Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Ryan C. Davis]

Someone smarter than me want to educate me on speaker time alignment?

I’ve got a pair Danley SH50’s used in a stereo config, and i just picked up 4 TH115’s I’m using in pairs as left and right subs. The amp is a danley dna20k. My sh50’s are setup on some scaffold and the Subs are on the ground about 6’ in front of the mains.

When i first setup the subs i was totally underwhelmed. No deep bass. I tried inverting polarity and playing with crossover points but it still sounded really bad on the low end. Crossover point is 80Hz at 24 dB/ octave BW. After messing around with about everything under the sun I tried delaying the subs by 6’ and everything came to life. I knew the time misalignment would impact the fidelity but  I really didn’t think I’d see so much interaction between the mains and subs that far below the crossover point. I’m stunned.

Maybe there’s nothing to explain but hopefully there’s something I can learn from it all.


Sent from my iPhone using Tapatalk

[Tim McCulloch]

Someone smarter than me want to educate me on speaker time alignment?

I’ve got a pair Danley SH50’s used in a stereo config, and i just picked up 4 TH115’s I’m using in pairs as left and right subs. The amp is a danley dna20k. My sh50’s are setup on some scaffold and the Subs are on the ground about 6’ in front of the mains.

When i first setup the subs i was totally underwhelmed. No deep bass. I tried inverting polarity and playing with crossover points but it still sounded really bad on the low end. Crossover point is 80Hz at 24 dB/ octave BW. After messing around with about everything under the sun I tried delaying the subs by 6’ and everything came to life. I knew the time misalignment would impact the fidelity but  I really didn’t think I’d see so much interaction between the mains and subs that far below the crossover point. I’m stunned.

Maybe there’s nothing to explain but hopefully there’s something I can learn from it all.


Sent from my iPhone using Tapatalk

Delay the SH50s, not the subs.

[Michael Lascuola]

I'm not sure why you'd want to delay the tops if they are behind the subs.

"I really didn’t think I’d see so much interaction between the mains and subs that far below the crossover point."

What do you mean that far below the xover point?

[Len Zenith Jr]

I'm not sure why you'd want to delay the tops if they are behind the subs.

The sound comes out of the tops waay before the subs. The horn path of the TH115 has to be over 7ish feet long and the phase lag on large woofers with big inducting voice coils is greater than small woofers with smaller inducting voice coils.

Counter intuitive but it is the same reason you need to delay the horn on a 2-way box and not the woofer even though the woofer is about 6" in front of the compression driver.

[Tim McCulloch]

I'm not sure why you'd want to delay the tops if they are behind the subs.

That's a mighty big "IF" you've got there, pal. 

As Len Z points out there is path length to the horn and even if there weren't, there is still group delay from the subwoofer processing as well as group delay from the inertia of the sub transducer itself.

Now the homework question:  Ryan says his "deep bass" improved when he delayed the subs by 6ms.  I don't doubt that it sounded better to him where he was standing at the time.  Does anyone want to venture an explanation?  Hint:  it involves lots of "wheel trips".

My question for Ryan - how did the subs sound by themselves, without the tops?

[Ivan Beaver]

Someone smarter than me want to educate me on speaker time alignment?

I’ve got a pair Danley SH50’s used in a stereo config, and i just picked up 4 TH115’s I’m using in pairs as left and right subs. The amp is a danley dna20k. My sh50’s are setup on some scaffold and the Subs are on the ground about 6’ in front of the mains.

When i first setup the subs i was totally underwhelmed. No deep bass. I tried inverting polarity and playing with crossover points but it still sounded really bad on the low end. Crossover point is 80Hz at 24 dB/ octave BW. After messing around with about everything under the sun I tried delaying the subs by 6’ and everything came to life. I knew the time misalignment would impact the fidelity but  I really didn’t think I’d see so much interaction between the mains and subs that far below the crossover point. I’m stunned.

Maybe there’s nothing to explain but hopefully there’s something I can learn from it all.


Sent from my iPhone using Tapatalk

Do you have any measurements?

There are many things that can affect the sound quality/perception.

Physical placement in the room is a big one.  Room modes will easily "overpower" any loudspeaker.

Why are the subs so far in front of the mains?

Proper setup is a matter of "divide and conquer".  First of all, make SURE that all of the wiring is correct.  I often see a polarity reversal, even from people who have done it for years. 

Make sure all the processing in the amps is correct, polarities, delays etc.

It seems really strange that you delayed the subs 6ms, while they are already about 6ms ahead of the mains.

The actual HP and LP freq will affect the delay times.

The listening position in the room can be a issue, due to room modes and physical placement of the cabinets.

The best way is a deliberate course of actions, not pressing buttons and hoping for a good outcome.

Depending on the system configuration (subs on an aux for example), you may want to raise the HP of the SH50s, or lower the LP on the subs.

There are many variable that can affect the overall outcome.  Some of which you can control, others you cannot do anything about.

[Ryan C. Davis]

I'm not sure why you'd want to delay the tops if they are behind the subs.

"I really didn’t think I’d see so much interaction between the mains and subs that far below the crossover point."

What do you mean that far below the xover point?

Well I haven’t done any testing with TEF yet, this was all seat of the pants. But the upper bass was decent, the lower bass-50 ish hz is what seemed so light. The reason I said this is the slope of the high pass is steep enough that I figured the top boxes would have almost no interference with the bottom boxes at 50 Hz. In my logic that interference was going to affect the upper bass way more than the deeper bass.


Sent from my iPhone using Tapatalk

[Tim McCulloch]

Well I haven’t done any testing with TEF yet, this was all seat of the pants. But the upper bass was decent, the lower bass-50 ish hz is what seemed so light. The reason I said this is the slope of the high pass is steep enough that I figured the top boxes would have almost no interference with the bottom boxes at 50 Hz. In my logic that interference was going to affect the upper bass way more than the deeper bass.


Sent from my iPhone using Tapatalk

One would think.  Go back and check Ivan's reply.

I think there was something else at play in addition to the top/sub alignment.

[Ryan C. Davis]

Do you have any measurements?

There are many things that can affect the sound quality/perception.

Physical placement in the room is a big one.  Room modes will easily "overpower" any loudspeaker.

Why are the subs so far in front of the mains?

Proper setup is a matter of "divide and conquer".  First of all, make SURE that all of the wiring is correct.  I often see a polarity reversal, even from people who have done it for years. 

Make sure all the processing in the amps is correct, polarities, delays etc.

It seems really strange that you delayed the subs 6ms, while they are already about 6ms ahead of the mains.

The actual HP and LP freq will affect the delay times.

The listening position in the room can be a issue, due to room modes and physical placement of the cabinets.

The best way is a deliberate course of actions, not pressing buttons and hoping for a good outcome.

Depending on the system configuration (subs on an aux for example), you may want to raise the HP of the SH50s, or lower the LP on the subs.

There are many variable that can affect the overall outcome.  Some of which you can control, others you cannot do anything about.

Understood, I double checked the wiring to each 115 and the polarity. And I’d inverted the polarity of the subs in the processing - tried it both ways.

I’m pretty new to delays- you say it seems strange that I would delay the subs when they are already ahead of the mains. This doesn’t make sense to me and I could use a little explanation.

If the subs are physically 6’ in front of the mains (not accounting for any physics pathway or group delays) it seems to make perfect sense that I’d want to delay the subs so their arrival time is matching the mains.


Sent from my iPhone using Tapatalk

[Mac Kerr]

I’m pretty new to delays- you say it seems strange that I would delay the subs when they are already ahead of the mains. This doesn’t make sense to me and I could use a little explanation.

If the subs are physically 6’ in front of the mains (not accounting for any physics pathway or group delays) it seems to make perfect sense that I’d want to delay the subs so their arrival time is matching the mains.

Not accounting for any physics or group delay it would seem straightforward, unfortunately we do have to account for them. Subs in general appear to be delayed because of phase shift that occurs due to the nature of low pass filters, and horn subs have the additional "delay" of the horn path length.

Mac

[Ryan C. Davis]

Not accounting for any physics or group delay it would seem straightforward, unfortunately we do have to account for them. Subs in general appear to be delayed because of phase shift that occurs due to the nature of low pass filters, and horn subs have the additional "delay" of the horn path length.

Mac

So is the only way to get the delay right to measure? No assumptions? If you were to make assumptions and a guess would you say delay the mains or the subs?


Sent from my iPhone using Tapatalk

[Mac Kerr]

So is the only way to get the delay right to measure? No assumptions? If you were to make assumptions and a guess would you say delay the mains or the subs?

You can start by getting the subs closer to the mains. Is there a reason they are not under the scaff? Once you have them close together any alignment you do will work for a larger part of the coverage area. Once they are close together you can try playing a tone that is in the center of the crossover region with the subs out of polarity. Then delay the tops till you get a good null and the tone cancels out. If you have gone farther than 1 wavelength of the tone (about 10' or 10ms at 10Hz), start over and delay the subs. Wherever you get the best null with the least amount of delay is as good a rough guess as you are likely to get. Set the polarity of the subs back to normal and give it a listen.

Best method is to look at the phase trace in an FFT analyzer and adjust till the phase slop matches.

Mac

[Len Zenith Jr]

you say it seems strange that I would delay the subs when they are already ahead of the mains. This doesn’t make sense to me and I could use a little explanation.

Voice coils don't move until there is a magnetic field induced in them. That magnetic field is created by the CURRENT flowing in them, not VOLTAGE. In an AC circuit with an inductor (low pass filter, voice coil, etc) the current lags behind the voltage in time because inductors resist change in current flow.

Think of a water wheel spinning with the flow of a river passing underneath. Now we very quickly reverse the flow of the river to go the other way. It takes a bit of time for the water wheel to stop and change directions due to the inertia of the water wheel. That water wheel is your inductor. The bigger it is the longer it takes to change directions.

Likewise, AC circuits with a capacitor in them (high pass filter), the opposite happens and VOLTAGE lags behind CURRENT because the voltage doesn't get to maximum until the capacitor fills up with charge. Think of this one like a bucket and a hose. The hose (current) flows  normally but the voltage (height of water in the bucket) doesn't reach maximum until the bucket is full. The bigger the capacitor the larger the bucket.

[Tim McCulloch]

So is the only way to get the delay right to measure? No assumptions? If you were to make assumptions and a guess would you say delay the mains or the subs?


Sent from my iPhone using Tapatalk

You're wanting a definitive when Mac gave you the info you needed to DIY.  Did you *comprehend* our responses?  I'm asking because if we need to drill down in our explanations, we can.  On the internet it's hard to tell when the tech level is too little, too much or just right.

For subs with a fairly steep HPF filter you can *usually* count on a later arrival than the LF pass band in the accompanying top box from the sub (and processing) group delay alone.  Note that it is the filter itself that creates the initial delay and it doesn't matter if that filter has big capacitors and inductors (passive) or is done in DSP, the group delay is the same.

Story time -

A couple of years or so ago we had a BE that was having a rough time doing his system alignment on one of our rigs.  We align them but many BEs want to do their own and we certainly encourage them to do so.  Anyway, he finally said "man, I think that cardioid sub is messing with my measurements" so I muted the reversed subs for him.  A few minutes late he was still puzzled.  "I can't find a delay time for the subs."  "Uh, lets take a look at your impulse response instead of phase response..."  We did separate measurements of subs & tops and low and behold, the tops were *ahead* of the subs by almost 5ms (and the subs were physically closer to the measurement mic by about 1 ft.).  I'm not sure he'd ever looked at impulse response in Smaart before... And this was on a rig with conventional bass reflex subwoofers.

/Story time.

I knew what was wrong as I was watching him make his measurements but it was better to guide him to enlightenment than simply say "put 4.6 ms delay on the left & right".

Have fun, good luck.

Tim Mc

[Scott Holtzman]

Voice coils don't move until there is a magnetic field induced in them. That magnetic field is created by the CURRENT flowing in them, not VOLTAGE. In an AC circuit with an inductor (low pass filter, voice coil, etc) the current lags behind the voltage in time because inductors resist change in current flow.

Think of a water wheel spinning with the flow of a river passing underneath. Now we very quickly reverse the flow of the river to go the other way. It takes a bit of time for the water wheel to stop and change directions due to the inertia of the water wheel. That water wheel is your inductor. The bigger it is the longer it takes to change directions.

Likewise, AC circuits with a capacitor in them (high pass filter), the opposite happens and VOLTAGE lags behind CURRENT because the voltage doesn't get to maximum until the capacitor fills up with charge. Think of this one like a bucket and a hose. The hose (current) flows  normally but the voltage (height of water in the bucket) doesn't reach maximum until the bucket is full. The bigger the capacitor the larger the bucket.

Len, you ever use Eli the Iceman mnemonic to remember this?  Voltage (e) leads current (i) in an inductor (l)

[Ryan C. Davis]

You can start by getting the subs closer to the mains. Is there a reason they are not under the scaff? Once you have them close together any alignment you do will work for a larger part of the coverage area. Once they are close together you can try playing a tone that is in the center of the crossover region with the subs out of polarity. Then delay the tops till you get a good null and the tone cancels out. If you have gone farther than 1 wavelength of the tone (about 10' or 10ms at 10Hz), start over and delay the subs. Wherever you get the best null with the least amount of delay is as good a rough guess as you are likely to get. Set the polarity of the subs back to normal and give it a listen.

Best method is to look at the phase trace in an FFT analyzer and adjust till the phase slop matches.

Mac

This makes sense to me, thank you Mac!


Sent from my iPhone using Tapatalk

[Ryan C. Davis]

You're wanting a definitive when Mac gave you the info you needed to DIY.  Did you *comprehend* our responses?  I'm asking because if we need to drill down in our explanations, we can.  On the internet it's hard to tell when the tech level is too little, too much or just right.

For subs with a fairly steep HPF filter you can *usually* count on a later arrival than the LF pass band in the accompanying top box from the sub (and processing) group delay alone.  Note that it is the filter itself that creates the initial delay and it doesn't matter if that filter has big capacitors and inductors (passive) or is done in DSP, the group delay is the same.

Story time -

A couple of years or so ago we had a BE that was having a rough time doing his system alignment on one of our rigs.  We align them but many BEs want to do their own and we certainly encourage them to do so.  Anyway, he finally said "man, I think that cardioid sub is messing with my measurements" so I muted the reversed subs for him.  A few minutes late he was still puzzled.  "I can't find a delay time for the subs."  "Uh, lets take a look at your impulse response instead of phase response..."  We did separate measurements of subs & tops and low and behold, the tops were *ahead* of the subs by almost 5ms (and the subs were physically closer to the measurement mic by about 1 ft.).  I'm not sure he'd ever looked at impulse response in Smaart before... And this was on a rig with conventional bass reflex subwoofers.

/Story time.

I knew what was wrong as I was watching him make his measurements but it was better to guide him to enlightenment than simply say "put 4.6 ms delay on the left & right".

Have fun, good luck.

Tim Mc

Tim- awesome, The story is helpful and this makes more sense now.

My TEF skills are a little rusty but I’ll bust it out and take some measurements. I’ll probably mess around with macs method too. Super clever, love it, makes perfect sense.

My tech skills are pretty solid- I have an engineering degree and worked in speaker design for several years when I was in my 20’s. No further explanation needed, just trying to learn something I’ve never messed with before.

Thanks everyone for sharing your knowledge.

Oh, BTW, this system is in my shop- i do sound gigs on the side but I’m not a full time (live) sound man. I do permanent installs. The SH50’s are on some sections of pallet racking and the subs are that far head of the SH50’s because I don’t have any room in the racking to push the 115’s in to match the 50’s. It’s not going to stay like this permanently, I was just excited to hear all of the speakers together so I put the subs where I could. Turned out to be a good way to further my knowledge.


Sent from my iPhone using Tapatalk

[Tim McCulloch]

Ryan-

I suspect you were getting a cancellation from the distance between subs & wall.  The delay you applied "moved" the subs further from the walls relative to the tops.

[Ivan Beaver]

Ryan-

I suspect you were getting a cancellation from the distance between subs & wall.  The delay you applied "moved" the subs further from the walls relative to the tops.

Yes.

Physical positioning of cabinets with relation to each other and physical barriers is one of the harder things to deal with.

You can use delay to adjust some things, but there are others that the physical realm simply takes over and dominates.  There is no adjusting for that-except physical positioning.

[Kevin McDonough]

hey

yeah as people have said, group delay of low frequency drivers on its own usually means the sound comes out a little "slower" than the accompanying top, and when they're pretty close together it's almost always the top that needs delayed to match the subs. Add in the horn path that the sound inside the tapped horn needs to travel round, and the subs and tops could be quite a big physical distance apart and it'd still be the tops that needed delayed.

If you have somewhere you can go, I'd suggest making a basic starting preset where your speakers are time aligned to each other when physically next to each other. Find a big open outdoor area and perform ground plane measurements on each speaker. If you have Smaart, Systune, SatLive etc (or know someone who can help) you can use that, or there is the free Room Eq Wizard (REW) which will also let you look at the phase traces of the speakers.

Set both speakers (sub and top) on the ground next to each other. Put the mic a metre or two away, and adjust the levels so that they're sitting at the normal level you would set them up to. This could be done by looking at the measurement screen with some pink playing (some people like the subs to be a few db above the tops 3 or maybe 6. People doing lots of EDM or low frequency HipHop more), or you could play a well known track and balance it to what you think sounds good.

Take a first basic frequency range measurement of each speaker, sub first (with the top muted) and then top (with the sub muted). This is just so you've got the acoustic crossover roughly in the correct place. The point where these two lines crossover is the actual acoustic crossover point of the speakers, based on the SPL you're intending to play them at in relation to each other. This may well be quite different from your "electronic" crossover, i.e. the points you've set the crossover filters to in your speaker processor (which after all are just fancy EQ points, not where the sound actually crosses over from one speaker to another as such).

Now you're gonna change to looking at the phase graph. Whatever frequency point those two lines crossed over at is your area of interest.

Use the delay finder in the software to account for the sub's delay time, and then maybe tweek it by a few ms either way so that you're getting a fairly flat phase line at the top end of the sub's range, just at this crossover point you've identified. Then play the top speaker, and delay it until your getting the same flat line at the bottom end of it's range, where they overlap, and make sure both phase traces match each other as closely as possible in that crossover region.

That'll be your basic delay time set so that you know they're in phase when together.

After this, now that they are in phase, you'll probably notice a change in sound and want to make a few EQ or crossover tweeks. Put the speakers on stands or however you normally set them up, and adjust the EQ or corssovers slightly. However unless you're using FIR filters these tweeks will come with a matching phase shift, and you'll need to make small adjustments to the delay of the tops again to keep them as perfectly matched as you can. Continue to tweek the delay and EQ etc until it's all sounding good to your ears, and all measuring as good as you can get it through the crossover region. 

With all this done, at each gig this gives you a starting point to work from, where you know things are in phase. If the subs are in front of the tops, or vice versa, you know they've started at a common place so you can just adjust it by that distance to keep them roughly in phase (though always best to measure again at the venue etc while setting up), and if there are any funny room acoustics or nasty reflections that are causing you problems you at least know you started at a common in phase set of settings and can start to take whatever steps are needed to adjust for it.

K

[Ryan C. Davis]

hey...........

K

Thanks Kevin

[ProSoundWeb Community]