Is this how a Subwoofer Crossover is supposed to work?


I bought two Starke SW12 subwoofers that I installed.  So far I'm not particularly happy with them.  They are way too loud even with the volume set almost to off.  More importantly, I'm having trouble integrating them into my system and I'm wondering if that is because their crossover setting is really functioning as I understand a crossover should. Attached please find measurements from Room Equalization Wizard with SPL graphs of the two subs (no speakers) taken at my listening position with the crossover set at 50 Hz, 90 Hz, and 130 Hz. Ignore the peaks and dips which I assume are due to room nodes.  All of those settings appear to actually have the same crossover point of 50 Hz. All that changes is the slope of the rolloff in sound levels. This isn't how I thought a properly designed crossover was supposed to work.  I thought the frequency the levels would start to roll off would change, i.e. flat to 50 hz then a sharp drop, flat to 90 hz then a sharp drop, etc. etc..  But Starke says this is how a subwoofer crossover is supposed to work.

https://www.dropbox.com/s/8x4cr32pagwg48i/Two%20Subs%20Different%20Crossover%20Points%20No%20Speaker...
Any experts on here with an opinion about this?  Is it possible to buy an inexpensive active crossover that I could use in place of what is built into these subs?
pinwa
Again, thanks to those of you that are providing constructive input.

1) I've put the system back together, moving one sub to a different location and setting phase and crossover frequencies individually for each sub.  I am still using the DAC to provide fine tuning of volume with it connected by XLR inputs to the subs and by RCA to the tube amp, but now that I have a better understanding of what is going on the volume on the tube amp is set higher and the volume on the DAC is lower which helps offset the 7 dB gain of the XLR inputs vs RCA.  You can see a chart here. NOTE this was measured before adjusting Amp volume up and DAC volume down so the low end bass response isn't actually exaggerated in the system.

https://www.dropbox.com/s/j5a418zmk99zqbu/Moabs%20with%20Starke%20Subs%20Green%2C%20without%20Orange...

Adding the subs provides significant low end extension and a slightly smoother response.  I have significant room nodes at 61, 77, and 99 Hz with steep 10-15 dB valleys that I'm not sure what to do about.  The subs don't help with that and moving the Moabs around hasn't done much either.  I don't hear them so that is good.  Any suggestions about smoothing those room nodes?  I've tried a couple of rockwool panels without much impact.

2) Active crossovers or DSP all seem like good suggestions but things that I would explore at a later time.  I do use the EQ in Roon to adjust for some things but I haven't figured out how to use REW to measure the changes Roon makes.

3) My assumption has been using the DAC to control volume is fine and that there is no change in signal quality with volume set at 100 vs 60. Anybody have any idea if that is true?

4)  mijostyn's idea of subwoofer placement on either side of the speakers is interesting but, as many people do, I have my tube amp right in the middle.  It is on springs and isolation pads etc. etc. but it seems risky to have a tube amp in between, and close to, subwoofers.  Thoughts?

In conclusion, mostly I've roundtripped to where I started.  The system sounds great, the subwoofers extend the bass and increase low end impact without sounding exaggerated or noticeable.  At some point in the future I may explore active crossovers or DSP but not now.  I would still love to see measurements for how changing the crossover frequency works on a Rel or other high end subwoofer.



Yes! That i show adjustable sub crossovers should work, except for the volume problem.
That is how adjustable sub crossovers should work, except for the volume problem.
hogwash. Why would anybody design an electronic lowpass filter with a variable slope and a fixed crossover point? Thats useless. What you want is a variable crossover point and fixed slope. Even better would be both variable. 
kenjit "Why would anybody design an electronic lowpass filter with a variable slope and a fixed crossover point? Thats useless. What you want is a variable crossover point and fixed slope." 

That is what I thought also and that is what started this whole thread.  But Starke says the subwoofer is designed properly.  And I haven't been able to find any charts that show how other subwoofers respond with changes in the crossover.

Surely some enterprising soul on Audiogon has performed similar measurements on their sub and can post them or knows of a source showing how a "quality" subwoofer works?
You can make anti-modes, or dips controllable with bass traps like GIK Soffit traps, however as Duke would surely advocate, the point of the 2 subs is that the second sub should fill in the big dips, and it doesn't seem like it's happening.  This usually requires non-symmetrical placement.

I suggest you plug your main speaker ports, and re-measure. See if that helps control the peak at ~38 Hz.

Not sure how REW works, but if you are using Roon what I normally do is rip my test signals to my music library, and play the test signals in Roon. This allows me to adjust the Roon PEQ’s accurately.

All of this however is made much harder due to not having a good crossover in place already. This feels like using gum and duct tape and I really think the miniDSP on your sub, plus plugging your mains is your way to glory. Everything else is merely better than before.