[chargedmr2]

Nuance - I did try reversing phase as an earlier solution, but the change did not affect the peak.  Also, I'm using a Denon 3312 for the mains and it doesn't have an adjustable crossover (that I'm aware of).  So I'm stuck with the 12db slope for now. 

I agree that the room mode can be EQed, but since it's caused by the mains, my sub EQ is of no use.  It's already set to cut the maximum amount (15dbs) at 63.4Hz.  I will try using the graphic EQ that is built into my AVR to see if that can help.  I was able to confirm that when Audyssey is set to manual mode, it simply acts as a graphic EQ and does not apply any of the Audyssey corrections from the measurement procedure.  If this works out I'll repost the frequency response.

[ricardojoa]

It's a room mode and messing with the ports will do nothing positive.  Besides, I'm fairly certain the ports are tuned far below the 63Hz area, so plugging them makes no sense.

There is nothing wrong to plug the port for the SAKE of TESTING. Pluging the ports can make the fronts to roll much ealier naturally, essentially taking unwanted bass. Chargedm2 could try or not, Im just giving a suggestion. And by the way, even if the port are tuned far below 63hz, doesnt mean there is no information comming from them, even if it is HP.

[Nuance]

Nuance - I did try reversing phase as an earlier solution, but the change did not affect the peak.  Also, I'm using a Denon 3312 for the mains and it doesn't have an adjustable crossover (that I'm aware of).  So I'm stuck with the 12db slope for now. 

I agree that the room mode can be EQed, but since it's caused by the mains, my sub EQ is of no use.  It's already set to cut the maximum amount (15dbs) at 63.4Hz.  I will try using the graphic EQ that is built into my AVR to see if that can help.  I was able to confirm that when Audyssey is set to manual mode, it simply acts as a graphic EQ and does not apply any of the Audyssey corrections from the measurement procedure.  If this works out I'll repost the frequency response.

Bummer. 

So how is it that the peak is higher/larger with the subwoofer engaged than without it?  If I look at the response from 90Hz and above it appears you didn't change the volume, so if the peak was caused only by the speakers why did it increase when you put the subwoofer into the mix?  They shouldn't be overlapping that much at 63Hz unless the mains are still running full range.  Weird...

It is possible that all the positions you've tried the speakers in are still causing the peak.  Try measuring the speakers full range with no crossover engaged, then set the crossover and measure them again.  Don't change anything but the crossover implementation, then post the graph here.  Your Denon should allow differing crossover points, so try 90 or 100Hz as well and see if you're able to EQ the peak.  If not then something is wrong.  Oh, and this might go without saying, but be sure your speakers are set to "small" on the receiver. 

[jsalk]

What is the distance from your floor to your ceiling?  Keep in mind that at the frequencies involved, you are not only measuring the speakers, but the room as well (you probably knew this).

- Jim

[chargedmr2]

Bummer. 

So how is it that the peak is higher/larger with the subwoofer engaged than without it?  If I look at the response from 90Hz and above it appears you didn't change the volume, so if the peak was caused only by the speakers why did it increase when you put the subwoofer into the mix?  They shouldn't be overlapping that much at 63Hz unless the mains are still running full range.  Weird...

It is possible that all the positions you've tried the speakers in are still causing the peak.  Try measuring the speakers full range with no crossover engaged, then set the crossover and measure them again.  Don't change anything but the crossover implementation, then post the graph here.  Your Denon should allow differing crossover points, so try 90 or 100Hz as well and see if you're able to EQ the peak.  If not then something is wrong.  Oh, and this might go without saying, but be sure your speakers are set to "small" on the receiver.

I'll give this suggestion a try--I think I did this already, but I didn't save the graph.  I also tried different crossover points up to 120Hz and the peak decreased a couple db, but did not disappear entirely. 

[chargedmr2]

What is the distance from your floor to your ceiling?  Keep in mind that at the frequencies involved, you are not only measuring the speakers, but the room as well (you probably knew this).

- Jim

Hi Jim,

The ceilings are 10ft high. 

Yes, this is definitely all room effects.  I did take some close-up measures of my sub and the response looks absolutely perfect.  And it's been a while, but I have also taken close-up measures of my SongTowers--of course they too measure beautifully.  So, I hope I am not detracting from Salk speakers by posting the effects of my room on them.  I should also mention that things are sounding very good, and I'm really enjoying the learning and experimentation. 

[chargedmr2]

OK, I applied a 6db cut at 63Hz to my main speakers (via my AVR in manual EQ mode).  As I expected, this solved the problem.  Here are my latest measurements, all taken with my mains crossed to the sub at 80Hz, just as before (Nuance, I didn't take the measures you were looking for since things appear better now):

No smoothing applied (+/- 3db from 10Hz through 80Hz):


Same measure with 1/6 smoothing applied:


And here is the full range measure with 1/3 smoothing applied:

[Nuance]

15-200Hz looks fantastic now - congrats!  I would leave it alone and move your focus to the area between 300 and 5000Hz.  Although, I still think something is wrong if even moving the crossover to 100Hz didn't do anything.  It still seems as if the crossover simply isn't working, but if it looks and sounds good...

I know you probably don't want to use EQ on the speakers above the Schroeder Frequency, and I am not a proponent of it, but if moving the speakers around a bit doesn't fix 300Hz-5KHz, you may want to give it a shot and see if it sounds better (it should).

Happy listening!

[chargedmr2]

Although, I still think something is wrong if even moving the crossover to 100Hz didn't do anything.  It still seems as if the crossover simply isn't working, but if it looks and sounds good...

I agree that my results aren't immediately intuitive.  The peak actually was reduced by small amounts when moving the crossover progressing from 80 to 120Hz, so the crossover was doing something for sure.  But I see your point...why would the peak remain at all if the crossover was moved up substantially?  I'll find some time to take more measures and solve this mystery once and for all.