Science on low frequency quality?

[youngho]

Some certainly claim to be, but I have no direct proof either way.

Me, neither, but I'm open to the possibility...

Higher room modes not handled by the main speakers is one of my main arguments for sometimes using a higher crossover (the other being speakers too limited in LF response). The problem is that the subs definitely get localized IME with high (>100 Hz) crossovers, so you have to use either stereo subs or place one on the centerline to prevent (reduce) that problem, and that can compromise smooth response.

Yes...that would be interesting to me--why could you localize a 10 ft wavelength but not a 14 ft one?

Young-Ho

 

[anotherhobby]

The argument about locating a bass or other instrument (or tone generator) is a common one but misleading. A bass produces a range of signals far above the fundamental, well into the kHz region for the bow or plucking fingers on the strings, so we can localize based upon that. Similar for other instruments.

This is also a good argument for why stereo subwoofers isn't really necessary, almost everything have harmonics high enough that your speakers will reproduce them, thus giving a stereo image despite running mono subs. I typically run dual subs in mono, and I have atmosphere / perception of room / spaciousness in heaps.

Localization is interesting to me. I can localize a single sub with an 80 Hz crossover in both of my listening environments depending on the content. A jazz recording with an upright bass that isn't centered is usually the easiset, because if it's opposite the sub I'll hear the bass "bend" toward the sub. For the longest time I just assumed like everbody else in this hobby that maybe I had more "golden ears" or whatever that was more sensitive to LF. Recently I came across a slick localization test here, and it explains the whole concept nicely with demo sweeps, and the end result convinced me. I get that it's not the LF that I'm localizing, it's the harmonics, including those that are above 80 Hz but still coming thru the crossover at a reduce level.

That said, I am currently running stereo subs in one of my setups, and about a week after running the above test I was listeing to a friends album with a lot of LF content where I immediately noticed the deep bass was all mixed in stereo. I could tell because with the stereo subs in my small office you can feel the left sub pressurize the left ear, and the right sub pressurize the right ear, especially at high volumes. No, not a chance I can localize them, but they absolutely are easily identified as R and L due to the pressure. It's not exaclty directionality or localization, but it's also clearly not "bass is coming from everywhere." I actually don't like the effect and will be redoing them as mono subs in the coming weeks.

 

[DonH56]

Yes...that would be interesting to me--why could you localize a 10 ft wavelength but not a 14 ft one?

What I have read and quasi-remember is that it relates to the difference in phase we can resolve, so in a sense the wavelength is a "symptom" but not the "cause".

Localization is interesting to me. I can localize a single sub with an 80 Hz crossover in both of my listening environments depending on the content. A jazz recording with an upright bass that isn't centered is usually the easiset, because if it's opposite the sub I'll hear the bass "bend" toward the sub. For the longest time I just assumed like everbody else in this hobby that maybe I had more "golden ears" or whatever that was more sensitive to LF. Recently I came across a slick localization test here, and it explains the whole concept nicely with demo sweeps, and the end result convinced me. I get that it's not the LF that I'm localizing, it's the harmonics, including those that are above 80 Hz but still coming thru the crossover at a reduce level.

That was my conclusion, but there are those with differing opinions.

That said, I am currently running stereo subs in one of my setups, and about a week after running the above test I was listeing to a friends album with a lot of LF content where I immediately noticed the deep bass was all mixed in stereo. I could tell because with the stereo subs in my small office you can feel the left sub pressurize the left ear, and the right sub pressurize the right ear, especially at high volumes. No, not a chance I can localize them, but they absolutely are easily identified as R and L due to the pressure. It's not exaclty directionality or localization, but it's also clearly not "bass is coming from everywhere." I actually don't like the effect and will be redoing them as mono subs in the coming weeks.

Interaural time difference, ITD, is the accepted reason for localization of lower frequencies (<1 kHz or whatever). Interaural intensity difference, IID, is accepted as the primary way we localize higher frequencies. What "some" state, and I was unable to prove either way in the past, was how a sudden pressure wave might influence our ability to localize a LF source. Again, not my field, so I do not know if there are papers on that. Be nice to know...

My experiments were about 40 years ago, and my career took a different direction, so I am not an expert by any means.

 

[anotherhobby]

Interaural time difference, ITD, is the accepted reason for localization of lower frequencies (<1 kHz or whatever). Interaural intensity difference, IID, is accepted as the primary way we localize higher frequencies. What "some" state, and I was unable to prove either way in the past, was how a sudden pressure wave might influence our ability to localize a LF source. Again, not my field, so I do not know if there are papers on that. Be nice to know...

My experiments were about 40 years ago, and my career took a different direction, so I am not an expert by any means.

That's very interesting. Given the small size and layout of my office, you can see why I experience the effect of stereo subs pressurizing each side of the room. The subs can be seen on the floor under and slightly out from each main speaker.

 

[youngho]

What I have read and quasi-remember is that it relates to the difference in phase we can resolve, so in a sense the wavelength is a "symptom" but not the "cause".

Interaural time difference, ITD, is the accepted reason for localization of lower frequencies (<1 kHz or whatever). Interaural intensity difference, IID, is accepted as the primary way we localize higher frequencies. What "some" state, and I was unable to prove either way in the past, was how a sudden pressure wave might influence our ability to localize a LF source. Again, not my field, so I do not know if there are papers on that. Be nice to know...

You might be interested in reading about this model, which holds that low frequency localization in closed spaces depends on location of source, listener, and room characteristics like dimensions and reverberation time: https://www.researchgate.net/profil...n-as-a-function-of-closed-acoustic-spaces.pdf. Additional information and references can be found here: https://www.researchgate.net/public...ation_as_a_function_of_closed_acoustic_spaces

It appears that Holman was wrong about the country in the link I provided above, actually Finland: https://secure.aes.org/forum/pubs/conventions/?elib=11465. I found this reference and very interesting history of research to date including a mention of Welti ("Welti (2004) has set out to compare the audibility of two channel versus single channel bass reproduction in a small room environment (6.4 m x 7.3 m. x 2.7 m). Four different subwoofer configurations were tested, namely: one mono subwoofer placed at centre front; two summed mono subwoofers located at left and right front corners; two discrete left and right channel subwoofers located at the same position as the previous configuration; and two channel subwoofers located at +/- 90° relative20 to the listener. The stimuli used comprised of four short program loops and three music programs. The test was conducted using an ABX triangle method and in total, f ive trained listeners participated. The author concluded that the only noticeable difference occurred when comparing the centre mono subwoofer and the stereo configuration located to the left and right of the listener. Contrarily, no difference could be heard when comparing the front mono or stereo configurations. While Welti’s test partly dismissed that audible differences would be heard across the different subwoofer configurations, the author did suggest that this test does not prove in any way that there are no merits to stereo subwoofer reproduction") in this fascinating thesis, which addresses my question about low frequency localization in relatively free space: https://aaltodoc.aalto.fi/bitstream...tasa_Anamaria_2022.pdf?sequence=1&isAllowed=y

This thesis is really worth skimming! A lot of fascinating papers come out of Aalto University, which also has one of the most sophisticated auralization setups in the world. The original poster asked about science, so I hope the links that I've provided suggest that there may be more to bass quality than flat frequency response at the listening position, that bass <80 Hz may be difficult but not necessarily impossible to localize, and that different approaches might be considered to enhance spatial perception at low frequencies if desired.

Young-Ho

 

[jsilvela]

This thesis is really worth skimming! A lot of fascinating papers come out of Aalto University, which also has one of the most sophisticated auralization setups in the world. The original poster asked about science, so I hope the links that I've provided suggest that there may be more to bass quality than flat frequency response at the listening position, that bass <80 Hz may be difficult but not necessarily impossible to localize, and that different approaches might be considered to enhance spatial perception at low frequencies if desired.

Young-Ho

I've been following this discussion with a lot of interest, yes. Thanks!

In a previous post in this thread (https://audiosciencereview.com/foru...e-on-low-frequency-quality.41064/post-1452577)
was the comment that headphones are the "beginner package" for low frequency. Totally right.
That ties in with @anotherhobby 's comments about the feeling of different pressurization on the ears with two subs at the sides.
Headphones give you Left and Right low frequency. Wonder if the frequency for "LF becomes mono" is the same with headphones.

Speaking of localization in general, not just for LF, it's curious that with headphones we can get stereo imaging, but the sounds don't seem to come from "in front".
There was a comment on that in Griesinger's video, I think, or maybe his website.

 

[witwald]

... you might expend effort and money to "go with the science" and then find you don't like it.

And by the same token, you may expend a modicum of money and end up going "with the science" and find that you actually like it. The savings can be substantial, and the end results much better than would likely have been obtained "without the science". But, it's always been that way.

The same is always true of course, but at ASR you see routine advice for people to go for Genelec / Revel / Kef etc, without much caveating.

"Much caveating" is not a particularly precise term. There happens to be a lot of technical information on loudspeakers from those brands published here on ASR. The technical information itself is the caveat; to be ignored at one's peril.

 

[witwald]

For low frequency, I see a lot less clarity. For <100Hz, the room intervenes.

And what happens if a musician brings in their double bass and plays it in a domestic listening room scenario? Is the sound from the double bass somehow going to be immune to the room intervening? So how are we going to tune the room? Get two or three or more double bass players and distribute them at various points in the room?

 

[jsilvela]

And what happens if a musician brings in their double bass and plays it in a domestic listening room scenario? Is the sound from the double bass somehow going to be immune to the room intervening? So how are we going to tune the room? Get two or three or more double bass players and distribute them at various points in the room?

Not sure if this is a rhetoric question.
Assuming not ... I imagine small rooms are bad for live performances with significant LF component, for the same reasons they are for reproduction.
From what I've read and learned, I think bass traps would need to be huge for frequencies <100Hz, so are not really feasible, domestically.
So, "Get two or three or more double bass players and distribute them at various points in the room?"
Is that what is done? Or are small rooms altogether avoided for recitals?

 

[jsilvela]

And by the same token, you may expend a modicum of money and end up going "with the science" and find that you actually like it. The savings can be substantial, and the end results much better than would likely have been obtained "without the science". But, it's always been that way.

"Much caveating" is not a particularly precise term. There happens to be a lot of technical information on loudspeakers from those brands published here on ASR. The technical information itself is the caveat; to be ignored at one's peril.

That comparison of the advice at ASR for subs with the advice at ASR for bookshelf/floorstanding loudspeakers has gotten me in hot water.
I wish I had not used that as a discussion point.
But I think it's fair to say that low frequency reproduction *is* more complicated.

 

[jsilvela]

Going to make a single post here, so I can just refer to it next time I’m taken to task, which has been happening quite a bit:

With this thread I did not intend to disrespect the work of Welti / Geddes / Toole. I did not intend to disrespect the findings of ASR members who have been kind enough to share their setup, preferences, measurements and suggestions.

I also was not looking for an excuse to cheap out. Believe me I feel OK doing that, and don’t need anyone’s blessing. Nor was I intending to prevaricate nor waste anyone’s time.

I have recently bought my first sub, and noticed indeed that the response is very spatially uneven.
I find the Welti / Geddes / MSO / miniDSP route more complex than I like. I may end up going that route, but I still find it over-complicated.

I am enjoying the science, papers and discussions that have been shared here, quite a lot.
EOM

 

[bodhi]

I find the Welti / Geddes / MSO / miniDSP route more complex than I like. I may end up going that route, but I still find it over-complicated.

You are making it overcomplicated. You already know how to make measurements with REW. All you now need to do is get another sub, borrow one if you don't want risk buying one. Then just follow advice from MSO tutorial to the letter. Look at the predicted response and if you like it, buy a minidsp.

With the time you have used writing all these messages you would be done already.

 

[jsilvela]

With the time you have used writing all these messages you would be done already.

I've been trying and failing not to make this thread about "should I (jsilvela) buy XYZ".
FYI I placed the order on a second sub yesterday (not a REL, before I get further dunked on).

I think we all have things that appeal or don't appeal to us at the esthetic level.
For me, what I'm not liking about the single sub solution is it's finicky. So placement-dependent.
I'm a computer programmer. In that industry we'd talk of "robustness". Single sub -> non-robust.
I want something more robust and less fussy. The Welti / MSO approach feels complex and fussy to me.

You're right with the comment on the time spent, but ... only if you assume my goal was to save time.
In general, most of the members of ASR are not here to *optimize* for time.

 

[Willem]

I agree that MSO etc involves a fair bit of effort but will probably give the best result. A DSpeaker Antimode 8033 avoids most of the effort for a "pretty good" result.

 

[bodhi]

I've been trying and failing not to make this thread about "should I (jsilvela) buy XYZ".
FYI I placed the order on a second sub yesterday (not a REL, before I get further dunked on).

But that is all that is left to talk about. The theory is solid, the countless personal experiences all agree. Only thing that is uncertain is whether or not you deem the improvement worthwhile.

As you mentioned software development: one of the most important things I have learned during my career as an software architect is to avoid falling into analysis paralysis. It's often ten times more efficient to just try something out when the theory is solid enough. You get some result and if it works it's usually good enough. If it's not, you have more information to continue the speculations. This situation falls directly into that category.

 

[Sokel]

May I ask something silly about localization?
Is there a chance we feel the (lows) pressure wave the same way we fell the air and can tell it's direction?

(I hate to lower the scientific level of this discussion and I'm gratefull about the knowledge but I was always curious)

 

[jsilvela]

May I ask something silly about localization?
Is there a chance we feel the (lows) pressure wave the same way we fell the air and can tell it's direction?

(I hate to lower the scientific level of this discussion and I'm gratefull about the knowledge but I was always curious)

I don't think you're lowering the scientific level of this discussion.
I'm curious to hear about this question too. Have no idea myself.

 

[youngho]

May I ask something silly about localization?
Is there a chance we feel the (lows) pressure wave the same way we fell the air and can tell it's direction?

(I hate to lower the scientific level of this discussion and I'm gratefull about the knowledge but I was always curious)

Do you mean like with a ported subwoofer where you can potentially feel air moving through the port? Otherwise, sound is a wave or cyclical phenomenon, so air particles move back and forth a certain number of times per second, also known as frequency. If you have ever seen a woofer or bass driver moving back and forth, as in many YouTube videos of subwoofers, you get the idea. Because the air particles move forward and backwards, the net displacement should be zero, since they end up basically where they started, not including effects of Brownian motion, thermal convection, or air currents from ventilation or moving bodies in the room. Wind is a vector phenomenon with a direction and magnitude, so there is net displacement of air particles, meaning they end up further from where they started.

The difference in the perception of air particle velocity can be understood if you play a test tone corresponding to a room standing wave (mode) and stand in the null point (node), which is where air particle velocity is greatest but you perceive a suck out and no directionality. Now blow a fan at your face. Very different.

 

[DonH56]

You might be interested in reading about this model, which holds that low frequency localization in closed spaces depends on location of source, listener, and room characteristics like dimensions and reverberation time: https://www.researchgate.net/profil...n-as-a-function-of-closed-acoustic-spaces.pdf. Additional information and references can be found here: https://www.researchgate.net/public...ation_as_a_function_of_closed_acoustic_spaces

It appears that Holman was wrong about the country in the link I provided above, actually Finland: https://secure.aes.org/forum/pubs/conventions/?elib=11465. I found this reference and very interesting history of research to date including a mention of Welti ("Welti (2004) has set out to compare the audibility of two channel versus single channel bass reproduction in a small room environment (6.4 m x 7.3 m. x 2.7 m). Four different subwoofer configurations were tested, namely: one mono subwoofer placed at centre front; two summed mono subwoofers located at left and right front corners; two discrete left and right channel subwoofers located at the same position as the previous configuration; and two channel subwoofers located at +/- 90° relative20 to the listener. The stimuli used comprised of four short program loops and three music programs. The test was conducted using an ABX triangle method and in total, f ive trained listeners participated. The author concluded that the only noticeable difference occurred when comparing the centre mono subwoofer and the stereo configuration located to the left and right of the listener. Contrarily, no difference could be heard when comparing the front mono or stereo configurations. While Welti’s test partly dismissed that audible differences would be heard across the different subwoofer configurations, the author did suggest that this test does not prove in any way that there are no merits to stereo subwoofer reproduction") in this fascinating thesis, which addresses my question about low frequency localization in relatively free space: https://aaltodoc.aalto.fi/bitstream...tasa_Anamaria_2022.pdf?sequence=1&isAllowed=y

This thesis is really worth skimming! A lot of fascinating papers come out of Aalto University, which also has one of the most sophisticated auralization setups in the world. The original poster asked about science, so I hope the links that I've provided suggest that there may be more to bass quality than flat frequency response at the listening position, that bass <80 Hz may be difficult but not necessarily impossible to localize, and that different approaches might be considered to enhance spatial perception at low frequencies if desired.

Young-Ho

Thanks, I'll have to look into those papers as well.

The conclusions you state match my own experiences, particularly with using a center sub vs. on the left or right. My tests showed a single sub between the front speakers was best, and if you had a pair with one on either side it did not matter if they were mono or stereo. I still think it is the higher-frequency cues that provide localization and that fits that theory (hypothesis).

It's times like these I wish my day job was less demanding and I had more time to devote to "fun" research!

 

[FeddyLost]

Not only do you need to get >1 subwoofer, you need to place them carefully, do mic readings, generate individual filters for each sub, etc.
Not for the faint of heart.
If you're not willing to do that kind of homework, is >1 sub still recommended? Does it at least serve to even out the FR a bit and the spatial variation a bit? a lot?

Without proper tuning and aligning few subwoofers may cause more harm than single. Anyway, results will be unpredictable.
If you are not sure about your abilities of tuning such system, you can pay for turnkey services as often people do.

Also, I see from the Psychoacoustic filter in REW that for low frequencies, human hearing seems to be a lot less sensitive.
In that case, is a flat FR still the goal? Can some deviations from flatness pass unnoticed?
How flat is flat enough?

Unfortunately, flatness of the graph is not the only one characteristic of room's transfer function.
Typically, peaks on graph means loooong decay times at these frequencies and this is not what considered "good sound" throughout the industry.

And then we arrive to next station of subjective justification of investments.
Human hearing is not only less sensitive in LF and have very different personal qualification and preferences, but also very adaptive within "personally acceptable" area.
So, there is no single simple answer even without discussion about Fletcher-Munson curve and different target curves for different SPLs.

Personally I'd say that if you are not professional mastering engineer but just audio entusiast, most of the cases for single listening point can end with single decent subwoofer and good tuning/EQing applied. Narrow notches, even very deep ones, without rattling peaks and boomy bass trails below 100 Hz is absolutely OK for "customer listening". If you have relatively flat FR within +- 3 Db below 100 Hz, it's already great.
For sure, there might be some improvement if you go further down this rabbit hole, but I doubt that no one will tell you where is the point of diminishing returns at which you should stop. Even experienced studio/HT builder barely will predict exactly when you'll find that you invested too much for such result.

So, IMO it's better to visit some pro studios and check which sound do you prefer more. Then measure it and ask someone qualified if that's possible in your room.

In the worst case you'll finally find that you strongly prefer flat and dry bass and need it equally good in few spaced listening points. Then you'll definitely will have to work hard for this goal.

I'm sure many at ASR will have had the experience of showing their new setup, with one extra sub, 5 PEQ filters / a new component / whatever ... to a friend, wanting to get a WOW reaction, and have gotten a bemused "sounds nice. To be honest your system already sounded great".

Actually, correct sound almost never induces WOW because for most untrained listeners it's "nothing special".
Only trained professionals can really understand sound better than not having obvious barely tolerable flaws and good dedicated stereo setup will always be "great" for people listening soundbar as main music source.