs and tops has changed.
Does this make sense to you?
Last edited:
That's a good point. (Where did my brain go?)
- And yet I keep reading about this being done. Maybe it's explained wrong, and they are delaying the tops?
Please let me know if you find out the reasoning behind doing that.
From what I have come to understand about psycho-acoustics, the minimum delay that humans can detect at low frequencies is around 1 period or slightly less. At 100Hz one period corresponds to 1/100th second or 10 milliseconds, 50Hz 20 milliseconds, etc. At 1 foot per millisecond that is tens of feet of difference, meaning that unless the relative distance is more than 10 feet@100Hz / 20 feet@200Hz you don't perceive the delay.
From what I have come to understand about psycho-acoustics, the minimum delay that humans can detect at low frequencies is around 1 period or slightly less. At 100Hz one period corresponds to 1/100th second or 10 milliseconds, 50Hz 20 milliseconds, etc. At 1 foot per millisecond that is tens of feet of difference, meaning that unless the relative distance is more than 10 feet@100Hz / 20 feet@200Hz you don't perceive the delay.
Thanks, Charlie.
I guess I won't worry about it unless someone else posts a strong opinion to the contrary.
FWIW - I found a pretty technical article that (in part) deals with this subject, written by Linkwitz Lab, but I can't understand most of it:
https://www.linkwitzlab.com/frontiers_5.htm
I guess I won't worry about it unless someone else posts a strong opinion to the contrary.
FWIW - I found a pretty technical article that (in part) deals with this subject, written by Linkwitz Lab, but I can't understand most of it:
https://www.linkwitzlab.com/frontiers_5.htm
I've read through this thread and could not understand why people would delay their subs if they were physically aligned with the tops.
Maybe you could post a few links to sites where you have read this so that people on this forum can see where you are coming from ?
Rob.
I have a 3ms delay on my mains with the drivers stacked on top of the subs all in vertical line, it makes a big difference in focus/clarity.
Lots of guys do the wrong thing.
They can phase align the subs a cycle or two or more behind by adding delay to them.
Merlijn Van Veen has covered the alignment process in great detail, check it out:
https://www.merlijnvanveen.nl/en/st...gnment-the-foolproof-relative-absolute-method
That said, if you put your subs in the center of a stereo PA, they can only aligned in one location.
Art
Thanks, Art. Yeah I meant to write that they delay their TOPS a bit. My memoery 'aint what it used to be. (or maybe it is, I can't quite remember....)
I'll read that Van Veen article tonight. (Although I suspect it's going to give me a nasty headache. 😳 )
I'll read that Van Veen article tonight. (Although I suspect it's going to give me a nasty headache. 😳 )
One thing to consider is where to place the subwoofer(s) to achieve the best frequency response at the listening position. If you don't consider this, any steps you take to align the subs with the mains will be pointless because the room modes will overwhelm any benefit gained by aligning the subs. Of course, the easiest way to do this is to place the subwoofer at the listening position and then find the place at the front of the room, or wherever you want to place the subwoofer, with the flattest frequency response (it is easier to move yourself than the subwoofer). If you have multiple subs, you can place those in the best locations to achieve a flat frequency response in the bass. Then you can worry about aligning the subwoofer with the mains.
Another thing to consider is if you are using an inductor to low pass the subwoofer, it will change the phase at the crossover point, and change how you align the subwoofer with the mains. The crossover can affect the mains as well, which also can affect the alignment with the subwoofer. If you are using a second order crossover for both your subwoofer and mains at the crossover point between the two, the signals will be out of phase at the crossover point and this can be fixed by swapping the speaker wires for your subwoofer to make the crossover phase consisistent.
One more thing to consider is whether the amplifier changes the phase of the signal to the subwoofer (some subwoofers do this), providing that that amplifier for the subwoofer is different than the one used for the mains.
Getting all this worked out is worth it in much better quality bass.
Retsel
Another thing to consider is if you are using an inductor to low pass the subwoofer, it will change the phase at the crossover point, and change how you align the subwoofer with the mains. The crossover can affect the mains as well, which also can affect the alignment with the subwoofer. If you are using a second order crossover for both your subwoofer and mains at the crossover point between the two, the signals will be out of phase at the crossover point and this can be fixed by swapping the speaker wires for your subwoofer to make the crossover phase consisistent.
One more thing to consider is whether the amplifier changes the phase of the signal to the subwoofer (some subwoofers do this), providing that that amplifier for the subwoofer is different than the one used for the mains.
Getting all this worked out is worth it in much better quality bass.
Retsel
Most of the people that delay the tops in PA is because they use TH subs, even if you put the tops over the subs cabinets ,the output of the 2 cabinets will not sum in time cause the TH have the horn path, so sound will need to travel a bigger distance, i use 2 to 3ms delay when using my horns, BR do not need delay if you stack the tops over your subs.
In the other hand even if you use your BR subs but the are not aligned with the tops , example of this is on a 3 feet high stage were tops are on the sides of it and subs ground level but on front of the stage not flushed below.
There are several ways to time align, with REW, with a SPL meter and by ear just to name a few.
Personally I play by ear most of the time but sometimes I like to pull the SPL meter.
I do the phase flip trick and throw a sine at the xover freq of the subs.
Hope this help.
In the other hand even if you use your BR subs but the are not aligned with the tops , example of this is on a 3 feet high stage were tops are on the sides of it and subs ground level but on front of the stage not flushed below.
There are several ways to time align, with REW, with a SPL meter and by ear just to name a few.
Personally I play by ear most of the time but sometimes I like to pull the SPL meter.
I do the phase flip trick and throw a sine at the xover freq of the subs.
Hope this help.
That's probably the easiest way - flip POLARITY on either the sub or the mains, adjust TD on the mains for largest null at the location you want it to sound the best, then flip it back.
More advanced: Adjust x-over/filter settings so the null is symmetrical, and isn't replaced by a peak when you flip polarity back 🙂
Yes, there is a simple answer and simple method. But it is specific to "small" rooms. I don't know how big of a room it will work in.
The idea is to use delays combined with room reflections to get many room modes, which results in "even" bass response. This is my pet project!
The method is to take measurements in REW using a timing reference. Then use the time alignment tool to calculate the delays. To find the tool, you go to the "all spl" tab, click tools menu, and select "alignment tool". In the tool you select the 2 measurements you want to align, and it calculates the resulting combined response in real time so you can see the result as you adjust the delays. Or you can align at a specific frequency if you want.
Keeping in mind that I'm doing this in my living room....I get my highs and mids dialed, then align 1 sub to the highs and mids. Then align the second sub to the first sub. And then align the 3rd to the first two, and so on. I find it's helpful to start with a small delay on all channels, because REW might give you a negative delay. And you might end up inverting something too. You just have to play with it until you get something you like.
I am doing this with what is in effect a pair of DIY cardiod subwoofers...but not being operated as cardiods. They have a front facing 12" woofer (sealed) and a rear facing 10" woofer (ported). Stacking subs with one front facing and one rear facing is essentially the same thing.
More thoughts on this issue:
I recently decided to try and figure out if there's an "ideal" physical sarting position, before dealing with time alignment. Based on some rough diagrams I made, there definitely is:
This is for my particular setup, where the tops will always be firing at about 7' above the floor, the non-horn subs centered and low, and only calculating for an audience "in front" of the stage. Looking at the physical distance of tops vs subs, for people both standing and sitting and at distances of ten, twenty, and 30 feet.
The setup that seems to have the least variation, overall, is with the subs firing about 2.5' more forward than the tops. (This is ballpark, and only rough-calculated.) So it seems to me that, no matter how I compensate with delay or phase, this is where I should place my speakers. I can then correctly compensate for that entire area. (room reflections notwithstanding.)
-----------------------------------------
EXCEPT: Let's say that setup is problematic for ergonomic reasons. (This is possible in some smaller rooms) and I need to put the tops as far forward as the subs. Thus I can only time-delay for one area of the audience. It raises the question:
HOW MUCH TIME DIFFERENCE IS REALLY AUDIBLE?
With an 80 Hz crossover to my tops, is it really going to matter if the LF hits a listener's ear 2ms sooner than the HF ?
I recently decided to try and figure out if there's an "ideal" physical sarting position, before dealing with time alignment. Based on some rough diagrams I made, there definitely is:
This is for my particular setup, where the tops will always be firing at about 7' above the floor, the non-horn subs centered and low, and only calculating for an audience "in front" of the stage. Looking at the physical distance of tops vs subs, for people both standing and sitting and at distances of ten, twenty, and 30 feet.
The setup that seems to have the least variation, overall, is with the subs firing about 2.5' more forward than the tops. (This is ballpark, and only rough-calculated.) So it seems to me that, no matter how I compensate with delay or phase, this is where I should place my speakers. I can then correctly compensate for that entire area. (room reflections notwithstanding.)
-----------------------------------------
EXCEPT: Let's say that setup is problematic for ergonomic reasons. (This is possible in some smaller rooms) and I need to put the tops as far forward as the subs. Thus I can only time-delay for one area of the audience. It raises the question:
HOW MUCH TIME DIFFERENCE IS REALLY AUDIBLE?
With an 80 Hz crossover to my tops, is it really going to matter if the LF hits a listener's ear 2ms sooner than the HF ?
Last edited:
A related question regards an interesting idea I recently read about, where some folks put the subs BEHIND the band, against the back wall.
The idea is that you minimize out of phase reflections off that wall, plus you have the ability to make a center cluster that isn't visually in the way.. You can even make a vertically stacked center cluster, for a wider horizontal dispersion with slightly less bounce off of the ceiling. (Which of course is the closest boundary.)
- And since most subs are omni directional, this doesn't much change how much LF energy is on stage.
The BIG possible negative of this setup is that it makes the physical distance to the audience, of subs vs tops, much more different as you go from close audience to far audience. You'd have to time-delay for the middle of the audience, and then hope for the best in the front & back.
So again, how much of a difference is acceptable, with an 80 Hz crossover?
The idea is that you minimize out of phase reflections off that wall, plus you have the ability to make a center cluster that isn't visually in the way.. You can even make a vertically stacked center cluster, for a wider horizontal dispersion with slightly less bounce off of the ceiling. (Which of course is the closest boundary.)
- And since most subs are omni directional, this doesn't much change how much LF energy is on stage.
The BIG possible negative of this setup is that it makes the physical distance to the audience, of subs vs tops, much more different as you go from close audience to far audience. You'd have to time-delay for the middle of the audience, and then hope for the best in the front & back.
So again, how much of a difference is acceptable, with an 80 Hz crossover?
Last edited by a moderator:
At around 100hz, the threshhold for audibility is around 50ms. This is according to Geddes.
This will be a hotly debated topic, so I'd say take Geddes' opinion and run with it. In the anecdote he was talking about is the delay applied to a subwoofer via DSP, which would be separate from any group delay inherent to the subwoofer.
We hear low frequencies different from high frequencies. This is why we can get away with taking a stereo signal and crossing it to a mono sub at ~100hz. And why we can place multiple subs around a room and not notice any issues.
I think it is far more important as a starting point to indentify if you have an issue of room interaction at low frequencies. You will notice very uneven bass response, like +- 20dB when doing a frequency sweep. If you do have a room problem, moving the subs won't fix it.
If you have the classic "small" room issues, you need multiple subs + DSP. You can use Multiple Sub Optimizer (MSO) if you choose to spread the subs throughout the room. Other solutions are possible, but you're more or less on your own to implement them.
Speaker delays are measured in milliseconds (ms).
Please refer to the following video. It starts with Geddes being asked about audibility of delays. He says the threshhold of hearing a delay at 100hz or below is 50ms...or 30-50ms...he says both in his answer. Point is low frequencies don't fit in a small room, so they play by different rules.
I love this video because he directly contradicts a bunch of audiophile/ audio beliefs. And if you don't know who Geddes is, it is worth looking him up. Him and Floyd Toole. They are the OGs of a lot of audio science.
Please refer to the following video. It starts with Geddes being asked about audibility of delays. He says the threshhold of hearing a delay at 100hz or below is 50ms...or 30-50ms...he says both in his answer. Point is low frequencies don't fit in a small room, so they play by different rules.
I love this video because he directly contradicts a bunch of audiophile/ audio beliefs. And if you don't know who Geddes is, it is worth looking him up. Him and Floyd Toole. They are the OGs of a lot of audio science.