Locating bass <80Hz?

[respice finem]

Just take a generator I suggest
Szynalski is a good one, because you can dial in directly and then sweep slowly by hand, and so pinpoint your individual frequency, at which you can locate the sub. Just watch the volume.

Online Tone Generator - generate pure tones of any frequency

www.szynalski.com

 

[Oddball]

You got a whole bunch of interesting clues. I was experiencing similar issues. 2 subs located symmetrically between FW and Surrounds solved the problem in the smaller setup/room. 4 subs in the larger setup/room did the job as well. Brute force is not to be underestimated.

 

[Kyle_PL]

Who sings 1 meter away from the mic?

Who listetning singer under 1 meter from his mouth?

 

[Sokel]

Who listetning singer under 1 meter from his mouth?

All of us probably:

 

[pablolie]

The cutoff isn't a clean straight line at all. If you xcover at 85Hz, you will have content above that, hence you could have content that can be "located".

 

[Sokel]

The cutoff isn't a clean straight line at all. If you xcover at 85Hz, you will have content above that, hence you could have content that can be "located".

The few studies we have about it have eliminated that, it's trivial I think.
Same with tests, one must take care not to built up sound nor use test signal above range.

Low distortion goes without saying so some of the usual subs which prioritize output vs distortion or make weird sounds because they are badly made must be excluded too.

 

[Oddball]

Lots of words, not any measurements. Is anyone going to post any measurements?

 

[Curvature]

Just take a generator I suggest
Szynalski is a good one, because you can dial in directly and then sweep slowly by hand, and so pinpoint your individual frequency, at which you can locate the sub. Just watch the volume.

Online Tone Generator - generate pure tones of any frequency

www.szynalski.com

Warbles and shaped noise are more useful. Sines are not a natural sound and are difficult to localize since they lack an envelope.

Griensinger has some available on his website. http://www.davidgriesinger.com/laaes2.pdf

https://www.dropbox.com/scl/fo/s09j...pRZg?rlkey=9mczd805df3r8ywlmu6dlp1w1&e=1&dl=0

 

[No_hair_left]

This is the exact reason I sold my sb 3000 and bought dual sb 1000 instead, no matter where I placed the single sub I could tell where it was the moment anything sent a signal to it, very disconnected to the screen, duals solved my problem and while they don't compare sound wise to the 3000 it was worth it.

 

[respice finem]

Warbles and shaped noise are more useful. Sines are not a natural sound and are difficult to localize since they lack an envelope.

Griensinger has some available on his website. http://www.davidgriesinger.com/laaes2.pdf

https://www.dropbox.com/scl/fo/s09j...pRZg?rlkey=9mczd805df3r8ywlmu6dlp1w1&e=1&dl=0

Interesting. But is there really much difference, specifically regarding locating a sub, proven by proper blind testing?

 

[LSPhil]

The core of the issue is the breakdown of the harmonic relationship, and the root cause lies in the inherent physical differences between the speakers.

A subwoofer is designed to move a lot of air for a very narrow bandwidth—often just two octaves. This demanding task, combined with the steep low-pass filter required to confine it, results in a significant inherent processing delay (group delay). In contrast, a satellite speaker, handling a much wider bandwidth of over seven octaves, is a far more agile system with much lower latency.

This timing discrepancy is critical. The higher harmonics (2nd, 3rd, 4th) from the fast satellites arrive at our ears first. Our brain initially attributes the entire sound to them. The fundamental frequency from the delayed subwoofer then arrives later, disconnected from its harmonics.

Because this low-frequency fundamental arrives as a separate, late event without its harmonic context, our ears can localise it to the subwoofer's position. The brain fails to fuse it with the sound from the satellites.

Therefore, the delay from the subwoofer isn't just a minor offset; it's a fundamental timing error that breaks the perceptual link between a note's foundation and its character, allowing the subwoofer to be localised. Time-alignment is essential to compensate for this and re-unite the fundamental with its harmonics.

 

[Kyle_PL]

All of us probably:

... but this mic dont take 20Hz.

The cutoff isn't a clean straight line at all. If you xcover at 85Hz, you will have content above that, hence you could have content that can be "located".

Yes. That's why I prefer lower crossover between sub and speakers, than 85Hz, when I have one sub.

 

[levimax]

Lots of words, not any measurements. Is anyone going to post any measurements?

OK here is a post with measurements as well my experience trying to recreate the OP's issue. The measurement is of the distortion at 80 dB at the LP taken near field from one of my subs which has a very large enclosure with an 18" professional driver so distortion is quite low so "hearing" harmonics is probably not happening.

I have a "spinning" office chair so I tried the OP's method of spinning while listening to tones and it was very interesting with the sound noticeably changing as I spun around until I got to ~45 Hz. At 45 Hz I could still spin around and stop myself when facing the speaker. At 40 Hz the sound was completely homogeneous as I spun around. At 50 Hz and above there was a large difference as I spun around and it was trivial to stop myself facing the speaker.

I am not sure how to interpret this as I never listen to music while spinning in a chair with my eyes closed.

 

[ChrisG]

Hi everyone,

I'm pleased that there's been some interesting discussion, and it looks like I'm not alone in this. I'd like to clear up a few things:
- As noted in the first post, if I play sine tones, the usual rules of standing waves apply, and it's difficult for me to locate the origin of the sound. If I play time-varying signals (music, tone bursts), the location of the subwoofer is immediately obvious.
- The SE7261A is rated for <3% each of 2nd and 3rd harmonic distortion at 90dB@1m in half-space. I'm probably running closer to 60-70dBSPL (at 3m), with the sub in quarter-space, and with the room helping out with some LF reinforcement. Distortion should be low.
- I have implemented a 60Hz 4th order lowpass using the RME interface, and there is an additional 85Hz lowpass built into the subwoofer itself. I'd estimate that the sub's lowpass is 4th order as well, so there's a 48dB/octave slope above 85Hz.
- The overall system response is flat from LF to HF. No sub boost.
- I experimented with time alignment to get the subs & mains working nicely together. Integration/summation is good, although it's possibl

It's getting late here, but I will re-do some measurements tomorrow and post those up. I'm thinking:
- Response of each of L/R/Sub/summed at LP
- Close-up of sub only
Let me know if any others should be added.



Finally, on to the discussion of the frequency range of the human voice.
Plosives have content down to the single-digits of Hz. If you talk face-to-face with the wrong person, you'll sometimes feel the wind blasts on your face. I'll leave it as an exercise for the reader to determine how we might recreate that.
Every vocal mic that gets used close-up will have some LF rolloff built-in, so those plosive sounds might only sound like they're happening 80-160Hz (for example). The reality of that situation is that 80Hz is about as low as most stage vocal mics will go.
My favourite stage vocal mic (I own six of them, they get used for every show I run sound for) is the Sennheiser e965, which extends down to about 30Hz - much more useful for those voices that do hit the bass register. They get much more "support" from the mic compared to the usual suspects from Shure/Neumann etc which roll off two octaves higher.

If anyone would like to discuss such things further, I'd appreciate a separate thread being started. I'd be very happy to contribute from my hands-on experience.

 

[ChrisG]

OK here is a post with measurements as well my experience trying to recreate the OP's issue. The measurement is of the distortion at 80 dB at the LP taken near field from one of my subs which has a very large enclosure with an 18" professional driver so distortion is quite low so "hearing" harmonics is probably not happening.

I have a "spinning" office chair so I tried the OP's method of spinning while listening to tones and it was very interesting with the sound noticeably changing as I spun around until I got to ~45 Hz. At 45 Hz I could still spin around and stop myself when facing the speaker. At 40 Hz the sound was completely homogeneous as I spun around. At 50 Hz and above there was a large difference as I spun around and it was trivial to stop myself facing the speaker.

I am not sure how to interpret this as I never listen to music while spinning in a chair with my eyes closed.

Thanks for trying! We can call this method: Spinorama V2.0.

Could you try with repeated tone bursts, rather than continuous tones?
I think there's something in the first arrival that's giving our hearing system a location clue.

 

[Curvature]

Interesting. But is there really much difference, specifically regarding locating a sub, proven by proper blind testing?

Just the 1980s AES papers on the topic as far as I know.

Edit: Which to be clear established the 80Hz rule we tend to follow and the had the general observation that below 200Hz it's hard to hear direction. These are good papers, but the topic needs more precise work because of what has been discussed by Griesinger and Lund.

 

[Curvature]

I am not sure how to interpret this as I never listen to music while spinning in a chair with my eyes closed.

Head movement allows one to pick up timing cues at low frequencies. Actually without head movement things start failing: front-back, vertical position, side position (cone of confusion).

Very interesting, by way. I'd like to try that, too.

 

[Curvature]

I think there's something in the first arrival that's giving our hearing system a location clue.

First arrival is relevant only above 1.5kHz, where ILD dominates.

 

[Sokel]

First arrival is relevant only above 1.5kHz, where ILD dominates.

Sometimes I feel that is as simple as the "air" that lows inevitably push toward us.
Of course we can argue that by the different designs (downfiring, etc) but...

 

[Curvature]

The core of the issue is the breakdown of the harmonic relationship, and the root cause lies in the inherent physical differences between the speakers.

A subwoofer is designed to move a lot of air for a very narrow bandwidth—often just two octaves. This demanding task, combined with the steep low-pass filter required to confine it, results in a significant inherent processing delay (group delay). In contrast, a satellite speaker, handling a much wider bandwidth of over seven octaves, is a far more agile system with much lower latency.

This timing discrepancy is critical. The higher harmonics (2nd, 3rd, 4th) from the fast satellites arrive at our ears first. Our brain initially attributes the entire sound to them. The fundamental frequency from the delayed subwoofer then arrives later, disconnected from its harmonics.

Because this low-frequency fundamental arrives as a separate, late event without its harmonic context, our ears can localise it to the subwoofer's position. The brain fails to fuse it with the sound from the satellites.

Therefore, the delay from the subwoofer isn't just a minor offset; it's a fundamental timing error that breaks the perceptual link between a note's foundation and its character, allowing the subwoofer to be localised. Time-alignment is essential to compensate for this and re-unite the fundamental with its harmonics.

It's an interesting thought, but I don't believe this is valid to VLF for a few reasons. I'd say it makes more sense above 1.5kHz and is a decent argument for linear phase correction.

Group delay measured in subs is below audible thresholds in all examples I've seen. Josh Ricci was especially good about including these in his reviews https://data-bass.com/ And I believe Audioholics also makes group delay measurements.

The ear has both forward and backward masking (in time), and VLF have a long integration window. In-room VLF decays take seconds, and with real music the decays will be persistent, almost never fully dissipating. That adds to masking of upper partials. "Partials" because any signal outside of a sine wave, distortion will be spectrally complex and not truly harmonic.

A speaker and sub may be time-aligned and fire simultaneously, but in multisub setups the listener will usually not be equidistant from each sub. Yet listener reports of directional bias don't happen. Even the sub being tens of milliseconds off doesn't cause appreciable issues if the response is reasonably flat. I agree it's important to have simultaneous driver firing so that we are, say, mathematically closer to the ideal, but in a practical sense it doesn't seem to matter if you have reasonable phase alignment (i.e., sub and satellite FR does not show nonminimum issues such as cancellation).

Finally, the ear is not sensitive to first arrival and instead relies on phase locking below 1kHz. (Between 1-2kHz there is a crossover of sorts between these mechanisms.)