EQ concept for speaker spread/postion from listener

[Jimp540]

It has been said that the proper spread for speakers is an equilateral triangle form by the listener position ang the two speakers. This would indicate a speaker spread of 60 degrees or 30 degrees to the left and right of center. Most say that spreading them further apart can deliver a wider sound field at the expense of being able to localize instrument positions as well and that too narrow of a spread will decrease the sound field and even muddy the sound. My room layout limits my speaker position to an angle closer to 45 degrees. I have used REW and acheived a nice flat response, a room curve result, and numerous variations yet the perceived sound always seems a bit dull. Through limited research online I have learned that our ears polar frequency response varies a lot with higher frequency. Very high frequencies are most susceptible to this phenomenon. The varied frequency response has to do with the shape of our ears and our abilities to determine the direction of sound and... But the bottom line is that as a sound source becomes closer to the center position in front of us, high frequencies are greatly reduced as far as what we hear, this rolloff seems to begin around 3-4KHz and increases with higher frequencies. If setting up a set of speakers with flat response at the ideal 60 degrees sounds great, then because of our directional frequency response, those same speakers will sound less bright the closer to center they are moved. From the plots I have seen, that 30 degree point does seem to be the location beyond which the rolloff can increase quite noticeably. Going wider not so much. I have tried adding a gently sloped EQ boost starting just below 3KHz adding up to about +3db by 16KHz and it seems to bring back a transparency and sparkle to the music that is lost simply due to positioning. You can experience the change in sound yourself by playing one speaker with music or even pink noise and rotate your head back and forth from looking directly at the speaker then turn to about 45 degrees or more away from it and you will hear how much the highs change especially when they are nearer to center. This is totally different than loudness EQ where the volume of the sound changes our perception of bass and treble. This is more of a geographical physical adjustment that I think might need further discussion when it comes to room layout and EQ. Room reflections can also be involved to a degree but our human ears are frequency directional; and in a way that doesn't FULLY lend itself to a calibrated mic and REW. Thoughts? Maybe there could be an EQ table constructed for a recommended offset to accommodate speaker positions outside of the 60 degree spacing, especially when going narrower, (based on human ear polar response), where the hearing change is greatest. This info may exist somewhere but I cant find it!

 

[ddude003]

Maybe your looking for something along the lines of the Fletcher Munson Curves only using speaker axis... Yes, speaker placement also includes on and off axis response... It might be that what you are trying to figure out is a highly subjective and a personal preference... You might also explore more about DSP solutions for stereo width and imaging manipulation... Maybe something along the lines of EQ plugins that include stereo width and imaging manipulation... Here is a little information that may take you in the direction your looking for... https://www.masteringthemix.com/blo...X89BFNWlwMlot3H_PxD32kcshmDqQ5csM2gTjDd8UmtzL

I personally use Room Correction DSP (REW + HangLoose Convolver) to get a "flat" baseline to which I add an additional EQ plugin that includes some of these techniques...

REW is a great tool for what it is... And there are many more interactive plugin tools that will allow you to add the spice to taste...

 

[Jimp540]

Maybe your looking for something along the lines of the Fletcher Munson Curves only using speaker axis... Yes, speaker placement also includes on and off axis response... It might be that what you are trying to figure out is a highly subjective and a personal preference... You might also explore more about DSP solutions for stereo width and imaging manipulation... Maybe something along the lines of EQ plugins that include stereo width and imaging manipulation... Here is a little information that may take you in the direction your looking for...

I personally use Room Correction DSP (REW + HangLoose Convolver) to get a "flat" baseline to which I add an additional EQ plugin that includes some of these techniques...

REW is a great tool for what it is... And there are many more interactive plugin tools that will allow you to add the spice to taste...

Thanks for the link to that info. The one thing the mixer usually has going for him/her is that they are in an acoustically treated room with speakers/monitors placed in optimal positions and angles. What I was referring to is the fact that if you are sitting directly in front of a speaker at a normal listening distance, facing the speaker head on, the speaker will sound noticeably duller than it would if you turned your head to a 30 degree angle. The directivity of the speaker will not have changed, just the direction that sound enters your ear. Our ears are more sensitive to high frequencies to the sides of our heads. So taking two speakers into an anechoic room and adjust the toe in to point to the listening position, those two speakers will sound different if the spread between then is less than the ideal 60 degrees. Ignoring imaging and soundstage, perfectly flat speakers will not sound flat if they are too close together simply by being further “off axis” to our ears. What Im hoping to get a handle on are some numbers that describe the change in frequency response that we hear as a source moves in position around our human heads so I can have a starting place for an EQ to compensate for narrow speaker placement as it pertains to freq response. I do use REW and miniDSP to achieve flat response then make additional tweaks to that but would like a little more science to supplement my adjustments.

 

[JStewart]

Through limited research online I have learned that our ears polar frequency response varies a lot with higher frequency. Very high frequencies are most susceptible to this phenomenon. The varied frequency response has to do with the shape of our ears and our abilities to determine the direction of sound and... But the bottom line is that as a sound source becomes closer to the center position in front of us, high frequencies are greatly reduced as far as what we hear, this rolloff seems to begin around 3-4KHz and increases with higher frequencies.

for what it’s worth, ChatGPT indicates it’s the other way around. There are some sources referenced.

Yes, human hearing ability for high frequencies (above ~3 kHz) decreases as the sound source moves from 0 degrees (directly in front) to 90 degrees (to the side of the head). This is primarily due to the directional sensitivity of the outer ear (pinna) and head shadowing effects:


1. Pinna Effects: The outer ear (pinna) enhances certain frequencies, particularly in the 2-5 kHz range, when sound comes from the front. As the sound source moves to the side, this enhancement decreases, reducing sensitivity to those frequencies.


2. Head Shadowing: Higher frequencies have shorter wavelengths and are more easily blocked by the head. At 90 degrees, the sound has to diffract around the head, leading to a reduction in intensity, particularly for frequencies above ~1.5 kHz.


3. Interaural Level Differences (ILDs): For high frequencies, the head creates an acoustic shadow on the far ear, reducing the amplitude of sounds reaching it when the source is at 90 degrees.

I then asked if there was underlying data showing the effect by frequency.

Yes, there is data showing how sound pressure level (SPL) changes with angle due to head-related transfer functions (HRTFs) and directional sensitivity of the outer ear. The effect varies by frequency, but some general trends can be observed from studies on HRTFs and binaural hearing research.

Typical Attenuation of High Frequencies from 0° to 90°

Below 1 kHz → Little to no attenuation (less than 3 dB).
3 kHz Mild attenuation (~5 dB).
8 kHz More significant attenuation (~12 dB).
Above 10 kHz Severe attenuation (~20 dB or more).

Key Data Sources

1. HRTF Measurements (Blauert, 1997)
Studies show that frequencies above 4 kHz can drop by 10–15 dB SPL as the source moves from 0° (in front) to 90° (side of the head).

2. Shaw (1974) - Pinna & Head Effects
• Around 3–5 kHz, the pinna boosts frontal sounds by 5–10 dB.

• At 90°, this enhancement is lost, leading to an apparent drop in SPL.

3. Gardner & Gardner (1973) - Head Shadowing
• At 8 kHz, attenuation is 12–15 dB at 90° due to head diffraction.

• Above 10 kHz, attenuation can be 20 dB or more.

It appears the effect is the same with one ear vs 2.

Yes, the effect would still hold true for a person who is deaf in one ear, but with some important differences:

1. For the Hearing Ear (Same Side as the Sound Source)
If the sound source moves from directly in line with the ear (0° on-axis with the ear) to in front of the face (90), the pinna effects would still apply. High-frequency sensitivity would increase as the ear captures more frontal reflections and filtering effects.

If the sound source moves from in front of the face (0° relative to the face) to the side of the head (90° relative to the face, in line with the ear), then high frequencies would decrease due to loss of frontal pinna gain.

No head shadowing occurs for the hearing ear when the sound source is on that same side.

 

[Jimp540]

Interesting stuff! It will take me a while to sort this stuff out. It appears contrary in some ways to polar plots of hearing at various frequencies that I had seen as well as my own personal observations. More to learn! Thanks!!

 

[Jimp540]

Just ran a quick test with a few fixed frequencies. It is going to be complicated. Around 5kHz the sound does indeed decrease as it moves from front to side. 7kHz does the opposite. 11kHz has several nodes, one at an angle about where a speaker might be and another to the side, near 90 or more degrees, possibly where a nice surround speaker might sit!?!? More investigating to come.

 

[Jimp540]

I have done a bit more subjective testing and I have found that our ears do increase the gain of frequencies in the 2k to 7kHz range as the source moves from 90 degrees toward the front of our heads however this gain reaches a peak at around 30-45 degrees (which corresponds to recommended speaker placement) but then drops off significantly as the source moves beyond that point towards zero degrees or straight ahead. So with my speakers being located tighter together than ideal I used a parametric EQ to apply a low 0.7Q boost centered around 4300Hz of 2-3db and it made a great improvement to what I was hearing. IMO. More testing to come

 

[JStewart]

I have done a bit more subjective testing and I have found that our ears do increase the gain of frequencies in the 2k to 7kHz range as the source moves from 90 degrees toward the front of our heads however this gain reaches a peak at around 30-45 degrees (which corresponds to recommended speaker placement) but then drops off significantly as the source moves beyond that point towards zero degrees or straight ahead. So with my speakers being located tighter together than ideal I used a parametric EQ to apply a low 0.7Q boost centered around 4300Hz of 2-3db and it made a great improvement to what I was hearing. IMO. More testing to come

Interesting finding. Thanks for sharing.

 

[ddude003]

Very interesting... That is an interesting place to have issues in, 2k to 7kHz... With 4.3kHz to 4.5kHz as a center right near the border of what would be mid to highs and even extending into the air band... If you are using your E/Vs that would seem like a touchy area... I saw a "stock" frequency response measurement, and they seem to roll of pretty fast up there... This is an area that I use dynamic EQ and even a dynamic air shelving filter in my kit... I wonder what a shelving filter would do for your kit? And interesting enough 432hz is a magic number... Wondering if there is some interplay between those highs and lows... Are you driving your horns with a Power Paradigm amp?