Using the EQ function: proper approach for flattening mains?

[cvanc]

I have fullrange mains that I cross over at 80Hz to subs. I also use the EQ function in REW to flatten the mains.

Here's the question: In such an application should you let the EQ function process a fullrange sweep or a sweep that includes the 80Hz crossover? Many thanks.

 

[ddude003]

That would depend on your use case... If mostly stereo music I would suggest crossing over at about 10Hz above where main speakers roll off at the low end... If mostly for surround sound AV movies etc then your crossover at around 80Hz is pretty standard... You might try +/- 15Hz around 80Hz and use what sounds best to your ears...

If you tell us more about your kit and use case we could probably make some better suggestions...

 

[JStewart]

Here's the question: In such an application should you let the EQ function process a fullrange sweep or a sweep that includes the 80Hz crossover? Many thanks.

I shouldn’t matter from a sound quality perspective. More important is not using EQ to extend a speaker’s range beyond its limits (which is unlikely to apply here if a good XO point has been chosen) or trying to over boost a non minimum phase dip.
If EQ is to be done with the XO in place the XO type and slope will need to be known and incorporated into the target for both subs and mains. A benefit to not baking the XO into the EQ filters is that different XO points can be tried/compared using the same filters.

Why Can't I Fix All my Acoustic Problems with EQ?

Minimum Phase

(Below is for Dirac Live, but the principles still apply.)

Setting crossovers in an AVR/pre-pro with Dirac Live 2 room correction

 

[cvanc]

Appreciate your input, thanks. I'm trying to figure out "rules of thumb" here - the processes and procedures that are independent of the hardware you own. That's why I asked it the way I did.

But to answer the questions, my mains have a low frequency power bandwidth (-3dB point when crazy loud) of 62Hz. They are electrostatic so I want to roll them off rapidly (4th order). An 80Hz XO actually is about right, it's fairly conservative, and it helps reduce THD when playing loud.

This gets me to my next question: Please see the parameters circled in red in the following photos. Should these all be set to match the XO the mains are using?

Again, many thanks.

 

[JStewart]

In the 1st screenshot the control above should be circled ones should also be changed.
In the 2nd one, for the mains I would go EQ to at least an octave below the XO, in this case 40Hz, so long as they seem capable. If not I would use a higher XO.
Not all XOs are the same, Butterworth and Linkwitz, being the most common. Do you now what your kit is using and their slopes (order)?

Edit: you mentioned the speakers are capable to 63. I would go higher with the XO and I would not EQ with XOs in the target, because you’ll get the same results in the end and can try different XOs with the same EQ without them baked in to the EQ. Perhaps that’s what you intended.

 

[ddude003]

Still not saying much about your kit... I have ML ESLs and an SVS sub, so I get that there are some interesting issues with crossovers... However, without specifics about your ESLs and your Subwoofer and if you are using some sort of Digital Player Software with your kit, I will bow out...

 

[cvanc]

Apologies, I don't visit this forum often. Here's my kit.

Electronics are miniDSP: An SHD and also a Flex Eight
SHD output is configured 2.1 with an 80Hz 4th order XO
The 2.0 goes direct to my (amplified) speakers
The 0.1 feeds all bass to the Flex Eight, which in turn feeds all ten subs (I had to pair up a couple)

Speakers consist of electrostatic mains and a ten subwoofer swarm
Mains: Beveridge 2SW-1 (1978 vintage, built in OTL tube amps)
Midwoofers: Eight SUB-1200 from Parts Express
Subwoofers: Two Hsu Research VTF-TN1

Calibration is done in three steps
The "EQ" function in REW is used to flatten the mains
MSO is used to align all the woofers
...And then Dirac Live is the layer on top

 

[ddude003]

Interesting bit of kit, ... An article in stereophile.com suggests "Frequency response: 40Hz–15kHz ±2dB over 180°. Lowest useable output: 35Hz." I saw another article suggesting the crossover between the Stat panel and the separate Subwoofers (SW-1) was 100Hz... And your using MSO for all subwoofers including the SW-1s? And layering DiracLive last... Seems reasonable to me... When you say that you want to "flatten the mains", do you really want flat or are you going to use some target like Harmon or B&K house curve? With all those subs I hope you have some physical room correction (traps and absorbers)...

 

[cvanc]

I'm getting close to having the system I will live with forever. At least I think so ":^)

Across their (very small) production run, the 2SW-1 crossover was varied between 70 to 100Hz. I've experimented between 55Hz & 100Hz and always came back to 80. Going lower than 80 got me increased THD and going above 80 had no sonic benefit. And I am steep to protect my panels, 4th order Butterworth slope.

Yes I am using MSO on all of my subs (the Flex Eight exists in the system solely to hold 8 individual MSO cals, otherwise it's just set to unity gain)

And by flatten the mains I actually mean calibrate them to be flat at the MLP. No Harman or house curve. Actually flat. I realize that's not conventional but I think the unique radiation pattern of Beveridges rewrites some of the long standing beliefs.

I'd love to post the mdat file of my sweeps for all to chew on but it's ten megs big.