Bass direction is audible

[restorer-john]

There's a long thread about this here somewhere, couldn't find it right now. I think the general answer (disregarding very special situations and/or golden eared individuals) is that no, this is not a real world problem.

But you are somewhat biased aren't you, being a manufacturer and seller of sub/satellite type systems...

 

[sigbergaudio]

But you are somewhat biased aren't you, being a manufacturer and seller of sub/satellite type systems...

Yes, and/or informed. We tested this quite extensively before selecting the crossover for our systems, and obviously wanted it to work as well as possible. And stereo perspective / imaging / immersiveness are all qualities that are important design goals. So if that was ruined unless you had a very low crossover, I wouldn't have designed it the way I did.

But anyone is free to make their own informed decision based on available data and own experiences of course.

My experience is that this is one of many areas where people are affected by bias and expectation, assuming 80hz is the absolute maximum viable crossover due to that being the standard recommendation, and since everyone wants to have the very best, they go even lower to be sure, sacrificing dynamic range along the way.

 

[Duke]

There undoubtedly IS something anomalous going on if you need multiple subs to get a satisfactory result. You're addressing it with multiple subs. Fine. There are other ways.

I should have made it clear that my priority is getting good bass over a large listening area.

I concede that a single sub + EQ can give satisfactory results in one listening location, but this is often at the expense of making the frequency response worse elsewhere in the room (because the room-interaction peak-and-dip pattern will be different at other locations). Whether or not this matters depends on one's priorities.

Ime getting satisfactory results throughout a large listening area calls for multiple subs intelligently distributed, as this reduces the spatial variance in frequency response. So any remaining frequency response anomalies are likely to be global (throughout-the-room) rather than local issues, and therefore the use of EQ becomes more globally effective, rather than improving the frequency response in the microphone location at the expense of degrading it elsewhere. (Imo a double bass array is a special case of "multiple subs intelligently distributed").

My experience is that this is one of many areas where people are affected by bias and expectation, assuming 80hz is the absolute maximum viable crossover due to that being the standard recommendation, and since everyone wants to have the very best, they go even lower to be sure, sacrificing dynamic range along the way.

My understanding is that the 80 Hz recommendation arose from home-theater-centric studies which assumed a single subwoofer would be used, which would often be placed in a corner of the room, or maybe even off to one side.

I have distributed multi-sub customers who use crossover frequencies an octave or more north of the 80 Hz "upper limit", in situations where they have about two octaves of overlap from which to choose the best-sounding crossover frequency. These are customers who use CLIO and/or REW. I would have told them not to go north of 80 Hz but they didn't ask for my advice, and now I'm glad they didn't because I am learning from their boundary-pushing experiences.

 

[Sokel]

Yes, and/or informed. We tested this quite extensively before selecting the crossover for our systems,over due to that being the standard recommendation, and since everyone wants to have the very best, they go even lower to be sure, sacrificing dynamic range along the way.

The suspicious one would think that the selected x-over (about 100Hz I guess as I had a look at the satellites) would make nice charts without the usual dips lower x-overred system have.
That's also the main reason we see so many people going up there just by looking at their charts in REW,etc.
And that's fine in a tight configuration with subs under mains for example.

But further apart stuff start getting a lot more complicated,some music is just unlistenable,specially if room is not helping either.

 

[NTK]

I was sent this article: Auditory Localisation of Low Frequency Sound Sources (Nastassa, Pulki, Makivirta, AES Convention 154 2023). Unfortunately I only have access to the abstract. I have quoted it in full:



Note that the study was done in an anechoic chamber. The lowest frequency of 31.5Hz is about 10m long, surely this can not be audible in an average listening room?

If this study is correct, it would have implications as to how we should position subwoofers in our listening rooms. Maybe Griesinger's "stereo sub" idea may have some merit. Subwoofer positioning schemes such as Geddes' 3 sub method, or two subs in diagonal corners would need rethinking. Opinions?

The paper is available for download at Aalto University's site. They often provide free access to their papers.

Auditory localisation of low-frequency sound sources

research.aalto.fi

 

[sigbergaudio]

The suspicious one would think that the selected x-over (about 100Hz I guess as I had a look at the satellites) would make nice charts without the usual dips lower x-overred system have.
That's also the main reason we see so many people going up there just by looking at their charts in REW,etc.
And that's fine in a tight configuration with subs under mains for example.

But further apart stuff start getting a lot more complicated,some music is just unlistenable,specially if room is not helping either.

The speakers have a -6dB point at 90hz, while the subs have the same at 100hz (so slight overlap). The crossover is selected to maximize dynamic range while still avoiding localization issues. This works well even with a single sub that is not in a "tight configuration". It does not at all make any music unlistenable.

 

[Sokel]

The speakers have a -6dB point at 90hz, while the subs have the same at 100hz (so slight overlap). The crossover is selected to maximize dynamic range while still avoiding localization issues. This works well even with a single sub that is not in a "tight configuration". It does not at all make any music unlistenable.

OK,unlistenable is a heavy word,I probably have a urge to drama I admit.
And I didn't mean it that speakers of their own is like that but rather the combination,"difficult" music - speaker - difficult room - not very careful setup,etc.

There are a lot of things that can go wrong,and they usually are.The fun part is that there is a lot of room to play with.
On the other hand dynamic range is also a problem with small setups,specially mid-bass.
If I was in that dilemma I honestly don't know what I would choose.

 

[Weeb Labs]

It has been my experience that closely spaced modes in small rooms can make low frequencies feel directional; especially if the response differs significantly between the ears. This is not the same as localisation but I would imagine that it has the potential to be mistaken for it.

 

[Bjorn]

I experienced this same conclusion. I tried several location arrangements and what sounded best and surprisingly measured best at the MLP was front stage L&R subs just outside the the Main L&R front speakers. I suspect that this is very room dependent and my living space is all open floor plan that includes Living Room/Dining/Kitchen/Den and main hallway. I had a greater sense of Tactile energy and chest bass sensation. I lost that feeling when the subs were moved to other perimeter walls. Of course add in a sensible amount of listener & visual bias. The subs are pretty big @ 41″ x 20.5″ x 22.5″ (DxWxH) making them all but Impossible to hide.

Especially with subs in front and in rear, tactile sense, transient, kick and slam seems to greatly suffer.

At a few occasions I have turned front subwoofers the opposite direction of listening position to achieve a more even response. But it leads to similar issues, at least when crossover is high enough.

In most rooms it's possible to achieve a very linear response with one subwoofer if we simply move out of the room modes. But it shouldn't be a Micky mouse subwoofer that can't deliver the SPL nor low modulation distortion. Sealed subwoofers with 10" and 12" drivers is pretty much a joke to me.

 

[OCA]

Pink noise by definition is a mixed bag of all frequencies so it cannot be at a single frequency of 31.5Hz! Having been born and raised in Istanbul, I am quite familiar with Earthquakes and that's about the deepest and longest bass sound most people will ever hear. I can confirm, you can not detect its direction, it's like everywhere at once although the epicenter is always at a certain direction from you.

 

[terryforsythe]

The ideal subwoofer crossover frequency is heavily dependent on the speakers, the subwoofer(s), speaker and subwoofer locations, and the room.

I have ported bookshelf speakers. I ran into various issues trying to cross them over to my subwoofer:

1) Initially I could not get good phase alignment between the speakers and the subwoofer due to the speakers' widely varying group delay around the port tuning frequency.
2) The harmonic distortion of the bookshelf speakers significantly increases below 100 Hz.
3) Below 100 Hz the bass of the bookshelf speakers sounds sloppy, not tight.

I played with a lot of different crossover frequencies and slopes, ran too many frequency response sweeps to try to count, and spent countless hours listening to different configurations. What finally worked best for my system was plugging the speaker ports and crossing over at 100 Hz using 48 dB/octave slopes. With the subwoofer located at the front of the room, I cannot audibly detect from where the sub frequencies are coming; to me and anyone else who has heard my system, it all sounds like they are in the sound stage provided by the speakers.

 

[Sokel]

View attachment 374764

Pink noise by definition is a mixed bag of all frequencies so it cannot be at a single frequency of 31.5Hz! Having been born and raised in Istanbul, I am quite familiar with Earthquakes and that's about the deepest and longest bass sound most people will ever hear. I can confirm, you can not detect its direction, it's like everywhere at once although the epicenter is always at a certain direction from you.

They probably talk about band-limited pink noise,the same described by Klippel,etc.

 

[gnarly]

They probably talk about band-limited pink noise,the same described by Klippel,etc.

That's how I read it too.....that pink bursts were low-passed, and any low-pass above 31.5Hz, gave directional recognition.
I am left wondering how long the pink-bursts were.....
I've been using short sine bursts, a few cycles for localizations tests. REW makes it easy.

 

[dualazmak]

I tried several location arrangements and what sounded best and surprisingly measured best at the MLP was front stage L&R subs just outside the the Main L&R front speakers.

Interestingly, in my case, what sounded best and surprisingly measured best at the MLP was front stage L&R subs just inside! the the Main L&R front speakers.
You can find my latest system setup here, if you would be interested. My heavy-large L&R subwoofers, YAMAHA YST-SW1000, mainly cover 16 Hz to 55 Hz.

Spoiler

 

[bmc0]

The paper is available for download at Aalto University's site.

Interestingly, the PDF contains the wrong paper: Impact of standing waves on human auditory perception of low-frequency direction. Still relevant though (perhaps more so).

Anecdotally: In my listening room, I think I can hear spatial effects down to around 40Hz with sharply low-passed (72dB/octave) stereo pink noise or beat signals (sum and difference of sines 2-4Hz apart). Additionally, I've found quite a few recordings where stereo reproduction in the subwoofer range (below 85Hz in my setup) seems to result in a different spatial impression than summed mono. These were simple sighted tests, however, so hardly definitive.

 

[Chrispy]

Strange thread title for a limited application. I'd not consider anechoic all that important in the application of lower frequencies in-room....

 

[JSmith]

Interestingly, the PDF contains the wrong paper

I'd be happy to post the whole paper here... but I think it breaches copyright. I do think it's about time AES makes more papers freely available... especially older ones.


JSmith

 

[Justdafactsmaam]

View attachment 374764

Pink noise by definition is a mixed bag of all frequencies so it cannot be at a single frequency of 31.5Hz! Having been born and raised in Istanbul, I am quite familiar with Earthquakes and that's about the deepest and longest bass sound most people will ever hear. I can confirm, you can not detect its direction, it's like everywhere at once although the epicenter is always at a certain direction from you.

Yeah but an earthquake literally is all around you. Doesn’t matter where the epicenter was. The ground is undulating all around you.

 

[krabapple]

How is that conclusion, that it hardly seems relevant to home audio applications not jumping to a conclusion?

Iit's a conclusion arrived at by "reading the paper carefully, paying attention to the test conditions, then considering how the results may or may not be applicable to other situations."

How is that 'jumping'?

 

[olds1959special]

Putting my sub in between my left and right speakers, facing forward, sounds best to me. I tried facing sideways and placing it under the left or right side, and it never sounded right.