Need Advice on Room Treatment

[Darra Pharma]

Hello All and Happy New Year,

The first attachment shows a diagram of my listening room. I know, other more traditional speaker and listener locations would be better, but this is also a family living room, and this just works best for us. The speakers are Magnepan 3.7i’s, with tweeters on the outside. I have been tweaking this arrangement for a few years now, moving speakers a foot or two in various directions, and adjusting toe-in. This arrangement provides pretty good sound, except, on most recordings, the stereo image seems to pull to the right. The second attachment shows the room’s response as measured with a calibrated mic and REW (Each trace representing the average of 9 measurements around a 28-inch diameter circle centered on the head of a seated listener). I noted the approximate 5dB difference in bass response between the left and right channels, and I am wondering if this might be causing the rightward pulling of the stereo image? If so, would putting a bass trap in the corner behind the right speaker improve the situation?

Thanks for any comments.
Darra

 

[suttondesign]

boundary reinforcement is occurring on the right. hard to fix the room, easier with EQ.

 

[Ogoript]

My guess is the stereo image is partially related to the first reflection. The wide open space to the left of the left speaker vs the right which has earlier reflection points. Are you able to move something to cover the left speaker early reflection (furniture, room divider etc) temporarily to test that theory?

 

[Darra Pharma]

suttondesign, thanks for your comment, I have thought that signal processing might be the way to go. Of course, the next thought is analog or digital? Any advice on that subject?

ogoript, thanks for your comment. By using the mirror technique, I have found the first reflection points for both the right and left speaker. For the right speaker, it is 5-feet from the corner, moving along the right-hand wall. For the left it is 4-feet from the corner, moving along the left-hand wall. There is a couch just to the left of the listening position, it is possible to hang an absorption panel at that point, is that the kind of thing you are suggesting?

Cheers,
Darra

 

[suttondesign]

digital. minidsp flex. it allows eq for one channel. other dsp plugins do as well.

 

[Ogoript]

The absorption might work on the left speaker, but it is more how wide open it is to the dining room vs right which has walls and windows closer to the speaker. Those reflections will bounce back to you first, possibly creating a timing differential and that could impact the imaging. I'm not familiar with he Magnepan's though. Try treating abosorbing the first reflections on each as a test of that theory.

 

[wunderkind]

You can be courageous and move your audio set up into the dining area. Place speakers by the 9' wall area and listening spot at the opposite end towards the hallway.

Dining is overrated.

 

[Darra Pharma]

suttondesign, thanks for that suggestion, I will look into it.

Ogoript, OK, now I understand what you are suggesting. From the impulse response data for these room curves, I have from 5 to 12 reflections in the first 20ms for both channels. I will delve into that data a bit more to see if the reflections from the right speaker are stronger.

wunderkind, I think I might lose my family if I did that! Are families overrated also?

Cheers,
Darra

 

[Darra Pharma]

I decided to go the DSP route. I purchased a MiniDSP Flex/Digital. As I see it, I have three options for dealing with the channel imbalance:

1 – pull up the left channel by +5dB between 70 and 200 Hz.

2 – reduce the right channel by -5dB between 70 and 200 Hz.

3 – pull up the left channel by +2.5dB and reduce the right channel by -2.5dB

Then there is the question of the 40 Hz room mode. I listen to mostly classical music, so there is not much content down there, so will I do more harm than good by knocking it down by 5 or 10 dB?

Thanks for any comments or suggestions.
Darra

 

[kemmler3D]

I decided to go the DSP route. I purchased a MiniDSP Flex/Digital. As I see it, I have three options for dealing with the channel imbalance:

1 – pull up the left channel by +5dB between 70 and 200 Hz.

2 – reduce the right channel by -5dB between 70 and 200 Hz.

3 – pull up the left channel by +2.5dB and reduce the right channel by -2.5dB

Then there is the question of the 40 Hz room mode. I listen to mostly classical music, so there is not much content down there, so will I do more harm than good by knocking it down by 5 or 10 dB?

Thanks for any comments or suggestions.
Darra

My approach would be this:

Looks like you're already using REW. Check out a room correction with REW tutorial here, there are a good number. Basically, REW will automatically generate some filters that will get both channels closer to a single target.

This will have more or less the same effect as what you've proposed manually, but you'll get more filters that result in a smoother response for both channels.

Also, DEFINITELY get rid of the 40hz room mode. "There is not much content down there" is a bit of a myth. Depending on how it's recorded and mixed, all percussion can have some amount of content to 40hz and below. Either way, you invested in the MiniDSP and that's a great use for it!

 

[Darra Pharma]

kemmier3D, thanks for your comments.

I am in the process of reading-up on the REW approach, but will take your advice on knocking-down the 40 Hz peak.

Cheers,
Darra

 

[Darra Pharma]

Just a follow up on my DSP room correction adventure.

I investigated using the REW software to generate the EQ filter, but I found only settings for increasing bass response, and no way to affect a decrease, so I gave up on that approach. I decided to just try decreasing the right channel by 5dB between 60 and 200Hz.

I quantified the subjective pulling of the sound stage to the right; by playing the bass decade warble-tones from track 17 of the Stereophile Test CD3, and noting the perceived location of the sound on a 1 – 10 scale, 1 representing the left speaker and 10 representing the right speaker.

The first attachment shows the results of these evaluations. The first pane in this table shows the perceived soundstage with no DSP corrections. The other three panes show the perceived soundstage with increasing attenuation of the 160 to 63 Hz frequencies. I am sure the precision of these evaluations is fairly low, but it is clear that it takes a full 5dB of attenuation to pull the soundstage to the center. The second attachment shows the MiniDSP transfer function with -5dB of attenuation . I was a bit surprised at the magnitude of the attenuation, but I guess that it is to be expected with the overlap between frequencies when using a Q of 3. I think other Q settings might produce better results, maybe someone can confirm that, and suggest a more appropriate Q.

Thanks for your thoughts.
Cheers
Darra

 

[Darra Pharma]

Concluding my DSP journey.

The first attachment shows the room response with the DSP set for a gain of -5dB and a Q=3 (green is L+R, red is R, and blue is L). As shown previously, these settings will correct the perceived Left-Right imbalance, but the right channel is now pulled down enough to also pull the L+R down more than I want. I tried, without success, various combinations of gain and Q, looking for a combination that corrected the perceived soundstage, and did not adversely affect the overall frequency balance. I did find that a gain of -2dB, rather than -5dB, with a Q=3 does nicely correct the left-right bass imbalance, and only marginally affects the overall frequency balance, see attachment two. Since this does not correct the perceived Left-Right imbalance, I decided to just live with it and just correct the 40Hz hump. The last attachment shows the final room response with just the 40Hz correction.

Cheers,
Darra