Electrical Crossover and Acoustical Response

[JCsAudio]

If you measure the speakers together and then flip the phase on one of them then you will get cancellation were they merge and it will show up as a dip in the response at the crossover point when measured with a microphone, which means they are out of phase. If the response has the opposite effect then they are in phase. Electrically two speakers with 12 db slopes should be 180 degrees out of phase at the crossover point but if the net result is 24 db for each driver (just guessing here now so Nick correct me if I’m wrong) then they would be back in phase at the crossover point. For each 6 db increase in electrical crossover your phase changes 90 degrees until it gets back to 360 at 24 db. It’s been a while but I think I have that correct.

[Justin Zazzi]

I used LR/24 in Jazzi's tuning companion for REW when I ....

I'm not sure if it's clear, but I am Jazzi and I post here now. I'm glad you enjoy the tool!

Do we swap polarity/invert according to the electronic crossover or the acoustic crossover?

Acoustic is the only response that matters. All rules of thumb apply to the acoustic response. Ignore the electrical response, it is just a means to an end.

If you measure the speakers together and then flip the phase on one of them then you will get cancellation were they merge and it will show up as a dip in the response at the crossover point when measured with a microphone, which means they are out of phase. If the response has the opposite effect then they are in phase. Electrically two speakers with 12 db slopes should be 180 degrees out of phase at the crossover point but if the net result is 24 db for each driver (just guessing here now so Nick correct me if I’m wrong) then they would be back in phase at the crossover point. For each 6 db increase in electrical crossover your phase changes 90 degrees until it gets back to 360 at 24 db. It’s been a while but I think I have that correct.

That sounds right.

[jrwalte]

I'm not sure if it's clear, but I am Jazzi and I post here now. I'm glad you enjoy the tool!

I use your tool for every tune as well! Great tool. Thanks for making it.

And I've just learned from this thread that you need to make the acoustical response match to a 24db cross over slope to achieve an in-phase cross over. This should help a lot in my next tune.

[Brego]

When I used LR/24 as my electrical XO's, ((and LR/24 in Jazzi's excel)) the RTA response always measured about a 33db down-slope on my High-Pass side, and about a 30db down-slop on my Low-pass side. This made me boost the EQ Freq's in those areas to match up to the Speaker curve, with big cuts in the center freq's, and by doing that, I always had a harsh, thin sounding system. The only way I could offset the horrible down-slopes was to ‘Overlap’ my XO’s.. and that NEVER seemed to work very well.
When I used electrical LR/12 XO’s, the RTA response matched to a natural LR/24 roll-off on the HP & LP sides a lot better. I did have to ‘underlap’ my XO’s quite a bit to get it all to line up,(before Eq’ing) but after some Eq’ing, the results of the sound of my system is so dramatic, it’s like a whole new system was installed. ALL the harshness is gone, the tonality of the system is so much deeper, dynamic, crisp, tight stage, with sub&base up front….. hell… even XM radio sounds pretty good….I finally have a base tune good enough that I can spend some time to ‘fine’ tune it…..

[Justin Zazzi]

When I used LR/24 as my electrical XO's, ((and LR/24 in Jazzi's excel)) the RTA response always measured about a 33db down-slope on my High-Pass side, and about a 30db down-slop on my Low-pass side. This made me boost the EQ Freq's in those areas to match up to the Speaker curve, with big cuts in the center freq's, and by doing that, I always had a harsh, thin sounding system. The only way I could offset the horrible down-slopes was to ‘Overlap’ my XO’s.. and that NEVER seemed to work very well.
When I used electrical LR/12 XO’s, the RTA response matched to a natural LR/24 roll-off on the HP & LP sides a lot better. I did have to ‘underlap’ my XO’s quite a bit to get it all to line up,(before Eq’ing) but after some Eq’ing, the results of the sound of my system is so dramatic, it’s like a whole new system was installed. ALL the harshness is gone, the tonality of the system is so much deeper, dynamic, crisp, tight stage, with sub&base up front….. hell… even XM radio sounds pretty good….I finally have a base tune good enough that I can spend some time to ‘fine’ tune it…..

You saw the light my friend! Very cool!

This is a great example of if you get the acoustic response to match the targets then it doesn't matter if you "underlap" the electrical filters or whatever. Asymmetric and non-conventional filters are not the enemy! I need to make tshirts or something.

[Brego]

That light is still dim... but as I experiment more, the light is getting brighter....


Hey, I'd buy your T-shirt... but, I gotta google what the hell "Asymmetric" means first.....


Thank You for creating that brilliant piece of work “Jazzi's tuning companion for REW”.. I never would have gotten this far without it !

[Brego]

I going to try Crossover Butterworth/18 next,,,, to determine if I can reduce the 'underlap' a bit.....

[Hillbilly SQ]

So am I reading this right? The electrical crossover doesn't matter as long as the acoustical crossover is where it needs to be? As in you don't need the electrical crossover and acoustical crossovers to match as long as the acoustical crossover is good?

I remember back in 2008 Nick Wingate (everyone should know who he is) got in my truck and almost threw up when he heard about 5 seconds of it. He told me my acoustical phasing was off really bad and started going to town with the crossovers in my Alpine 9833. That was a very popular way to go active back when a standalone processor was still unobtanium for most budgets. He got acoustical phasing lined out best he could and bailed out. He also had me make adjustments with my PG tld66 linedriver that I had in there for independent level control that couldn't be done with the headunit or amps. Then Mark Eldridge and David Seal did some eq work. That truck sounded AWESOME for what it was after those three were done with it. Our very own Jorge (doiter) was there too. I don't think Nick Wingate was nearly as concerned about the electrical crossover as he was the acoustical crossover. I keep going back to what he was telling me when I think about acoustical slope vs electrical slope and phase.

[Jdunk54nl]

So am I reading this right? The electrical crossover doesn't matter as long as the acoustical crossover is where it needs to be? As in you don't need the electrical crossover and acoustical crossovers to match as long as the acoustical crossover is good?

I remember back in 2008 Nick Wingate (everyone should know who he is) got in my truck and almost threw up when he heard about 5 seconds of it. He told me my acoustical phasing was off really bad and started going to town with the crossovers in my Alpine 9833. That was a very popular way to go active back when a standalone processor was still unobtanium for most budgets. He got acoustical phasing lined out best he could and bailed out. He also had me make adjustments with my PG tld66 linedriver that I had in there for independent level control that couldn't be done with the headunit or amps. Then Mark Eldridge and David Seal did some eq work. That truck sounded AWESOME for what it was after those three were done with it. Our very own Jorge (doiter) was there too. I don't think Nick Wingate was nearly as concerned about the electrical crossover as he was the acoustical crossover. I keep going back to what he was telling me when I think about acoustical slope vs electrical slope and phase.

Justin, Chris, and I just had a good chat about this late one evening.

Yes, you are reading it right. Nothing electrical ultimately matters for the acoustical response at the listening position. That means if your speaker naturally rolls off at 12db/oct at 80hz, you may only need to apply a 12db/oct filter to get it to match the 24 db/oct acoustical slope. Or you might need to apply a 6 or 18 or even 24 to match your curve.

Now it might make a difference if you need to lower the crossover beyond the speakers limitations to match what you ideally would like....which means you need to raise your crossover point to fit your speaker and not make your speaker fit your curve.

For example. I was attempting to get my wifes suv to cross at 2000hz in her car. The tweeters couldn't do it unless I dropped the crossover to around 1700-1800hz. I wasn't comfortable dropping that far due to limitations of the tweeter and I thought I would then be pushing it too hard and could cause it to blow. So I changed the crossover point to around 2500hz based on the tweeters frequency response.
You just need to remember if you were to measure the frequency response close to the speaker and you lower your crossover, the speaker will be playing those lower frequencies, but at your listening position something has interfered with them.

[Hillbilly SQ]

What about for example on a pair of midranges where they act different around the intended cross points? Can one be at 24db for example and one at 18 or 12? What about mixing Linkwitz Riley and Butterworth slopes? I'm personally not a fan of odd order electrical slopes but probably more of a superstition than anything.