Fighting to EQ +20dB room gain below 60Hz

[rajdude]

Hi Folks,
I could use a little bit of help here. Please take a look at my subs measurements. I got way too much room gain.

I want to have a smooth curve, maybe with a little rise at the low end, something like the Harman curve, but first I should be able to make it flat, correct?


For the past few hours, I am trying to generate filters which can tame this huge, over 20 dB hump below 60 Hz, but have been mostly unsuccessful. Here is what I have tried so far:
Reducing the output of the subs, but that also reduces mid-bass region.
I have time aligned the 2 subs using the time alignment tool. That does nothing.
Created manual filters to try to boost the 60 to 80 Hz range, it helps a bit.

Here is what the curve looks like with the auto generated EQ filters

Thank you in advance for your help!

 

[moedra]

Typically I avoid boosting anything in the filter. Try lowering your target or adding more EQ bands. Either that or, if you use a computer to run audio, follow my guide. You can add a small bump (low shelf) in the target if you'd like.

Digital Room Correction with REW | Meridian Soundworks

nverxion.wixsite.com

 

[Wayne A. Pflughaupt]

What are you using as an equalizer?

Have you tried different loctions for the subwoofer?

What are the main L/R speakers?

Regards,
Wayne

 

[rajdude]

Typically I avoid boosting anything in the filter. Try lowering your target or adding more EQ bands. Either that or, if you use a computer to run audio, follow my guide. You can add a small bump (low shelf) in the target if you'd like.

Digital Room Correction with REW | Meridian Soundworks

nverxion.wixsite.com

Thanks for your reply.
I have tried lowering the target to almost 65 dB, but then the dip becomes even more, I mean with the automatic EQ filters REW generates.

I think next I should try adding all filters manually and yes, more filters!

 

[rajdude]

What are you using as an equalizer?

Have you tried different loctions for the subwoofer?

What are the main L/R speakers?

Regards,
Wayne

Hello Wayne,
Thanks for chiming in.

This is my current rig:
Nvidia Shield TV >
Emotiva XMC-1 >
Bryston monoblock amps>
Magnepan 3.6R fronts & a CC3 center + some Polk boxes for surrounds
2 Rythmik 15" servo subs ported

This rig is in a traditional colonial house's family room which is open to the kitchen area (think large rectangular hall). I cannot move the subs, my lovely wife would kill me ;-) She's already upset at the size of these beasts, LoL. Right now the 2 subs sit right next to the Maggies and the 85" TV sits on top of them.


PS: I have tried Dirac on the XMC-1 (I have the full license), but it ends up destroying the beautiful native sound of the Magnepans. Hence this REW EQ route.

 

[rajdude]

ok I am reading this now :-)

On Minimal EQ, Target Levels, and a Hard-Knee House...

www.hometheatershack.com

 

[Wayne A. Pflughaupt]

Not seeing an equalization device for the subwoofer in your equipment list? Does the Emotiva do parametric EQ for the subwoofer?

Regards,
Wayne

 

[rajdude]

Not seeing an equalization device for the subwoofer in your equipment list? Does the Emotiva do parametric EQ for the subwoofer?

Regards,
Wayne

Yes, the Emotiva XMC-1 has 11 PEQ bands and delays for two subs, totally independent of each other.

 

[Wayne A. Pflughaupt]

Thanks for the info.

For starters, you don’t want individual filters for the subwoofers. Filters, if applied separately, should be matching.

Second, you should evaluate the subs’ performance with the main speakers. According to a Stereophile review I found, your speakers have strong bass output. As such, it should be easy to blend them with your subwoofers. I expect that a combined measurement will remedy the issue below 50 Hz once a proper crossover frequency is in place.

Lastly, do not apply any subwoofer filters below the crossover frequency, as they will be “blown out” by the main speakers once they’re added.

So – get a combined reading, and let’s see what that looks like.

Regards,
Wayne

 

[rajdude]

Thanks for the info.

For starters, you don’t want individual filters for the subwoofers. Filters, if applied separately, should be matching.

Second, you should evaluate the subs’ performance with the main speakers. According to a Stereophile review I found, your speakers have strong bass output. As such, it should be easy to blend them with your subwoofers. I expect that a combined measurement will remedy the issue below 50 Hz once a proper crossover frequency is in place.

Lastly, do not apply any subwoofer filters below the crossover frequency, as they will be “blown out” by the main speakers once they’re added.

So – get a combined reading, and let’s see what that looks like.

Regards,
Wayne

Thanks for your suggestions. Here is what I measured the 2 fronts + 2 subs last night.

I think I may have to lower the crossover frequency from 80 Hz to maybe 50 or 60 to get rid of the suck-out. I was thinking of having the 2 large subs take do the hard work of handling the bass and let the Maggies handle upper mid-bass etc. But somehow these subs seem to have no output above 45 Hz, which I cannot believe....must be my room mode or something.

Maybe I also should try to invert the phase of the subs and test, maybe the inversion of polarity somewhere in the connectors/circuitry is causing such a huge null in the crossover region....what do you say?

 

[Wayne A. Pflughaupt]

Yes, reversing polarity is certainly worth a try. If the Emotiva allows time adjustment between the mains and sub, might play with that too.

Regards,
Wayne

 

[rajdude]

Yes, reversing polarity is certainly worth a try. If the Emotiva allows time adjustment between the mains and sub, might play with that too.

Regards,
Wayne

Yeah, just found out an interesting thing. I was using the RTA to check in realtime, what is going on, what is causing the huge null. I noticed that without the subs running ( fronts in full range mode) there was little output below 80 Hz. Whereas these speakers are supposed to have good output down to 40-50 Hz.
So I did a quick test, swapped the L speaker wires. Suddenly the RTA seemed to be higher in the lower. Then I decided to run a sweep. Here are the results.

What is going on here? Do I really see that the front two main speakers work better together when they are polarity reversed??
If it is true, I would be damned! I have been running these speakers like this for nearly 20 years!

IS there a definitive test in REW to check the electrical polarity, I mean if these are correctly wired or not Positive to positive.
Should comparing the Impulse Response tell me something in REW?

 

[sm52]

What is going on here? Do I really see that the front two main speakers work better together when they are polarity reversed??
If it is true, I would be damned! I have been running these speakers like this for nearly 20 years!

I apologize for interfering. I see you are learning something new about your pets. Everything is more complicated. When the right and left sides are of different polarities, the lower bass may sound better, but the rest of the mids and highs will become so sharp that after 5 minutes you will want to return everything to the way it was.
If you upload here the measurements separately for the right side, separately for the left side, and separately for each subwoofer, in an mdat file, It will be possible to say something in more detail. Make all of the above measurements with acoustic synchronization turned on. You can select the right or left speaker as the synchronization reference. The reference must remain the same for all measurements. Microphone in one position.

 

[John Mulcahy]

You can use the sine generator and its harmonic distortion option to create a polarity test signal by adding 2nd harmonic distortion at 0 dB and -90 degrees phase. Play that signal with the mic close to the speaker being checked and look at the output on the scope.

 

[rajdude]

I apologize for interfering. I see you are learning something new about your pets. Everything is more complicated. When the right and left sides are of different polarities, the lower bass may sound better, but the rest of the mids and highs will become so sharp that after 5 minutes you will want to return everything to the way it was.
If you upload here the measurements separately for the right side, separately for the left side, and separately for each subwoofer, in an mdat file, It will be possible to say something in more detail. Make all of the above measurements with acoustic synchronization turned on. You can select the right or left speaker as the synchronization reference. The reference must remain the same for all measurements. Microphone in one position.

no no :-) thanks for "interfering"....actually you are not interfering, you are helping. Thanks for your tips.
I will measure the fronts separately and upload the mdat file here soon.

 

[rajdude]

You can use the sine generator and its harmonic distortion option to create a polarity test signal by adding 2nd harmonic distortion at 0 dB and -90 degrees phase. Play that signal with the mic close to the speaker being checked and look at the output on the scope.

View attachment 74166
View attachment 74167

Thanks for chiming in here, John, I appreciate it.

So I did the test you recommended above. Here are the results. I am not 100% sure how to interpret these graphs, so please enlighten me:

1. Question 1:
Since the measured sine wave graph starts by going downwards first (starting at zero), the speaker's output is inverted as regards to the signal fed to it. Am I correct in that observation?

2. Question 2:
Comparing the L and R graphs, it seems to me that the polarity of my front main speakers is actually correct and "in sync" with each other when the electrical connections are NOT reversed on one speaker. Note that in that configuration, both L & R measured sinewave (the taller of the peaks, not the distortions) are shown going downwards at zero.
Is that observation correct?

Here are the two screenshots of L and R, side by side...



But if I reverse the speaker wires of the left, it now measures like this, note that the bigger peaks are now above Zero line (is that the correct way to say it?):



Thank you, in anticipation!

 

[rajdude]

I have had an interesting observation, while I was doing the tests above:

I noticed that while the polarity of the left speaker was reversed, the 80Hz sine wave tone seemed louder in my room. So I placed a sound level meter I have, set it to C weighted and slow, put it in the center of the room and measured the sound levels in those two situations. SLM was not moved between the 2 measurements

Speakers in default polarity: 54 dB

Speakers in reversed polarity: 63 dB

So what is going on here?
I clearly can see an almost +10dB gain at 80 Hz by running the mains with inverted polarities. But the sinewave graph test seems to show that their polarities are correctly aligned at the usual polarity. Why is this happening?

 

[rajdude]

I apologize for interfering. I see you are learning something new about your pets. Everything is more complicated. When the right and left sides are of different polarities, the lower bass may sound better, but the rest of the mids and highs will become so sharp that after 5 minutes you will want to return everything to the way it was.
If you upload here the measurements separately for the right side, separately for the left side, and separately for each subwoofer, in an mdat file, It will be possible to say something in more detail. Make all of the above measurements with acoustic synchronization turned on. You can select the right or left speaker as the synchronization reference. The reference must remain the same for all measurements. Microphone in one position.

Here is the mdat file for the measurements you ordered

 

[sm52]

The polarity that was in the measurement that you showed is correct for all participants.
But there is actually a problem with subs. They've made war on the main speakers.To make things a little better, you'll have to make a lot of measurements. The essence of these measurements is to find such a combination of frequencies of the main and sub cutoffs so that they become friends as much as possible.

 

[rajdude]

The polarity that was in the measurement that you showed is correct for all participants.
But there is actually a problem with subs. They've made war on the main speakers.To make things a little better, you'll have to make a lot of measurements. The essence of these measurements is to find such a combination of frequencies of the main and sub cutoffs so that they become friends as much as possible.

Ummm......sorry I do not understand.....made war? Hmmm. So are you saying that I should try varying the main to sub crossover frequency in my pre-amp processor and take measurements....hoping to find a crossover frequency which results in a more flat transition?


This will become a hit-and-trial process. Should not there be a more scientific way to do this, since we have so many tools at our disposal now? For example, REW has an alignment tool, but wait, that it only for time alignment of sub-mains.

Maybe there is another tool or plugin which can analyze and recommend a more suitable crossover frequency?

 

[Wayne A. Pflughaupt]

Ummm......sorry I do not understand.....made war? Hmmm. So are you saying that I should try varying the main to sub crossover frequency in my pre-amp processor and take measurements....hoping to find a crossover frequency

Yes, that's what he's saying. However, I don't necessarily see the subs and mains fighting each other. The evidence for that would be that combined response would look worse than independent measurements. That wasn’t the case here – combined response showed the same problem with the sub that the separate sub measurement showed.

This will become a hit-and-trial process. Should not there be a more scientific way to do this, since we have so many tools at our disposal now?

Not necessarily. You've read my Hard Knee Housecurve article where I noted I didnt like the way my system sounded, despite the graphs looking wonderful. REW certainly helps - imagine just doing all this by ear and not even realizing there are problems. Of course ignorance can be bliss, but correcting something that might not be immediately obvious can get improved sound quality, once you are aware of the problem and are able to correct it.

I've never done a new setup upon moving to a new house that didn't require much trial and error. On at least one occasion I had to forgo conventional wisdom and run my main L/R full range to get an audibly satisfying blend with the subwoofer.

That said, why haven't you simply used the Dirac automatic correction your Emotiva has onboard? People seem to be pretty pleased with its results.

For example, REW has an alignment tool, but wait, that it only for time alignment of sub-mains.

Playing with that adjustment could make a difference as well - assuming the parameters could transfered to the XCM-1.

I’d be curious to see measurement of each sub separately.

Regards,
Wayne