pretty much the same patern. try what obi said, to stuff a big pillow or look for an old couch cushin stack in corners behind speaker, just for fun ,it would not be as good as bass trap. but your graph should show sign of that presence. if any better you know that bass trap in those corner , would help

Sorry for sounding harsh but I would be disappointed with those measurements even if they were white van speakers.

It might be time to check your measuring equipment (if you are using Realistic meter for example, check battery health, slow response time, C-weighting, error compensations, positioning (ie. is your meter on a tripod?)) and establish whether both speakers sound correct individually. Then proceed with the more complex.

Bob, a couple things...

Continuing on noway's comment: which mic or spl meter are you using for the measurement? Is the device on a tripod? It's very important that the measurement device be fixed. Holding it in your hand is going to create all sorts of errors.

Also, for now, take your subs out of the mix. Unplug the power cord and remove the RCA connector. Get the Maggies as flat as you can first.

I wouldn't bother with the fluffy pillows or cushions in the corners; rather, I would just buy traps and get them installed (from any of the vendors I suggested in an earlier post). Buy as many as you can afford, both financially and for space.

IF, and that's a possible IF, you get the Maggies and in the right spot AND the bass trapping flattens the response you MIGHT not feel the need to use the subs.

IF at that point you want some extension, then start to read about subwoofer placement - especially material on multiple subwoofer placement.

I think you've got high-frequency reflections giving you the boost above 5kHz, and you've got plenty of standing waves at low frequency giving you the boost down there.

Also, try moving your listening spot. Take measurements at different spots in the room.

_________________
"What you talking 'bout, Willis?."

The boost then cut at 160hz is front wall sbir cancellation. The valley centered on 2khz is left right speaker interference notch. This is a common symptom of a stereo configuration, as measured by a point source microphone with pink noise. Our ears/brains are able to account for this in program material and many other nasties we can measure. Front wall poly diffusion is very useful for aiding dipole issues in general terms. Dipoles have a lot of issues. Adding front wall corner bass absorbtion as mentioned defeats the purpose of owning dipoles in some ways. Rear corners are a better idea in this case.

Your subs should be placed at 1/4 points on front wall for 2ch application in a normal rectangular room. They should be setup in mono if possible. This will take care of the width nulls in your room, assuming your mlp is centered (which it should be.) Your seating position should be selected with respect to length axial modes in your room. I may have missed it, but room dimensions would be helpful in predicting what should go where in your setup.

If you have an Ipad I strongly suggest using "audio tools" by studio six to assess other very important factors such as room decay vs frequency and setting your subwoofer level, crossover and phase correctly. I think you will appreciate the speed and convenience of software assisted measurements.

I'm a bit of an acoustics nut, so I'll help you if I can.

What are sbir, left right interference notch and length axial modes ? Are you suggesting I not get front wall bass traps?

SBIR- Speaker boundary interference response. The destructive summing of delayed sound recombining with itself out of phase from the original signal.

The left right interference notch is described in Floyd Toole's book as the summing of a mono signal through a stereo playback system. This is usually manifested in centered vocals. The specific 6db 2khz dip has to do with the relation of travel distance between your ears and sound wavelength. It is refered to as the "Phantom Comb" in his book. Not much you can do about it. No need to worry about it.

Axial modes cause the most dramatic peak and null effects in a listening room. They are responsible for extreme dips and peaks in the bass region below the rooms transition frequency. The transition frequency is an imaginary dividing line where below it room modes dominate frequency response errors, above it reflections and comb filter distortion dominate frequency response errors. The transition frequency is based on room volume. The larger the volume, the lower the frequency. Most domestic sized rooms land somewhere between 300-500hz.

Front wall traps are a great idea, but they are in direct opposition to the radiation pattern of your speakers. The dipole figure 8 radiation pattern means 1/2 of the acoustic energy is sent to the front wall. If you treat this area with absorption you are effectively lowering the sensitivity of your speakers wrt sound power in the room. This will however eliminate some of the sbir and comb filter errors you see in your measurement plots. Its a trade off.

I would recommend diffusion up front instead as a means of preserving energy while reducing destructive summing you are seeing in your plots. A nice benefit of QRD's specifically is your sweet spot size improves considerably. Ideally, you would flip the system to the long wall adjacent the patio so you can treat your new front wall without worry of blocking the window with stuff. This will also yield weaker lateral reflections which are very colored with dipole setups.

To move the speakers to the front wall without moving the amps would require really long speaker cables. A huge expense. I think I'll build some traps and see what they do and then post some curves. Thanks a lot for your help, I really appreciate it.

Can you estimate what the curve would look like with panel base traps in the front corners and one in the corner behind and to the left (looking toward the speakers) of the listening position? I'm thinking 4 1/2 inches of dense fiberglass in galvanized channels from floor to ceiling two feet wide.

4 1/2" traps would be broadband absorptive down to about 100ish hz. If you place them on the sidewalls pushed tight into the front corners you will also cut down on destructive corner bounce. I hesitate to comment on low bass absorption, as pressure based membrane absorbers are best used in corners.

You should see an improvement with smoother response and weaker nulls. They wont go away though. Placement fixes that. The biggest improvement will be a general reduction in your room's decay time -yielding a more dynamic presentation as transients are better seperated.

I found your 18x33 dimensions. A good guideline is to place your listening position at either 0.2, 0.32, 0.45, 0.55, 0.68 or 0.8 times the rooms length. This translates to positions where length modes are neither at their peak or null points. How open are you to keeping the same general layout, but moving your chair and speakers around? I've done room mockups for others and a dipole mockup would be a fun challenge.

I tried something else. Pointing the backs of the speakers toward the front corners. I got a bit more refined curve. Any thoughts?

You still aren't even measuring correctly. Magnepans are nice speakers but any speaker can be made to look bad if measured incorrectly.

Noway. Can you tell me how to measure them correctly. I'd be grateful.

Change the meter from fast-->slow since you are better off measuring average not peak levels.

I don't use Rives Audio software but I understand from reading that there are more than one set of test tones, one set compensates for Realistic meter error. But there are several versions of the meter and I would expect each model number to have its own error profile. You should investigate that, based on the serial number of the meter. If the compensations aren't meant for your particular model, you may have to use the uncompensated tones in the software and manually make corrections based on internet info. I use Model 33-2050 and a compensation chart I got off the internet for that particular model. My meter is analog display so doesn't have all the bells and whistles of the newer model.

Measure at an appropriate volume level. I would suggest establishing 80dB at 1kHz and don't touch the volume control on the amp until you're done. Also I think there are better ways to measure than Radio Shack meter and test tones that give a more accurate picture.

I personally prefer to set speakers up by ear and use the measurement process only for adjusting bass frequencies (ie. 200 Hz and below).

Start by taking measurements of one channel at a time no subs. Use a test disk with dual mono pink noise or download a track. Stop using an spl meter. You need to look at an rta or fft plot. You are not getting very good resolution with your current method. 1 octave resolution perhaps? You need at least 1/3 octave resolution. 1/6 is pretty good and eliminates grassy noise in the measurements.

Once you find out each channel behaves differently you will have to find the asymmetry causing your speakers to have different responses. The common problem trends between speakers will be associated with the low end most likely. (Below transition frequency). Problems above can be treated out effectively. Problems below need placent of speakers/subs and your mlp looked at.

It looks like your response is +/- 8 db or so. This isn't terrible for a raw room with dipoles.