Stereo sound interference?

[Mishaiger]

Hello everyone, the short version of my problem is that two speakers play quieter than when only one speaker is playing.

Long version:
First i would like to describe the room i'm listening in. It's small very treated with absorption panels room, i.e. ceiling, window, door, every wall except floor, this results in ~90ms RT60 decay time, if this is of some value.

The problem seems to be in mid range, so we'll be looking at 20-1000 hz graphs.
Either left or right speaker plays fine. When measured alone they are fine:

But when they play at the same time it gets much worse:

This is LR - 6 dB. It matches closely up to about 200 hz, but then gets further away. Difference in loudness is really noticeable in female vocals.

I started investigating, replacing speakers to a different one yield similar results.
But using my current speakers in different untreated room with bare walls and some furniture results in almost perfect SPL doubling:

This is vector average of L and R + 6 dB vs real stereo.


My conclusion is that room treatment causes this effect, but i really like the sound with very short decay times. So, is there any way to fix this without heavily modifying the room?

 

[FeddyLost]

I'd ask on RMM and possibly add some diffusion to get at least "studio standard" RTs.
Also, you need to remember that your pair of ears is not an single small omni microphone, but something little more complicated.
For example, when I pointed acoustical axes of my speakers at mic, destructive interference in mids was very deep, but subjectively all was OK.

 

[Mishaiger]

Yeah, i understand that mic is not ears, but i've confirmed everything by ears. Most noticeable was quieter female vocal. I tested it with Eq APO, i subtracted Mid-Side and then applied eq to Mid, alas, Eq APO cannot perfectly mix it back together, mixed sound resulted in more mono than it should, but was just enough for me to listen and test.

Here is an update though; i removed 2/3 of absorption panels and that helped, but not completely eliminated the problem. So i guess that was the problem after all.

 

[wwenze]

The graphs aren't the un-normalized SPL values so I'm not sure what is happening...

Is double speaker softer or staying the same volume as single speaker in your first room?

Is double speaker double volume in your second room? (Also I'm pretty sure it's 3dB not 6 for SPL doubling from having 2 speakers. 6dB would mean further constructive interference)

 

[Mishaiger]

First, i'm sure that double SPL means +6 dB in speakers, i think you are referring to amplifier power, where double watt would indeed be +3 dB.

Second, double speakers play roughly at the same volume as single, i can clearly hear it in tests and music. In tests audio simply moves from side to mid, not becoming any louder.
In another room sound is definitely louder when both speakers are playing.

 

[NTK]

I think this is likely a case of the comb filtering that happens when 2 separated sources (separated by more than a wavelength of the sound) playing the same sound and the measurement point is not equal distant from both sources. This is also the cause of the "phantom center dip" of 2 channel stereo. The effect is much more dominant in heavily treated rooms where the direct sound is a lot stronger than the reflected sounds (which have random time delays and help even out the constructive and destructive interferences).

Below clip is from this Audioholics article by Dr Toole. (The dip frequency of ~2 kHz is due to the typical offset distance between our ears which makes equal distance to the 2 stereo speakers impossible.)

Room Reflections & Human Adaptation for Small Room Acoustics

Dealing with acoustics in small rooms is no trivial matter. This article focuses on how to treat early reflections based on listening habits, your loudspeaker choice. Absorb, diffuse, reflect; decide.

www.audioholics.com

 

[DVDdoug]

"Just for fun", try reversing the + & - connections to one speaker...

First, i'm sure that double SPL means +6 dB in speakers

Two speakers is twice the power so +3dB.

I would never say, "double the SPL" (since decibels are logarithmic).

 

[Suffolkhifinut]

"Just for fun", try reversing the + & - connections to one speaker...

Two speakers is twice the power so +3dB.

I would never say, "double the SPL" (since decibels are logarithmic).

The original unit for gain or attenuation was the Bell, decibel is log to the base 10 of a Bell. When telegraph systems were being installed in the US using the Bell calculating the line attenuation was too difficult for the people working on the installations. Replacing the Bell with Log to the base 10 made the calculations much easier.

 

[Mishaiger]

I think this is likely a case of the comb filtering that happens when 2 separated sources (separated by more than a wavelength of the sound) playing the same sound and the measurement point is not equal distant from both sources. This is also the cause of the "phantom center dip" of 2 channel stereo. The effect is much more dominant in heavily treated rooms where the direct sound is a lot stronger than the reflected sounds (which have random time delays and help even out the constructive and destructive interferences).

Below clip is from this Audioholics article by Dr Toole. (The dip frequency of ~2 kHz is due to the typical offset distance between our ears which makes equal distance to the 2 stereo speakers impossible.)

Room Reflections & Human Adaptation for Small Room Acoustics

Dealing with acoustics in small rooms is no trivial matter. This article focuses on how to treat early reflections based on listening habits, your loudspeaker choice. Absorb, diffuse, reflect; decide.

www.audioholics.com

View attachment 187032

Thanks, that correlates well with my findings. I indeed measure twice with gap about the size of human head and then vector average, because i find it to be more similar with what i actually hear.

"Just for fun", try reversing the + & - connections to one speaker...

Two speakers is twice the power so +3dB.

I would never say, "double the SPL" (since decibels are logarithmic).

I tried it "just for fun", similar results. And yeah, double seemed a bit off when i was typing, but couldn't find more appropriate word.
But, for example, one powered speaker = X spl, then two powered speakers with same power will be X+6 dB.