Which combination of mains & subs to EQ to house curve (for correct SPL)?

Porkchop

Registered
Thread Starter
Joined
May 14, 2020
Posts
1
TL;DR: When EQ'ing/leveling a stereo setup (L&R mains, 2 subs) to a particular target (e.g., 75db flat, Harman curve, etc.), do you match each speaker and each sub separately to the target level, match the combined subs to the target level, match the mains combined to the target level, or . . . ?

For a stereo setup (music), I have L & R mains and two subs (Lsub, Rsub). Everything goes through a miniDSP 2x4HD, the subs each get the combined L+R signal. My question is specifically which combination(s) of these should be used to EQ to a target house curve? I've searched for a few days, looked through the guides, and haven't been able to find a specific answer.

I'm specifically concerned about making sure the levels are correct across the range once I'm using all speakers together (listening to music). I've seen some guides say to measure all mains and subs separately, some guides say to measure all subs together. Most guides seem to suggest EQ'ing the subs together. However, measuring subs together leads to a higher SPL vs measuring them separately. Thus if I'm trying to match a Harman curve with a target of 75db (so ~81db at lowest frequencies), I'll set lower volumes for the subs (or apply a more aggressive EQ) in the miniDSP if I'm doing subs combined compared to subs separately, to reach that 75db target curve.

The same question applies to EQ'ing the mains to the house curve. I understand I should EQ the Lmain and Rmain separately, but should each be EQ'd to a target of 72db (so they combine to ~75db when both are played) or each to 75db? Is Trace Arithmetic A+B useful here?

What got me thinking is the guides that instruct to measure L+all subs and R+all subs, such as the AustinJerry guide. Wouldn't that lead to an SPL mismatch between the mains and subs, since the subs will have summed output compared to the single main speaker?
 
Last edited:
Not sure about the expectation that the information here from AV NIRVANA will be better or more accurate than all those “guides,” but I’ll play along. :)

I'm specifically concerned about making sure the levels are correct across the range once I'm using all speakers together (listening to music).
I do the adjustments with the L/R plus subs, ’cause that’s the way I listen to the system. Typically when things sound balanced, the subs will end up measuring a higher SPL than the L/R.

The same question applies to EQ'ing the mains to the house curve. I understand I should EQ the Lmain and Rmain separately…
No. The L/R should be equalized with matching filters. Otherwise stereo imagine sounds whacked in the frequency range where the filters mismatch. Equalizers work via phase changes, so that probably accounts for it.

Below the Schroeder frequency, you can EQ the L/R with separate filters. There are calculations for that, but basically the point where the signal can’t be localized (i.e. which speaker it’s coming from), that’s your room’s Schroeder frequency. It’s best to use filtered pink noise as a signal source for this, not sine waves.

Also, multiple subs should be equalized as a single unit – after all, that’s the way you hear them.

The same question applies to EQ'ing the mains to the house curve. I understand I should EQ the Lmain and Rmain separately, but should each be EQ'd to a target of 72db (so they combine to ~75db when both are played) or each to 75db? Is Trace Arithmetic A+B useful here?
Really, people fret too much over this stuff. Just eliminate the worst peaks and depressions in your mains and subs, adjust the level between them for a good-sounding blend, and go. :) It probably won’t get as much improvement as you expect by chasing every little ripple in response with a zillion filters. Good-quality mains shouldn’t need more than 2-3 filters, if that. If the subs need more than 3-4 filters, placement probably needs to be re-evaluated.

Regards,
Wayne
 
Last edited:
Back
Top