Subwoofer crossover with full range mains

The first gig with the new subs was a private party at a pub and we crossed the subs at 80Hz to keep the guitars out of the aux-fed subs.:guitar: The new subs have a more solid bottom,:spank: but I suspect that is due primarily to the new sub's internal DSP EQ flattening the response compared to the old subs with no DSP and a strong bump at 80Hz. If we had measurement capability we might have tuned the old subs well also, given time. Even with the new subs that have flatter bass, as usual we had to put lots of EQ on the system to get smooth bass in a pub anyway. I chalked it up to learning curve and moved on.
 
The second gig was outdoors and the sound was far more transparent, warts and all. We crossed the subs at 100Hz at the drummer's suggestion despite the attenuation in the guitars:guitar: on the E string from not being in the subs, but I barely noticed the attenuation of the E string while I was concentrating on other aspects.:listeningmusic:

In the absence of room acoustics I noticed a strong bump in bass output from the DJ's full range house music at 100Hz that I suspect might be due to the 24dB/octave filters in the DL1608 crossover we created for the aux-fed sub. For flat response through crossover they need to be Linkwitz-Riley cascaded stages with 6dB of attenuation in their cascaded response at 100Hz crossover so that they sum to 0dB with mutual coupling between the subs/mains woofers at crossover. When I look at the plot in Master Fader, at 100Hz the lowpass/highpass crossover filter on the sub/mains is down 3dB instead of 6dB, so there would be a 3dB bump in the combined output at crossover from the improper crossover implementation.:facepalm:

Anyone know anything about this, how to manage it, without using external crossover? We just tailored the EQ on subs and mains to fix it. I should have come up with a strategy before the gig, but oh well...

I also neglected to account for the internal crossover at 120Hz in the sub. That internal crossover cannot be defeated despite us crossing the aux feed at 100Hz, but if anything that extra crossover probably helped attenuate the 3dB bump at 100Hz from the lack of cascaded L-R crossover in the board. I added no additional 100Hz attenuation in the subwoofer.

I noticed that even with a bass boost contour on the subs amounting to 6dB of boost at 20Hz with the bass boost curve beginning at about 50Hz (about 6dB per octave bass boost but with a hook up at 20Hz rather than straight line), the low end was still a little weak as would be expected from a ported sub with 12dB/octave rolloff starting at about 50Hz but only 6dB/octave of bass boost.

I did not add any additional bass boost since I figured it was probably a bad idea to go chasing LFE outdoors with ported 1000W subs that plug into 120V 20A circuit. Maybe later, once we have time to experiment more, I can be more aggressive in the LFE for outdoor gigs, but I wanted to run it by the pros here also to see if they have an opinion on that.

What I did try was straightening the bass boost from a hooked curve more toward a straight line ramp, but that just made the bass muddy, so I reversed the change.

The gig was plenty loud without being over the top and overall it sounded good, but if I go adding another 6dB at 20Hz I will be multiplying the LFE power by a factor of somewhere between 10 and 20 over what it would be with flat EQ,:eek: and that seems like a really bad idea for a novice sound engineer to attempt during a first-time outdoor gig running brand-new killer subs on a borrowed power circuit of unspecified pedigree.;)

I also noticed that the 100Hz bump was still somewhat audible after these adjustments, though I did not directly attempt to compensate it beyond 2dB of cut in the mains, as part of the feedback notch we needed (the portable elevated stage was tiny and the mains were close to the mics). I suppose I could have put slightly more cut at 100Hz, but by the time I got this far I figured it was better to leave it alone, especially since 100Hz also seemed to vanish in the null beside the power alley along with the rest of the LFE.:(

I mixed from approximately 20 degrees off the centerline and the subs were approximately 15'-20' apart center-center, so I was about midway between the center lobe and the first LFE null of the power alley per my listening test. The mix was noticeably light on the LFE when listening from the left side past my seat at about 45 degrees off centerline, and it sounded a little boxy/muddy from directly on the centerline, and where I was mixing at near 20 degrees off centerline it sounded fairly flat and natural but there was still a touch of airy nullishness to the LFE, so I figured I probably got the compromise about right. The lobing was highly audible as I walked in an arc across the front, and this was my first opportunity to actually audit the power alley effect in person.

When I was done with the EQ it seemed to me that I pretty much hit the same flattish but slightly boxy sound that I hear at all outdoor gigs, only maybe slightly less so because this equipment seems to perform really well:thumbs up: and I tried to optimize it using my existing background in engineering. From what I have experienced so far, most people mixing sound in the wild (at bars and free outdoor festivals) tell me they have no expertise:o and learned by trial and error on their own from others who did the same.;) I guess I am sort of pleased/proud to have the opportunity to play with these guys,:thumbs up: because I am learning a lot doing this and having a great time using my background knowledge in the process.

I am not ready with the PC and REW so I did no measurements. I would not have had time anyway. Maybe at the next outdoor gig. I will need an analog mic to do the time alignment per your helpful feedback on this forum. That will have to wait until I am ready for it. Other priorities currently.

The DJ who spun the house music had a dual turntable with those videodisc/laserdisc-ish stylus that only serves as a guide groove and the audio on a pitted buried metal layer under the vinyl. He was scratching on one track while playing another track on the other turntable, and would swap back and forth between the two turntables like that, playing one and scratching on the other.

It was the first time I had personally witnessed either a DJ scratching, or those newfangled videodisc/laserdisc-ish audio tech. When the drummer informed me that the DJ was scratching on vinyl I hurriedly created a rumble filter but I took it out while debugging that 100Hz bump when I asked the DJ about his equipment, specifically his EQ, and learned about the disc tech he was using. I figured that the pitted substrate was not going to be susceptible to rumble amplification since it probably uses pulse-width modulation or some form of digital encoding rather than RIAA equalization, and might not even be sensitive to rumble at all. But that was just a guess. No woofers went flying across the parking lot.:laughings:
 
The prior reply (#42) got held for moderation. If this one appears first, the thread could be somewhat confusing.

The drummer wants to upgrade the mains to the K12.2 on pole mounts for better projection over the audience, and spare his back from hauling those extremely heavy Mackie 3-way mains around, but I advised him to either choose the KW122 from the same product line as the subwoofer, or contact QSC to verify that the K12.2 will integrate well with the KW181. The DSP is probably tuned for delay and we hope that QSC has been consistent across product lines but just in case...

It seems that the K.2 line has an improved tweeter for better dispersion characteristics up to a higher frequency, so it may be preferable overall as long as it plays nice with the KW181.
q_spk_k_12.2_ease-horiz.png
q_spk_kw_122_ease.r2v11-horiz.png

Once the mains and subs match, I should be able to ignore time alignment, but I still need to figure out that 3dB bump at crossover with aux fed sub and a crossover fudged in Master Fader with the DL1608.

If he does buy matching mains, we could use the internal crossovers in the speakers instead of fudging one in the mixer. The QSC powered speakers have the mains crossovers built into the mains instead of the subs, so that is an improvement over the Mackie subs/mains where the whole crossover is in the subwoofer and incompatible with aux fed sub.

The QSC crossover is set to 120Hz apparently, so that means either the guitars are going to need to use the aux fed sub, or the low E string is going to lose a quarter of its bandwidth.

I checked the EASE data on the K12.2 and the EASE data on the KW122, but for some reason the treble looks fishy bad (way too much attenuation above 5KHz).:(:wtf:
q_spk_k_12.2_ease.png
q_spk_kw_122_ease.r2v11.png

I am unfamiliar with EASE data/viewer so maybe I am just misinterpreting,:confused: but it looks to me as if the QSC frequency response data is weird. There is an 'input voltage' curve and it is not even close to flat.
q_spk_k_12.2_ease-input.png
q_spk_kw_122_ease.r2v11-input.png

So if I am supposed to ignore the input voltage curve and take the frequency response data at face value, the K12.2 and the KW122 seem to have terrible treble response,:eek: but the bass of the KW122 is down only 10dB at 20Hz and that also seems like improbably good, strong output that low.:confused: Did they take the measurement with DEEP bass boost enabled?

Even worse, the K.2 and KW lines seem to have different methodology where the 'transfer function' is concerned, since they look radically different and confusing:
q_spk_k_12.2_ease-xfer.png
q_spk_kw_122_ease.r2v11-xfer.png

I have no idea what I am looking at.

The sensitivity plots are also confusing:
q_spk_k_12.2_ease-sens.png
q_spk_kw_122_ease.r2v11-sens.png

I emailed QSC and they were unhelpful with figuring out why I am having difficulty understanding the frequency response data they published. What you see is what you get.

The K12.2 claims 1dB more peak output, but has a smaller compression driver on the tweeter (1.4 vs 1.75) and uses slightly less input current at 1/8 power, but otherwise appears equivalent to the KW122.

However, the KW series claims that the power is split equally between the woofer and tweeter at 500W each and that also seems improbable even though it has a somewhat larger compression driver on the tweeter. The K.2 line indicates that the 'peak' power is divided up between woofer and tweeter at 1800W and 225W respectively and that seems more reasonable.

Any thoughts? I suspect that this spec on the KW series is just plain wrong. I never heard of a 500W tweeter in any universe. Maybe that applies to the KW153 where the tweeter and midrange share the same amp, but it cannot be right for the 2-way speaker, unless they are also throttling the amp back to avoid frying the tweeter crisp like bacon.

I think this improbable power allocation is just a mistake in the documentation, and I think the frequency response plots are also wrong, but I would really like to know so I can advise the drummer properly.

I asked the drummer to also consider the KW152 or KW153 as potentially a better mate for the KW181, but I suspect that is too heavy/expensive and the KW152 only has 60 degree horizontal dispersion...

I could use some advice here. It is not like I am going to be able to figure this out with a listening test. On a good day I have trouble discerning quiet speech from across a conference room table. I have brick-wall loss at 10KHz.

It would be a shame for my friend to buy speakers that sound as bad as my hearing is. However, I have already mixed for a band that uses the K122 without a sub and the speaker sounds fabulous to my damaged hearing.

What do you think?

Sorry for the long update. I could probably have dribbled it out a little at a time in back-and-forth banter, but this was easier for me to compose it all in one sitting.

Thank you all for your helpful input. This has been a fascinating plunge into the night life. I look forward to your excellent suggestions!:cursing::RTFM:;)
 
You have to eq for two different reasons, to compensate for the speaker and to compensate for the room. Even if the speaker was perfectly flat, you would still have to compensate for the room.
 
Did it sound good? If so, then it is good. It's never going to be perfect so you need to accept that and get on with the job of making it fun for the audience and satisfying for the band. On the one hand, I admire and envy your attention to detail, but it's possible to be too obsessive. Focus on the show, mix the band.
 
I agree with BSG.

One thing did catch my eye, you were adding a bass boost below 50hz? I would almost guess the reason why you seemed bass-tight after doing that is because you were turning up useless rumble that was using all the headroom in the amp. I tend to filter out anything below 35hz, just to clean up the low end and save the power for the frequencies that matter.

Another thing to realize is that stringed instruments, like guitars, have most of their power an octave above the note they are playing. This means the low E on a guitar will have more power at 160hz than at 80hz. Same with bass.

It really does sound like you are way overthinking this whole thing. 80hz is a good standard sub crossover frequency and the audience will not notice that the guitars aren't coming out of the subs. Remember, the point is to entertain the audience, not create a mathematically perfect sound system.
 
Yeah, you got me pegged.;)

I am used to tuning circuitry.

I think it is OK to just run with the setup as-is, and concentrate on learning to EQ various instruments. There should be some notes here on the forum on that.

I could use some background info on setting up gates and compression also.

So I will do some reading. I think the mains integration is handled OK now.
 
Here is some video of my first live outdoor gig mixing sound with East Crescent.

The Mackie mains are crossed to the aux fed subs at 100Hz with a 'faked' non-Linkwitz-Riley crossover in the DL1608.

IMG_0285.PNG
IMG_0284.PNG

I added some LFE boost to the subs and added some minor graphic cut near 100Hz to compensate for non-L-R crossover added into the feedback suppression notches in the mains. I also had a -1dB/octave overall 'house curve' contour to tame the high end a little. This curve is stored as a preset and I recalled it for both mains and subs before tweaking the graphic, so that explains why the subwoofers have full band graphic contour on them.

IMG_0287.PNG
IMG_0286.PNG

I did not have instrumentation to set the delay so there may be significant uncompensated skew between the analog mains and the DSP subs, but I did not hear any ill effects from it. The major issue I noticed was the power alley and you can hear the effect it had on the bass from the location at the mixer/camcorder. The second major issue I noticed is that the subs lost bottom from being up on the stage, plus there was probably some boundary interference from the ground as well as from the building behind the camcorder.

Overall I was pleasantly surprised by the sound quality, and a little disappointed that the lead singer was having uncharacteristic difficulty hitting the higher notes that day. I took advantage of the outdoor setting to spend significant time tweaking the proximity effect in the vocal subgroup, but on the recording it still sounds to me as if I could have added another 2dB of cut in the low/mid end to fix the slight 'cupped hands' megaphone sound. Overall there was approximately 7.5dB of bass cut sloping in a drooping straightish contour from 30Hz up to about 1KHz.

Band 1 low shelf: -2.9dB at 271Hz
Band 2 peak: -2.1dB at 37.7Hz, Q=0.5
Band 3 peak: -3dB at 20KHz, Q=1.9
Band 4 high shelf: +2.5dB at 895Hz

IMG_0283.PNG

I always have difficulty setting the level balance between the keyboard and guitars for this band, plus I had to re-do the guitar EQ for the outdoor setting.

Original (indoor) lead guitar EQ:

IMG_0289.PNG

Revised (outdoor) lead guitar EQ:

IMG_0288.PNG

The rhythm guitar needed a similar adjustment for the outdoor venue.

I set the compression to be just barely triggering on the mains and the bass/kick, and left it pretty much disabled on everything else except the rhythm guitar. I did not really have time yet to experiment with the compression so it is primarily a safety factor to protect woofers. The drummer set up the rough cut on the compression and noise gates before I arrived and I made few changes. I noticed no major issues with the gating or compression but I would have preferred a more sophisticated approach.

I welcome your comments and suggestions. Driving a capable P.A. is somewhat intimidating for a noob.

EDIT: I credited both East Crescent plus Drunken Starfighter (karaoke) too and replaced the link:

East Crescent with Drunken Starfighter at Player's Ink 2017-06-11 (part 1) - YouTube

EDIT 2: The playlist is not showing up when I use this link I posted. You can find the playlist with both videos on my YouTube profile here:

East Crescent and Drunken Starfighter at Player's Ink 2017-06-11 - YouTube
 
Last edited:
Back
Top