Omni-directional bass - when does it start and why?

[Cloth Ears]

Why does bass become omni-directional? The oft-quoted point is 80Hz, but why would that be? And is it true everywhere?

 

It is due to the distance between your ears (>30cm, but probably <40cm) relative to the sound you're trying to localise? For me, the distance between my ears would be about 7% of the length of an 80Hz wave. Would that be a suitable yardstick?

 

Or is it due to the size of the room that you are in? Where the shortest dimension of your room determines when the localisation of your bass becomes difficult? So, say, in a 7 foot ceiling room, that half a wavelength of an 80Hz signal - and with a 9 foot ceiling, that becomes 65Hz? Or is it more complex than that?

 

And is there an absolute limit to where you cannot determine where something is coming from? One example from freespace is thunder, as after the initial clap, the rumbles (40-100Hz) are easily followed even if the eye did not see the bolt of lightening. This is obvious when you live in the Top-end, as the lightening is mostly sheet lightening, from cloud to cloud, and you don't always see it. But you can definitely follow it by ear. So it seems that it's easier to localise lower frequesncies outside, rather than inside.

 

Are there any technical articles on the subject?

 

Hugh Covill trying to explain treating subwoofers as an array Demystifying Directional Bass from Audio Technology (Hugh Covill)

An article on the Kii Three and that (but not how) it attempts  to make the speaker more directional at low frequencies by cancelling out some of the almost omnidirectional bass that comes from the main driver Does hi-fi end here? from The Rational Audiophile

Frequency Analysis of Thunder Features

 

Edited April 28, 2021 by Cloth Ears

[Godot]

“the Occasional Podcast” had an episode posted 6/4/21, “Subwoofers & 2 channel audio: everything you ever wanted to know “, a deep dive with SVS Director of Technology Ed Mullen. 
It wasn’t actually too technical but did have some interesting points (for me anyway). He did cover the questions in your post, and much else. I found it worth a listen. 

[davewantsmoore]

13 minutes ago, Cloth Ears said:

Why does bass become omni-directional?

 

Direction is figure out using differences in phase and level in each ear.

 

At high frequencies is is mainly level differences (not phase) .... part of this reason is that the head is too big.

 

Below about 1000 hz ... the head is small enough not to mess with the phase .... and so phase starts to dominate.

 

Below about 300 hz .... level differences are almost gone.... any sound will start to sound the same loudness in both ears.

 

Below about 80 hz .... phase differences are almost gone .... a sound arriving from (for example) hard left, has approximately the same phase in both ears.... and all ability to tell the direction is gone.

 

 

13 minutes ago, Cloth Ears said:

And is it true everywhere?

 

Yes.... but it starts to get complicated with headphones.... where level and phase differences can be achieved, at LF.

 

... but two things happen.    Your brain is not used to processing it (and so doesn't "hear" it as much as maybe expected).... and conduction of sound through the head is very significant.

 

13 minutes ago, Cloth Ears said:

80Hz signal

 

80Hz is often chosen because if the is the "crossover point" with a steep-sh filter, then the level is suitably low by ~200 and definitely ~300 Hz to be swamped by sources which are "in the right place".

 

13 minutes ago, Cloth Ears said:

And is there an absolute limit to where you cannot determine where something is coming from?

 

It can be difficult to do these experiments because is requires low distortion and "point sources"...... but error in the experiment tell us we can "hear bass to lower frequencies" than we actually can.

 

So it's somewhere around 200hz .... but most people don't believe that because when they try it that can "hear stereo" a lot lower .... but they're not testing it right  (distortion, rattles, etc.)

 

 

 

13 minutes ago, Cloth Ears said:

One example from freespace is thunder

 

It simple... thunder is easy to localise as it contains high frequencies.

 

 

 

13 minutes ago, Cloth Ears said:

So it seems that it's easier to localise lower frequesncies outside, rather than inside.

No.

 

Yes... Rooms if anything make it worse, as the room is relatively small compared to the sound wavelength.... but the effect still exists in full force outside.

 

13 minutes ago, Cloth Ears said:

An article on the Kii Three and that (but not how) it attempts  to make the speaker more directional at low frequencies

 

At low frequencies, the kii three operates all of it's woofers in phase.

 

At middle frequencies (eg. 100 to 800) it uses the side and back to cancel the sound, and increase directionality.

 

 

There is a LOT of merit to narrower coverage (increase directionality) in the range between ~200 and 1000..... it offers a HUGE improvement in clarity..... but avoids the "in your head" effect from headphones.

 

IMVHO it is the fundamental reason why speakers like horns, dipoles incl kii three (can) sound so much better.    (I say "can" as not all are well designed).

 

 

[RoHo]

So, Dave, to use a stupid example to illustrate a serious point, an alien with a metre wide head would be able to localise sound to a lower frequency than us simple humans?  Given that he has similar hearing instruments to us humans and not a thousand sucker-like nodules oozing green slime?

[Cloth Ears]

9 minutes ago, davewantsmoore said:

 

So it's somewhere around 200hz .... but most people don't believe that because when they try it that can "hear stereo" a lot lower .... but they're not testing it right  (distortion, rattles, etc.)
 

Dave, you keep saying that. But that means that every subwoofer I've ever heard has distortion and rattles enabling it to be located (if not pinpointed). Relative to the wavelength of the sound, you should be able to localise down to 10% (if not 5%). I mean, there are people on this site who can hear the difference between frequencies that are 1/64 tone different (that Music Lab test), so surely L/R at some lower frequencies should be possible.

 

But I've yet to see an article that explains it to me.

[Deepthought]

This paper certainly seems to suggest that people can perceive left to right movement of low freqs (30-60hz) in reverberant rooms - https://www.acoustics.asn.au/journal/2014/Vol42No2_DEAN.pdf

 

[davewantsmoore]

53 minutes ago, Cloth Ears said:

Dave, you keep saying that

Don't shoot the messenger?!   

 

53 minutes ago, Cloth Ears said:

that means that every subwoofer I've ever heard has distortion and rattles enabling it to be located

Shrug.

 

Many subwoofers produce high amounts of non-linear distortion when played at realistic levels.

 

53 minutes ago, Cloth Ears said:

Relative to the wavelength of the sound, you should be able to localise down to 10% (if not 5%).

 

As an example.... At 200Hz, the distance from ear to ear is extremely approximately 10% of one wavelength.... and there is basically zero level difference.

 

... and because when you move, or turn your head, it doesn't change.... your listening system has evolved to not consider it "information".

 

53 minutes ago, Cloth Ears said:

But I've yet to see an article that explains it to me.

 

https://en.wikipedia.org/wiki/Sound_localization

[davewantsmoore]

1 hour ago, deepthought said:

This paper certainly seems to suggest that people can perceive left to right movement of low freqs (30-60hz) in reverberant rooms - https://www.acoustics.asn.au/journal/2014/Vol42No2_DEAN.pdf

 

 

 

Quote

It is important to bear in mind (as detailed in the Appendix), that while the digital signal leaving the MAXMSP synthesis in each case contained very little energy at frequencies above 200 Hz, that generated by the subwoofers contained some energy detectable above the acoustic background of the studio, in some cases at frequencies up to about 600 Hz. Correspondingly, sending high dB sine tones to the speakers produced pitch-discernible audible sound at least up to 1000 Hz. Thus even ‘clean’ low frequency sounds as presented by these excellent speakers will always contain higher frequencies. The same observations held for larger (much more expensive!) Meyer subwoofers

 

 

Aside from the above (which doesn't necessarily explain what they observed) ....  it's important to undcerstand that the paper is not talking about "being able to hear where the subwoofer is"

 

.....  but about the creation of a spatial effect using multiple subwoofer, each producing different content... ie. being able to hear a different, or even "moving" spatiality (is that even a word, lol) to the sound, based on multiple tones.

 

I touched on this in the other thread (erplying to AndyR).    Yes, it can be possible to encode stereo effect into LF sound like this "articifically" (which isn't a negative connotation, just noting that you wouldn't get this effect from any typical "real" multi-mike recording).    If you play it back in your room, you won't necessarily get the same effect heard by the person creating it.... which is a big part of why it isn't typically done.

 

[Deepthought]

"I conclude that musically-experienced listeners have good capability in relation to location/lateralization and movement perception of low frequency sounds in our reverberant studio environment, though movement accuracy is much lower than location accuracy, as normally observed."

So his conclusions are not correct?

[Cloth Ears]

On 29/04/2021 at 11:19 AM, davewantsmoore said:

Aside from the above (which doesn't necessarily explain what they observed) ....  it's important to undcerstand that the paper is not talking about "being able to hear where the subwoofer is"

 

.....  but about the creation of a spatial effect using multiple subwoofer, each producing different content... ie. being able to hear a different, or even "moving" spatiality (is that even a word, lol) to the sound, based on multiple tones.

 

I touched on this in the other thread (erplying to AndyR).    Yes, it can be possible to encode stereo effect into LF sound like this "articifically" (which isn't a negative connotation, just noting that you wouldn't get this effect from any typical "real" multi-mike recording).    If you play it back in your room, you won't necessarily get the same effect heard by the person creating it.... which is a big part of why it isn't typically done.

 

OK - in that case, I think we might be in agreement (say it ain't so...LOL). Just created a test set of tracks, with ~32-37Hz signals in both left and/or right channels. And also with "in-phase" and "out-of-phase" frequencies - by "phase", I mean that the peaks of the sine waves were either in-sync, or exactly out of sync (peaks to troughs). Is there a name for that?

 

And, with only the bass amps on, I couldn't locate the sound with my ears as to whether it was left or right. Actually, I could, but only with practice of listening to the tracks more than once and then being able to know it's coming from one speaker or the other. Pretty obvious when the signal was both channels, as it would be slightly louder (if "in-phase") or weirdly lobing (if "out-of-phase").

 

Left and right was different. I could tell when the signal was coming from one or other channel, just not which one. And, I expect, I would have the same issue with the bottom end of my main speakers if I used a 60Hz signal. As my bass modules are directly below the main speakers, they should be time aligning properly. But I also think that if the signal was in mono, rather than stereo, I would be getting a different sonic picture than I am getting from having the stereo signal being sent to the correct speakers.

[davewantsmoore]

1 hour ago, Cloth Ears said:

OK - in that case, I think we might be in agreement (say it ain't so...LOL)

 

1 hour ago, Cloth Ears said:

Left and right was different. I could tell when the signal was coming from one or other channel, just not which one.

Yes!    If you measure your channels carefully.... they will not be the same.   They sound different.    Your brain can hear different.... and it tries to tell you information because of the difference.

 

1 hour ago, Cloth Ears said:

they should be time aligning properly

Time alignment of the bass, is only relevant as far as it relates to summed levels.....  You should make sure the overall levels of everything are the same/correct .... as difference in level is very audible.... (See above)

 

1 hour ago, Cloth Ears said:

And, with only the bass amps on, I couldn't locate the sound with my ears as to whether it was left or right. Actually, I could, but only with practice of listening to the tracks more than once and then being able to know it's coming from one speaker or the other. Pretty obvious when the signal was both channels, as it would be slightly louder (if "in-phase") or weirdly lobing (if "out-of-phase").

 

Left and right was different. I could tell when the signal was coming from one or other channel, just not which one.

 

Being able to tell them apart.... isn't the same as being able to actually hear the location of the woofer.   Your brain is more sensitive to chages of level in the bass.... or any difference between the two is very obvious.... and leads to "I can hear something".

 

If we do identical bass tones, we can't hear where thy are coming from.

 

If we do different bass tones (ie. stereo bass) .... then we can hear something.... but this "something" doesn't rely on the location of the subs.   The can be front/back.... left/back... wherever .... and we get the same(similar) effect.    The effect will be different for every room/setup.... which is why it isn't done  (and why it is standard mixing practise to make bass mono).... and why when artists try to make such bass effects, they usually stipulate they should be listen to on headphones, as hen the bass effect is the same for everyone.

 

For the rare occasion there is stereo bass in a recording, it is usually either:

 

Panned full range signal..... which doesn't provide such "bass effect" .... and so you can't hear anything (it just sounds mono) ... and you could send the bass to every subwoofer location, and it would sound the same.

 

Multi-mike "stero/milti" recording.... eg. mike stage left... mike stage right.....  but unless there is big separation, and no room (or huge room) the effect is not what you'd expect...... and then... you can't play it back (the same as recorded) in a small room, as the bass just mixes together..... which is why it isn't (or rarely) done (bass gets mixed mono).

 

[Cloth Ears]

2 hours ago, davewantsmoore said:

For the rare occasion there is stereo bass in a recording, it is usually either:

 

Panned full range signal..... which doesn't provide such "bass effect" .... and so you can't hear anything (it just sounds mono) ... and you could send the bass to every subwoofer location, and it would sound the same.

 

Multi-mike "stero/milti" recording.... eg. mike stage left... mike stage right.....  but unless there is big separation, and no room (or huge room) the effect is not what you'd expect...... and then... you can't play it back (the same as recorded) in a small room, as the bass just mixes together..... which is why it isn't (or rarely) done (bass gets mixed mono).

 

Except when you get recordings done using a crossed pair of mikes (like Clark Terry Live at the Village Gate). Now, in that one you can see (with the right tools) that the bass isn't mono'd, but also that there also aren't any fundamentals down low enough to really matter (in the context of this discussion). In other words, you can always tell that the bass is slightly to the left of centre stage and the drums are to the right of that.

 

Or maybe I've misunderstood you.

[davewantsmoore]

2 hours ago, Cloth Ears said:

Except when you get recordings done using a crossed pair of mikes (like Clark Terry Live at the Village Gate). Now, in that one you can see (with the right tools) that the bass isn't mono'd

 

How far apart are the microphones?    If they're not (quite) far apart then the bass entering each mike is practically the same level and phase  (so you can't record stereo bass).

 

... but it depends on how far, and what frequency.

 

2 hours ago, Cloth Ears said:

but also that there also aren't any fundamentals down low enough to really matter (in the context of this discussion). In other words, you can always tell that the bass is slightly to the left of centre stage and the drums are to the right of that.

 

Bass and drums both have lots of high frequencies (even in their "low frequency bits".... as they are short sounds, with wide bandwidths).   So unless you are filtering out everything above 80hz, then of course you can tell their direction.

[Cloth Ears]

40 minutes ago, davewantsmoore said:

 

How far apart are the microphones?    If they're not (quite) far apart then the bass entering each mike is practically the same level and phase  (so you can't record stereo bass).

 

... but it depends on how far, and what frequency.

 

As far as I'm aware, using a coincident pair of directional mikes for stereo recording needs them to be at 90 degrees (or greater for some types) and as close as possible without touching.

Edited May 7, 2021 by Cloth Ears

[davewantsmoore]

1 hour ago, Cloth Ears said:

and as close as possible without touching.

Right... so it cannot record stereo bass.... because it is not directional at low frequencies, and they are too close together.

[almikel]

5 hours ago, Cloth Ears said:

As far as I'm aware, using a coincident pair of directional mikes for stereo recording needs them to be at 90 degrees (or greater for some types) and as close as possible without touching.

Are you referring to the classic "Blumlein pair" recording mike config? 2 recording mikes co-located at 90 degrees?

 

Unfortunately not used much these days - most recordings are close miked...

...with a recording done on on a Blumlein pair you get a real sense of the space the recording was done in with a good playback system...

 

...but what @davewantsmoore is saying is correct - there's no stereo bass in anything captured by a Blumlein pair - the mikes are too close together compared to the long wavelengths of bass.

 

cheers,

Mike

[Cloth Ears]

10 hours ago, davewantsmoore said:

Right... so it cannot record stereo bass.... because it is not directional at low frequencies, and they are too close together.

 

6 hours ago, almikel said:

Are you referring to the classic "Blumlein pair" recording mike config? 2 recording mikes co-located at 90 degrees?

 

Unfortunately not used much these days - most recordings are close miked...

...with a recording done on on a Blumlein pair you get a real sense of the space the recording was done in with a good playback system...

 

...but what @davewantsmoore is saying is correct - there's no stereo bass in anything captured by a Blumlein pair - the mikes are too close together compared to the long wavelengths of bass.

 

cheers,

Mike

Yet, in this recording you can tell where everyone is on stage, including the double bass, the pieces of the drum kit and the way the piano is set-up. And you can see (looking at the digital stuff) that the bass notes of the drums, DB and piano are not in the 'centre' (i.e. not a mono signal), they are in the same 'location' as the attacks on the various instruments.

 

So, what has been recorded?

[sir sanders zingmore]

2 hours ago, Cloth Ears said:

 

Yet, in this recording you can tell where everyone is on stage, including the double bass, the pieces of the drum kit and the way the piano is set-up. And you can see (looking at the digital stuff) that the bass notes of the drums, DB and piano are not in the 'centre' (i.e. not a mono signal), they are in the same 'location' as the attacks on the various instruments.

 

So, what has been recorded?

I suspect your ears are getting the location information from the “attack” part of the sound and your brain is putting it together with the low part. 
If you applied a low pass filter my guess is that you’d lose the ability to locate those instruments. 

[Satanica]

Sorry if I missed it, but typically at what frequency is the transition from mono bass to stereo? 

[Pigpen]

I remember speaking to an audiologist during a medical check re this topic and she explained that there is "some evidence" to suggest that "around" 300Hz directionality is decreased /lost for those who do not speak a tonal language. For those who speak a tonal language, it is around 200Hz. 

 

My wife's native language is tonal (and her ears actually work) so if she tells me something sounds good, I nod wisely and agree - I use her to set up my system (particularly handy getting her to lift the big amplifiers......) and she has always got things pretty much right.

[Satanica]

1 hour ago, Pigpen said:

I remember speaking to an audiologist during a medical check re this topic and she explained that there is "some evidence" to suggest that "around" 300Hz directionality is decreased /lost for those who do not speak a tonal language. For those who speak a tonal language, it is around 200Hz. 

 

My wife's native language is tonal (and her ears actually work) so if she tells me something sounds good, I nod wisely and agree - I use her to set up my system (particularly handy getting her to lift the big amplifiers......) and she has always got things pretty much right.

 

Interesting, but my question was about what is contained within a typical recording. 

[Pigpen]

21 minutes ago, Satanica said:

 

Interesting, but my question was about what is contained within a typical recording. 

Are stereo recordings devoid of 200 - 300Hz?

[davewantsmoore]

7 hours ago, sir sanders zingmore said:

I suspect your ears are getting the location information from the “attack” part of the sound

If there is an "attack" (although what is "an attack" could be subjective, so it's hard to do this via the internet) then this means it contains high frequencies..... and so the low pass filter is some how insufficient.

[davewantsmoore]

10 hours ago, Cloth Ears said:

 

Yet, in this recording you can tell where everyone is on stage, including the double bass, the pieces of the drum kit and the way the piano is set-up. And you can see (looking at the digital stuff) that the bass notes of the drums, DB and piano are not in the 'centre' (i.e. not a mono signal), they are in the same 'location' as the attacks on the various instruments.

 

Becauae all of thse instruments contain lots of high frequencies.... a bang on a bass drum contains lots of high frequencies.

[Satanica]

55 minutes ago, Pigpen said:

Are stereo recordings devoid of 200 - 300Hz?

 

That's my question, at what frequency does the typical average recording stop being stereo and become mono?