SB1000e out-of-band EQ

[Helge A. Bentsen]

Anybody here got some experience with out-of-band EQ on SB1000e?



Current DUT, a Lake LM26(excellent piece of kit btw), means that I've probably made a gazillion different measurements on my current rig(sh50+sb1000e+ux8800). During some experiments with Mesa EQ on the L channel(2x2 sh50/sb1000) I noticed that I could eq the rig to sound pretty good and measure pretty smooth, but a lot of cut was needed around 200hz. Crossover was LR24@100hz. Because I did both array compensation and some taste-correction, the filter became quite wide and complex, but it got me where I wanted(thumbs up to Lab).



Upon further experimentation, I realised that the extra cut around 200hz wasn't necessary when I muted the subs, so I started to dig around with them. I quickly found that the complexity and cut needed in the Mesa filter was reduced by a significant amount if i applied some heavy OOB-eq on the subs. I ended up with a parametric cut on 170hz, bw=0.5, -9 dB on them.



At this point, I decided to redo my system tuning, and ended up with this:



Subs:

LR24 LP@100hz

-9db@170hx, bw=0.5

Delay 3ms



Tops:

LR24 HP@100hz

-3dB lo-shelf@300hz

-2db@200hz, bw=0.2



This sounded pretty smooth to my ears.



Have anybody else here experienced something similar?

 

[Tim McCulloch]

Re: SB1000e out-of-band EQ

How much harmonic distortion comes from the SB1000?



Inquiring minds, etc.



Tim Mc

 

[Art Welter]

Re: SB1000e out-of-band EQ

Out of band eq is used frequently when a speaker has peaks within an octave or so from the pass band.

If using less than 24 dB per octave filters, you may need to correct peaks a couple octaves out to get good conformity of the acoustic crossover between bands.

 

[John Chiara]

Re: SB1000e out-of-band EQ

I had Danely TH 215 subs with an out of band EQ in the DSP and the buss feeding the subs had a 4 band parametric that I always used as an out of band EQ that I could easily adjust by ear as the crowd changes as well as tuning for different musical styles. I think this is the reason so many systems sound ''tubby'' and indistinct. Little or no use of channel highpass filters can have the same effect.

 

[Evan Kirkendall]

Re: SB1000e out-of-band EQ

When I run SB1000's, I normally electronically cross them over around 63-70hz. Then I'll throw an out of band EQ around 100hz. Something like -18, Q15. Works pretty well. The acoustic crossover point still ends up around 90hz or so.







Evan

 

[Jeff Babcock]

Re: SB1000e out-of-band EQ

If you are mixing most modern styles of music with a relatively hot sub level, choosing 100Hz for both your top and your sub crossover points is very unlikely to result in an ideal response.



As Evan has already noted, when setting your sub LPF, forget about electrical response . You want to consider the actual acoustic crossover taking place. In this case, a lower frequency LPF is your friend.



That's not to say that you won't still need some out of band EQ - but let the crossover do its job.

 

[Helge A. Bentsen]

Re: SB1000e out-of-band EQ

As Evan has already noted, when setting your sub LPF, forget about electrical response . You want to consider the actual acoustic crossover taking place. In this case, a lower frequency LPF is your friend.

I've tried that in the past, and can't say I liked it. For some reason I prefer the sound of an symmetrical electric crossover. Right now the acoustic crossover point is somewhere around 120hz, if I lower the crossover point on the subs, they will sum flatter in the 100-120hz region, but I don't like it sound wise. It's probably because i prefer that extra addition. And if I move to a symmetrical electric crossover point around 80hz, the sb1000e sounds terrible IMHO

Some more experimentation is in order.

 

[Jeff Babcock]

Re: SB1000e out-of-band EQ

As Evan has already noted, when setting your sub LPF, forget about electrical response . You want to consider the actual acoustic crossover taking place. In this case, a lower frequency LPF is your friend.

I've tried that in the past, and can't say I liked it. For some reason I prefer the sound of an symmetrical electric crossover. Right now the acoustic crossover point is somewhere around 120hz, if I lower the crossover point on the subs, they will sum flatter in the 100-120hz region, but I don't like it sound wise. It's probably because i prefer that extra addition. And if I move to a symmetrical electric crossover point around 80hz, the sb1000e sounds terrible IMHO

Some more experimentation is in order.

You might prefer the way your symmetrical crossover point sounds, but this simply means that you like more addition/overlap, not that symmetrical is the ''right'' way to approach this. Based on this I would presume you might like a rig with dedicated midbass boxes to give you extra meat in that range.



If you are happy with the results, then don't change for the sake of change. But I don't want folks to get the wrong idea from what you are saying. For those reading...... acoustic crossover is king, not the electrical response. Non-symmetrical electrical response is usually what is required to achieve a good looking acoustic response when dealing with hot sub levels.

 

[RYAN LOUDMUSIC JENKINS]

Re: SB1000e out-of-band EQ

Non-symmetrical electrical response is usually what is required to achieve a good looking acoustic response when dealing with hot sub levels.

It may be that the OP is not dealing with really hyped subs. In the last few years I have been hyping my subs a lot less than I used to. I have found that a symetrical crossover has worked for my applications lately.



All this being said, I provided some additional subs for an EDM event at a venue that has a fairly nice house rig. We ended up with an electrical crossover point considerably lower than what I would use under my normal circumstances, further confirming your statement above.

 

[Bennett Prescott]

Re: SB1000e out-of-band EQ

As Evan has already noted, when setting your sub LPF, forget about electrical response. You want to consider the actual acoustic crossover taking place. In this case, a lower frequency LPF is your friend.

I've tried that in the past, and can't say I liked it. For some reason I prefer the sound of an symmetrical electric crossover. Right now the acoustic crossover point is somewhere around 120hz, if I lower the crossover point on the subs, they will sum flatter in the 100-120hz region, but I don't like it sound wise. It's probably because i prefer that extra addition. And if I move to a symmetrical electric crossover point around 80hz, the sb1000e sounds terrible IMHO

Helge,



The reason your subs need a big cut at 200 is because of your incorrect crossover settings.



Out of band EQ can be a handy tool (for reasons Art has elucidated), but what is really happening is that you ''like'' the overlap around 100Hz, but by the time it gets to 200Hz you've stopped liking the extra contribution from the subs and want their response to be gone. That's not to say you don't also need some out of band EQ, although that has not been my experience when aligning EAW systems run on the UX8800, but the real problem is your crossover selection.



Here is the problem: Your tops roll off somewhere in the 80Hz range (let's just say). Maybe they die a little higher or a little lower, but the point is that regardless they are rolling off near where you would set their high pass filter for system integration anyway. That means that when you apply your high pass it combines with their existing acoustic rolloff and produces additional rolloff and phase shift. With a 4th order filter, you may be seeing an 8th order acoustic response.



Your subwoofers on the other hand, if not effectively low passed, will go to 1kHz or so before they start to roll off. This is octaves away from where you would set your low pass. That means that, if they are equalized flat or are flat of their own accord, when you apply your 4th order L-R filter you are getting almost an ''electrical'' response. If you set it at 100Hz then your filter is down 6dB at 100 and falls 24dB/octave beyond that, with the expected phase response to match.



So while your tops are gone just below your crossover point, your subs are rolling off at half the rate and with half the phase shift. This means that while you can match phase at crossover, it won't stay matched for long. Furthermore, when you boost your sub level 10dB or so your sub response extends for half an octave or more before it equals your top response, and still contributes for another half an octave beyond that... this is where your 200Hz overlap comes from.



An ugly solution is to use an 8th order L-R filter for your sub low pass, that will at least make the slopes equal and make your phase alignment a lot easier. I don't like using high order filters, though, so I avoid this.



The best solution is to drop your sub low pass to 60-70Hz. This will account for the crossover overlap, and with a little care you can usually get the phase alignment to play nice. Sometimes I'll increase the order of the low pass filter just to make the phase slopes line up, but once you're playing with 4th order filters already you start to get into nasty sounding filters when you go much beyond that (or it could be argued that it sounds nasty already, and you would ideally use a lower order filter for your top HPF...)



I suspect the reason you don't like doing this is because of time. Your tops are incurring the phase shift of a 4th order filter at 100Hz. When you have the sub filter there as well it incurs the same phase shift, so they remain time aligned if they were to begin with. Now that you've shifted the LPF for your subwoofers almost an octave lower, however, they incur almost twice the phase shift, or almost twice the delay. What you need to do is add delay to your tops to make their delay at the crossover point equivalent to the subs... 2-4ms is what I find to be normal, probably on the high side of that for such steep filters.



Hopefully that is useful, I've had too much coffee and so I'm a little stream-of-consciousness.

 

[Art Welter]

Re: SB1000e out-of-band EQ

As Evan has already noted, when setting your sub LPF, forget about electrical response . You want to consider the actual acoustic crossover taking place. In this case, a lower frequency LPF is your friend.

I've tried that in the past, and can't say I liked it. For some reason I prefer the sound of an symmetrical electric crossover. Right now the acoustic crossover point is somewhere around 120hz, if I lower the crossover point on the subs, they will sum flatter in the 100-120hz region, but I don't like it sound wise. It's probably because i prefer that extra addition. And if I move to a symmetrical electric crossover point around 80hz, the sb1000e sounds terrible IMHO

Some more experimentation is in order.

Easy to see why a ''regular'' crossover does not work with an SB 1000 (and many other) subs.

The difference between 50 Hz and 200 Hz is over 10 dB, a 12 dB per octave filter would need to be applied just to flatten the response before considering a crossover point.

 

[Bennett Prescott]

Re: SB1000e out-of-band EQ

Easy to see why a ''regular'' crossover does not work with an SB 1000 (and many other) subs.

The difference between 50 Hz and 200 Hz is over 10 dB, a 12 dB per octave filter would need to be applied just to flatten the response before considering a crossover point.

Art,



Helge said that the subs were processed with a UX8800 which, in my experience, provides excellent equalization of the SB1000 (and any other system I've seen it connected to).

 

[Art Welter]

Re: SB1000e out-of-band EQ

Easy to see why a ''regular'' crossover does not work with an SB 1000 (and many other) subs.

The difference between 50 Hz and 200 Hz is over 10 dB, a 12 dB per octave filter would need to be applied just to flatten the response before considering a crossover point.

Art,



Helge said that the subs were processed with a UX8800 which, in my experience, provides excellent equalization of the SB1000 (and any other system I've seen it connected to).

7.1.2 GENERIC PROCESSING

When generic processing is selected for an Input Channel, the Input Channel and Output Channel

processing is configured for standard digital processing functions, including signal delay, gain, polarity,

EQ, limiting, and high/low-pass filters.



This allows users to fuck up settings in any way desired

.

 

[Bennett Prescott]

Re: SB1000e out-of-band EQ

Art,



I have been ASSuming from the start that Helge has the UX8800 configured somewhat correctly...

 

[Pascal.Pincosy]

Re: SB1000e out-of-band EQ

Hi Helge,



Can you post some of the measurements you took? Both full-range and post-processing would be great!

 

[Helge A. Bentsen]

Re: SB1000e out-of-band EQ

A lot of interesting things in this tread

Based on this I would presume you might like a rig with dedicated midbass boxes to give you extra meat in that range.

Valid point. My favorite RnR-pa is Meyer 650+Ds2+Msl3A. I've also had some good times with 700hp+Psw4+Msl4

Non-symmetrical electrical response is usually what is required to achieve a good looking acoustic response when dealing with hot sub levels.

It may be that the OP is not dealing with really hyped subs.

Again, valid point. Bought myself a decent home stereo a couple of years ago, have since started to turn down the subs in all the systems I mix on. I prefer adding low end on the channels that needs it, rather than using brick wall lo-cut on vocals etc.

Art,



I have been ASSuming from the start that Helge has the UX8800 configured somewhat correctly...

I've loaded the correct greybox for an SB1000, but I don't do any lp-filtering in the UX8800. I do all the filtering in the LM26, and feed te UX by AES. The LP-filter is set at 20k or something in the UX. I think this is the 17th UX-processor I've installed now

 

[Helge A. Bentsen]

Re: SB1000e out-of-band EQ

As Evan has already noted, when setting your sub LPF, forget about electrical response. You want to consider the actual acoustic crossover taking place. In this case, a lower frequency LPF is your friend.

I've tried that in the past, and can't say I liked it. For some reason I prefer the sound of an symmetrical electric crossover. Right now the acoustic crossover point is somewhere around 120hz, if I lower the crossover point on the subs, they will sum flatter in the 100-120hz region, but I don't like it sound wise. It's probably because i prefer that extra addition. And if I move to a symmetrical electric crossover point around 80hz, the sb1000e sounds terrible IMHO

Helge,



The reason your subs need a big cut at 200 is because of your incorrect crossover settings.



Out of band EQ can be a handy tool (for reasons Art has elucidated), but what is really happening is that you ''like'' the overlap around 100Hz, but by the time it gets to 200Hz you've stopped liking the extra contribution from the subs and want their response to be gone. That's not to say you don't also need some out of band EQ, although that has not been my experience when aligning EAW systems run on the UX8800, but the real problem is your crossover selection.



Here is the problem: Your tops roll off somewhere in the 80Hz range (let's just say). Maybe they die a little higher or a little lower, but the point is that regardless they are rolling off near where you would set their high pass filter for system integration anyway. That means that when you apply your high pass it combines with their existing acoustic rolloff and produces additional rolloff and phase shift. With a 4th order filter, you may be seeing an 8th order acoustic response.



Your subwoofers on the other hand, if not effectively low passed, will go to 1kHz or so before they start to roll off. This is octaves away from where you would set your low pass. That means that, if they are equalized flat or are flat of their own accord, when you apply your 4th order L-R filter you are getting almost an ''electrical'' response. If you set it at 100Hz then your filter is down 6dB at 100 and falls 24dB/octave beyond that, with the expected phase response to match.



So while your tops are gone just below your crossover point, your subs are rolling off at half the rate and with half the phase shift. This means that while you can match phase at crossover, it won't stay matched for long. Furthermore, when you boost your sub level 10dB or so your sub response extends for half an octave or more before it equals your top response, and still contributes for another half an octave beyond that... this is where your 200Hz overlap comes from.



An ugly solution is to use an 8th order L-R filter for your sub low pass, that will at least make the slopes equal and make your phase alignment a lot easier. I don't like using high order filters, though, so I avoid this.



The best solution is to drop your sub low pass to 60-70Hz. This will account for the crossover overlap, and with a little care you can usually get the phase alignment to play nice. Sometimes I'll increase the order of the low pass filter just to make the phase slopes line up, but once you're playing with 4th order filters already you start to get into nasty sounding filters when you go much beyond that (or it could be argued that it sounds nasty already, and you would ideally use a lower order filter for your top HPF...)



I suspect the reason you don't like doing this is because of time. Your tops are incurring the phase shift of a 4th order filter at 100Hz. When you have the sub filter there as well it incurs the same phase shift, so they remain time aligned if they were to begin with. Now that you've shifted the LPF for your subwoofers almost an octave lower, however, they incur almost twice the phase shift, or almost twice the delay. What you need to do is add delay to your tops to make their delay at the crossover point equivalent to the subs... 2-4ms is what I find to be normal, probably on the high side of that for such steep filters.



Hopefully that is useful, I've had too much coffee and so I'm a little stream-of-consciousness.

Very useful indeed. I rechecked my delay times after applying the oob-eq, so they should be somewhat correct.



I think I will:



- Go back and recheck my delay times, and especially pay attention to the acoustic crossover point.

- Try to configure an asymmetrical acoustic crossover that I like, by lowering the crossover on the subs.

 

[Helge A. Bentsen]

Re: SB1000e out-of-band EQ

Hi Helge,



Can you post some of the measurements you took? Both full-range and post-processing would be great!

Unfortunately, I only stored two sub measurements. I'll have to wait until Sunday to capture some new ones. Neighbours and pink noise isn't fund of each other....

 

[Bennett Prescott]

Re: SB1000e out-of-band EQ

Helge,



We'll need to see a capture of your mains as well, and your mains and subs together, to really determine what's going on... make sure to use the same delay compensation and mic placement for all three measurements.

 

[Helge A. Bentsen]

Re: SB1000e out-of-band EQ

Helge,



We'll need to see a capture of your mains as well, and your mains and subs together, to really determine what's going on... make sure to use the same delay compensation and mic placement for all three measurements.

Will do