• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Increase sweet spot of correlated pink noise

sfiruch

Member
Joined
Apr 14, 2024
Messages
5
Likes
4
Hello everyone! Long time reader, first-time poster here. I have spent a year or so reading research papers, listened to reviews and interviews, spent hours on this forum, spent some money on speakers, absorbers, diffusors and did a lot of experiments. As a result, not only did my friends and partner have to listen to me for hours talking about speakers and room acoustics, I also got an amazing listening room with excellent RT60 and a smooth frequency response. But I'm still learning new things every other week as I continue experimenting.

My most recent discovery is that imaging (sound sources distinctively appear between, or even outside, of speakers) strongly correlates with how correlated/mono pink noise sounds. When I listen from a position where pink noise sounds as if it emanates from a small point, everything sounds fantastic. When I'm listening from another position, I lose the ability to localize instruments. While the music is still stereo, the instruments are somehow elusive, instead of spatially localizable.

Questions: How can I increase the area where correlated pink noise seems clearly localized in the center? Which parameters affect my ability for sound localization? Why does correlated pink noise "widen" when I move a couple inches/centimeters forward or backward?
 
Welcome to ASR! Reading up is admirable.

An acoustic instrument produces fundamentals and overtones. The Klippel Spinorama produces a graph of frequency response against vertical and horizontal listening angle. ASR, reviewer Erin, and the Spinorama website publish those graphs.

I would speculate that within the Spinorama sweet spot, changes in cm in ear position would be room modes and reflections. That is why recording studios spend a lot of money on acoustic design such as "live end / dead end."

The recording space is never perfect. The microphones are picking up all kind of nearfield and farfield reflections from the instruments. We by convention say that the reflections are pleasing at Abbey Road or Carnegie Hall. Our ears are doing different localization in a complex acoustic environment like Carnegie Hall in relation to our complex home listening spaces.

Pink noise is not an instrument, acoustic or electronic, connected to a space where recording takes place.
 
Last edited:
pink noise sounds as if it emanates from a small point, everything sounds fantastic. When I'm listening from another position, I lose the ability to localize instruments. While the music is still stereo, the instruments are somehow elusive, instead of spatially localizable.
I think this is normal. (Personally, I mostly listen to rock which is "artificial" anyway so I just listen and enjoy it.)

I had a link to an article about panning (but now it's gone. :( ) The conclusion was that the mixing engineer shouldn't waste time trying to precisely locate all of the instruments/tracks because you'll lose the position if you move or if you listen in a different room.

I don't think we can "precisely locate" instruments & vocals with live music either (with our eyes closed) but it's been awhile since I've listened to live music, and closing my eyes I'd still remember where the instruments are.

A lot of it is related to the recording. My intuition is that a reflective room (and maybe omnidirectional speakers) can give a you a wider sound and a dead room should give you more precise location.

Why does correlated pink noise "widen" when I move a couple inches/centimeters forward or backward?
The soundwaves go in-and-out of phase as you move around. That includes the direct & reflected sounds. At 5kHz the wavelength is about 3 inches. With about a 1.5 inch distance-difference the waves will be out-of-phase and cancel. The waves can be in-phase in one ear and out-of-phase in the other and/or in-phase from one speaker and out-of-phase with the other.

You'll get similar "weirdness" with high-frequency test-tones. With music and normal sounds we are used to it and we usually don't notice it and the sounds are more randomized.

Have you tried with one channel inverted? You'll get a wide "spacey-phasey" sound. (And the bass gets almost completely canceled.)
 
I don't think we can "precisely locate" instruments & vocals with live music either (with our eyes closed) but it's been awhile since I've listened to live music, and closing my eyes I'd still remember where the instruments are.
For example, in Little Sadie by Crooked Still the cello (?) starting at 0:09 "obviously" plays right over there. The effect is almost eerie with binaural recordings. But it's totally plausible that it's all in my head :)

The soundwaves go in-and-out of phase as you move around. That includes the direct & reflected sounds. At 5kHz the wavelength is about 3 inches. With about a 1.5 inch distance-difference the waves will be out-of-phase and cancel. The waves can be in-phase in one ear and out-of-phase in the other and/or in-phase from one speaker and out-of-phase with the other.
Yeah, I noticed this effect. You can do amazing demonstrations by exploiting room modes. Listening for the same effects at higher frequencies needs a bit more training and time, but is very interesting.

What I don't understand how these effects could possibly affect pink noise. I would've expected that phase shifts affect all listening positions similarly when listening to pink noise. Why is there a small sweet spot where phase shifts are seemingly not an issue?
 
Hi, mono noise is great tool for listening your system! When two sound sources output same sound, and you are equidistant to both, both toed-in similarly and so on that sound from source to your ear(s) from both sources is as alike as possible, then you have the pin point phantom center. There are many things that break it, or prevent one hearing it in the first place. For example, if you move sideways not being equidistant anymore it gets loose immediately, good directivity and toe-in can help with this though ( time intensity trading ). Response of the speakers could be different (to any angle) due to manufacturing tolerances, not having DSP to equalize them as close as possible, for example. In addition, edge diffraction makes response vary to any angle. Speaker operation changes with excursion, so on some systems low frequency sound on either side could break phantom center just because the speaker now modulates so much the correlation between L/R is broken. In addition, early reflections from room boundaries and objects could be quite loud, further affecting things even if the speakers were fine. In reality, many people don't pay attention on any of this, and they never have the pin point imaging and it's fine to most.

The important thing is to realize, the pinpoint imaging of phantom center is formed into your reality, into your perception, by your own auditory system! I use this effect to be aware of my auditory system, which then allows to utilize logic on sound I perceive in order to adjust my speakers. the speakers are built by me and it is very important to be able to somehow reason with what I hear in order to try and improve on things with the system, in my room, with my own preference and liking. Believe or not I use the pin point imaging to utilize my own auditory system to be able to reason about the playback system. If you are interested, look for David Griesinger studies about proximity and localization, basically most of his studies seem to touch the subject although mostly in concert hall and live music context, but it's the perception he studies, which we all carry with us no matter if it's concert hall or livingroom.

In short, I've found out that all the rooms I took my speakers into there is some particular size of stereo triangle beyond which this pin point phantom center doesn't happen anymore. But, I've always found it just by moving forward closer to speakers, effectively increasing direct/reflected sound ratio, until at some point the pin point imaging happens quite suddenly. I think this is quite easy to perceive and I think is what Griesinger writes about, it's the auditory system switching state basically. Griesinger calls this distance, where the transition happens, as Limit of Localization Distance. Nice thing with this is, there is no uncertainty with it, it's either or basically and always present in all environments and sound sources as it's your auditory system that is responsible for it, so a lot of logic can be drawn from it.

For example, no matter how small your listening distance, and still not hearing it, then the loudspeakers aren't that great and something with them prevents it. Or, no matter how great your loudspeakers, but if you listen too far too loud room sound (early reflections) would prevent it happening. And so on. I think being able to hear it, finding the transition by ear, takes your listening skill to another level altogether, enables use of logic, confusion gone. You'd learn about yourself (perception), your recordings, your playback system and so on. In general, you can now utilize your own auditory system to your best. For example, some recordings sound weird with the pin point imaging, which to me indicates the engineer didn't have it while making the recording otherwise he would have vomitted doing that all day long. On the other hand, some recordings sound just phenomenal, either side of the transition, with or without pin point imaging. Nevertheless, you'd be always able to listen on either side, at will, just by moving yourself a little further or closer to speakers, utilize switch in your own auditory system. End of circle of confusion.
 
Last edited:
[...] If you are interested, look for David Griesinger studies about proximity and localization, basically most of his studies seem to touch the subject although mostly in concert hall and live music context, but it's the perception he studies, which we all carry with us no matter if it's concert hall or livingroom. [...]
Wow, what a fantastic suggestion. I just finished several of his papers - well written, and interesting! Here's one relevant excerpt from Griesinger, D., 1997. The psychoacoustics of apparent source width, spaciousness and envelopment in performance spaces. Acta Acustica united with Acustica, 83(4), pp.721-731:
[...] reflections in the 10ms to 50ms time range while listening for the apparent source width (ASW). We have done this experiment with several listeners. All report that reflections in this time range broaden the apparent width of the source,
I experimented with this myself, and can confirm the effect. So cool to actually experience it, rather than just reading and having to trust.

In short, I've found out that all the rooms I took my speakers into there is some particular size of stereo triangle beyond which this pin point phantom center doesn't happen anymore. But, I've always found it just by moving forward closer to speakers, effectively increasing direct/reflected sound ratio, until at some point the pin point imaging happens quite suddenly.
Hm... That's interesting, I don't hear it this way... ? As I leave the sweet spot, sources appear to grow slowly.

Thanks again for your reply! You/Mr. Griesinger/we found one factor: 10-50ms reflections. A 10ms reflection corresponds to just 3.4m (11ft). I suppose it might be easy to accidentally introduce those reflections when listening in a larger stereo triangle. On the positive side, this is predictable and measurable.


Will post some more if I find more answers.
 
I had a link to an article about panning (but now it's gone. :( ) The conclusion was that the mixing engineer shouldn't waste time trying to precisely locate all of the instruments/tracks because you'll lose the position if you move or if you listen in a different room.

I don't think we can "precisely locate" instruments & vocals with live music either (with our eyes closed) but it's been awhile since I've listened to live music, and closing my eyes I'd still remember where the instruments are
Don't forget that immersive audio systems (Dolby, DTS, etc) do precisely place sounds.
 
My most recent discovery is that imaging (sound sources distinctively appear between, or even outside, of speakers) strongly correlates with how correlated/mono pink noise sounds. When I listen from a position where pink noise sounds as if it emanates from a small point, everything sounds fantastic. When I'm listening from another position, I lose the ability to localize instruments. While the music is still stereo, the instruments are somehow elusive, instead of spatially localizable.

Questions: How can I increase the area where correlated pink noise seems clearly localized in the center? Which parameters affect my ability for sound localization? Why does correlated pink noise "widen" when I move a couple inches/centimeters forward or backward?

That large and spacious experience is one of my favorite attributes of well implemented stereo playback!

Sadly, I have never experienced any system/setup/technology that has a sweetspot wide enough to accommodate more than a single listener with the best experience. In my experience, systems that supposedly have a wide sweetspot actually have a very poor sweetspot (I.E. no sweetspot).

In hometheater sized rooms, Dolby Atmos is no exception; either you are in the sweetspot or you have a significantly degraded experience. Sure the sounds are anchored to a center channel, but the 'bubble of sound' experience is greatly degraded.
 
Wow, what a fantastic suggestion. I just finished several of his papers - well written, and interesting! Here's one relevant excerpt from Griesinger, D., 1997. The psychoacoustics of apparent source width, spaciousness and envelopment in performance spaces. Acta Acustica united with Acustica, 83(4), pp.721-731:

I experimented with this myself, and can confirm the effect. So cool to actually experience it, rather than just reading and having to trust.


Hm... That's interesting, I don't hear it this way... ? As I leave the sweet spot, sources appear to grow slowly.

Thanks again for your reply! You/Mr. Griesinger/we found one factor: 10-50ms reflections. A 10ms reflection corresponds to just 3.4m (11ft). I suppose it might be easy to accidentally introduce those reflections when listening in a larger stereo triangle. On the positive side, this is predictable and measurable.


Will post some more if I find more answers.
You may also take a look at these David Griesinger's paper and presentation. The spatial information is all encoded in the recording, and therefore how it is recorded is extremely important. Excerpt from the paper.
griesinger.png

 
Sadly, I have never experienced any system/setup/technology that has a sweetspot wide enough to accommodate more than a single listener with the best experience. In my experience, systems that supposedly have a wide sweetspot actually have a very poor sweetspot (I.E. no sweetspot).
You are welcome to come visit. ;)

Big speakers and for room treatment - diffusion as well as selective absorption are my secret sauce.
 
You are welcome to come visit. ;)

Big speakers and for room treatment - diffusion as well as selective absorption are my secret sauce.

I would take you up on that if I was closer!

Without any additional context, what you are describing generally sounds like the same "sauce" I'm enjoying.
 
You could even out the SPL differences between left and right with clever horizontal dispersion characteristics of the speakers, but the time domain difference has no solution.
the only way to make these differences smaller outside of the center line is to make the triangle bigger. In practice, the triangle/room would need to be gigantic to make a 3-person couch even.
 
For a more intense sweet spot, assuming you are already equidistant from each speaker, reduce reflected sound.

Here is an overlay of unsmoothed frequency response for beamy low reflections electrostats (black) vs wide dispersion JBL (red)

I interpret the dips as cancellations due to reflected sound, so many clues to the imaging may be missing.

index.php


That one has a lot of EQ applied.

Here it is without "room correction".

index.php
 
Back
Top Bottom