Stereo Bass using subwoofers

[DonH56]

May I know in any of your prior tests, did it involves artificial low sub bass in the 30hz?

See the description of the files: https://www.audiosciencereview.com/...tereo-bass-using-subwoofers.11034/post-928692

The files are all filter at 48db sloop at 80hz. The main tones are artificial bass in 32 to 34hz. The files are difference of intensity between left and right of 20db, 10db, and 3db.

It was tested in big room (5000+ cubic ft) that opens to the rest of the house with subs eq to deal with room modes. So it is not a typical closed room and not your typical non eq subs. More details here:

Stereo Bass using subwoofers

Hmm, this particular one is about difference in 32 to 34 Hz. If I turn off ARC, then right subs will have significant dB peak due to room mode, and left sub will have significant null due to room mode. The difference, based on the measurements, should be in the +10db range. This +10db imbalance...

www.audiosciencereview.com

I am pretty sure that none of your tests involved such 30hz artificial sub bass tones with such big db difference between left and right channels. I really hope you can perform tests with such files with a stereo subs yourself, and then come back to let us know about your own observations.

Test files available here:

Traust - Google Drive

drive.google.com

Start with files 9 and 10. If you cannot sense a different, no point going further.

I have not had my system on in a week except to watch news or a ball game whilst practicing, too busy at work.

The tests IIRC (it was a long time ago) used single and swept tones from 5 Hz to 250 Hz as well as pink noise and some demo tracks (like the old Sheffield Drum Tracks, an MF organ recording I have forgotten, a few others). The focus was less on stereo effects than localizing low frequencies particularly in-room. It was hard even back then finding recordings with stereo bass, and multichannel did not exist for all practical purposes (we did have the old Pink Floyd Quadraphonic album).

Obviously, or maybe not, if you place two subs widely apart in a field you can begin to localize very LF tones, though it was harder than we thought. I found an article back then talking about determining storm direction from thunder that was interesting, wish I had it and/or remembered more. All I recall is that the directional information was basically from the sharper "crack", and the rumbles were more difficult if not impossible to localize, but general direction could be determined from a distance.

The only thing I know for certain is that my opinion and my old essentially un-documented test results aren't going to mean anything to anybody. I'll try to ignore this thread.

 

[Chromatischism]

See figure 7.2 on pg161 in 3rd edition of Toole's book for more details. It was measured in a variety of ways and also confirmed by multiple studies and is pretty audible with pink noise comparisons as well. I don't think of it as new/interesting tbh. If you have music that is available and recorded very similarly in both stereo and mch formats, it's striking how much clearer a real center is.

Interesting. I don't doubt that a real speaker is better, but that test was done in an anechoic chamber. In a normally reflective room much of that interference dip would fill in.

Edit: in fact if you read on, Toole says exactly the same thing.

 

[Pdxwayne]

I have not had my system on in a week except to watch news or a ball game whilst practicing, too busy at work.

The tests IIRC (it was a long time ago) used single and swept tones from 5 Hz to 250 Hz as well as pink noise and some demo tracks (like the old Sheffield Drum Tracks, an MF organ recording I have forgotten, a few others). The focus was less on stereo effects than localizing low frequencies particularly in-room. It was hard even back then finding recordings with stereo bass, and multichannel did not exist for all practical purposes (we did have the old Pink Floyd Quadraphonic album).

Obviously, or maybe not, if you place two subs widely apart in a field you can begin to localize very LF tones, though it was harder than we thought. I found an article back then talking about determining storm direction from thunder that was interesting, wish I had it and/or remembered more. All I recall is that the directional information was basically from the sharper "crack", and the rumbles were more difficult if not impossible to localize, but general direction could be determined from a distance.

The only thing I know for certain is that my opinion and my old essentially un-documented test results aren't going to mean anything to anybody. I'll try to ignore this thread.

Thank you for the feedbacks! From what you said, you confirmed that you did not test stereo bass where left and right channel db difference is large.

If you have a way to switch to stereo subs easily and your setup can accept USB output from your laptop, the tests with file 9 and 10 won't take you more than 5 minutes. It was so obvious that it took me a little over 2 minutes to complete 10 trials.

Would love to have you add your abx results in this thread. That would add to peer review results to confirm or not confirm my ABX results.

Anyway, fyi, for me, 3db difference between channels is at the edge of detectability. 9db and 20db difference were very obvious.

 

[MakeMineVinyl]

My main problem with the binaural approach is it only captures what someone would hear in a precise position. Once you're used to regular multi-channel, headphone-like audio destroys immersion because even if you only move your head a half inch or turn it slightly, there is no accompanying shift in the sound that should be there.

I don't listen through headphones, but my speaker setup (full range horns) does reproduce binaural very well. Now I freely admit that my setup is extremely anti-social. If you aren't exactly in the prime listening position, you get what you get, which ain't much. But that's OK since I'm the only one listening. The only time my wife listens to music is so that she can pick songs for the dance class she teaches, and then she's out in the middle of the floor doing dance moves. The cat isn't interested - he's too smart to get sucked into the audiophile abyss.

 

[Soundmixer]

Another complaint I have with discrete multi-channel is that under any sane conditions, the surround speakers are smaller than the mains (and the mains must be large to be convincing with organ music), so there is a front/back disconnect. If the total soundfield is coming from two stereo speakers, this problem does not exist. My mains are refrigerator-sized, so you can see the problem there. I do listen to all types of music, but I'm using organ music as an example.

Your statement here is incorrect. It does not matter if the size of the front speakers is different than that of the surrounds, as long as the timbre matching between the two is very close. The problem with stereo is the only additional enveloping information coming from the sides and rear is in the form of room reflections. If your stereo speakers have a poor off-axis response, those reflections will have a different timbre than the direct arrival. That does not happen with multichannel.

 

[Sancus]

Interesting. I don't doubt that a real speaker is better, but that test was done in an anechoic chamber. In a normally reflective room much of that interference dip would fill in.

Edit: in fact if you read on, Toole says exactly the same thing.

Yeah, he says it helps but doesn't fix it completely. It's still audible. Even then, that only applies when the direct sound is not dominant. Some people try to make the direct sound dominant deliberately, or just end up that way. In my setup listening distances are quite close(1.5-2.5m depending on the speaker) so it would probably be more audible than a 3-4m setup with very wide dispersion speakers. Anyway, that's just one of several stereo issues that are resolved by multi-channel.

Your statement here is incorrect. It does not matter if the size of the front speakers is different than that of the surrounds, as long as the timbre matching between the two is very close. The problem with stereo is the only additional enveloping information coming from the sides and rear is in the form of room reflections. If your stereo speakers have a poor off-axis response, those reflections will have a different timbre than the direct arrival. That does not happen with multichannel.

I was trying not to make this the multi-channel v. stereo thread, but since discussion dropped off anyway... whatever. Agreed, smaller surrounds is not that big of an issue. The more channels you have, the smaller each speaker generally needs to be, especially with frontal-soundstage material. Adding the center speaker alone means 5-6dB more output from your system, and the surrounds(together, playing reflections/ambience) usually add like 3-6dB as well. The total output to achieve a certain SPL at the listening position is spread over more speakers. The exception to this is music where the surrounds individually play parts in isolation but this is sufficiently uncommon that it's not a big deal, and some people don't even like music recorded/mixed this way anyways. It's even more exceptionally uncommon for heights.

All that said, if you have the space for it, there is no problem with using large floorstanders for surrounds, and people certainly do it.

 

[Pdxwayne]

I am traveling this week but I will try it this weekend in a room that might give me pretty symmetrical L and R. I'll use different filtering and speeding up the tracks so I can use my mains speakers, as I mentioned previously.

One thought though: you mentioned that you heard distortion with the track when you initially tried an ABX with the subs. As part of your posted ABX tests, did you measure or hear distortion to make sure there was nothing significant being put out at higher frequencies? The distortion might help you to localize.

Since there are questions about harmonics, which could clue me in the directions, I decided to do one more ABX with these settings:

*Using a reasonable lower volume than before. Using RadioShack SPL meter, song #9 measured at left ear position is ~69db C weighted. Including adjustment for distance of ~12 ft, I would estimate the sub spl at 3 ft is about 78db. I would say at 78db, the sub's harmonics over 80hz likely too low to be audible. You can see the spectrum analysis below to confirm.

A bit more about my subs' measurements:

Paradigm Reference Seismic 110 Subwoofer - HomeTheaterHifi.com

The Paradigm Seismic 110 subwoofer is a freak. It is very small, lightweight as subwoofers go; yet capable of playing very loud, very low, and very clean. It will blow away subwoofers that are bigger, heavier, and in some cases, more expensive. Highly recommended!

www.google.com

Quote:
In fact, since the Paradigm was not even breathing hard at 31.5 Hz, I started turning it up. In this graph, the Seismic was up to almost 110 dB at less than 2.5% distortion, and sounded like it could handle much more.

So, at 78db at 3 ft, the distortion over 80hz should not be a concern. Also note that at this volume, the subs have no rattle issue.

*Remove test songs' signals over 80hz much more (more than what you already crossed at 80hz with 48db slope) by using the lowest crossover of the subs (which is 30hz).
Here is the spectrum analysis from mic captured (as captured from my left ear, 10 seconds section with 30hz bass) when playing song #9.

I will say anything over 80hz will be low enough to be masked by the 30hz tones. Thus, harmonics should not give any clue of directions.

I must admit that, with these adjustments, even using 20db difference files, it is much harder to tell the directions. But, I can still hear the directions.

Still got 10/10 ABX.

 

[pjug]

Since there are questions about harmonics, which could clue me in the directions, I decided to do one more ABX with these settings:

*Using a reasonable lower volume than before. Using RadioShack SPL meter, song #9 measured at left ear position is ~69db C weighted. Including adjustment for distance of ~12 ft, I would estimate the sub spl at 3 ft is about 78db. I would say at 78db, the sub's harmonics over 80hz likely too low to be audible. You can see the spectrum analysis below to confirm.

A bit more about my subs' measurements:

Paradigm Reference Seismic 110 Subwoofer - HomeTheaterHifi.com

The Paradigm Seismic 110 subwoofer is a freak. It is very small, lightweight as subwoofers go; yet capable of playing very loud, very low, and very clean. It will blow away subwoofers that are bigger, heavier, and in some cases, more expensive. Highly recommended!

www.google.com

Quote:
In fact, since the Paradigm was not even breathing hard at 31.5 Hz, I started turning it up. In this graph, the Seismic was up to almost 110 dB at less than 2.5% distortion, and sounded like it could handle much more.

So, at 78db at 3 ft, the distortion over 80hz should not be a concern. Also note that at this volume, the subs have no rattle issue.

*Remove test songs' signals over 80hz much more (more than what you already crossed at 80hz with 48db slope) by using the lowest crossover of the subs (which is 30hz).
Here is the spectrum analysis from mic captured (as captured from my left ear, 10 seconds section with 30hz bass) when playing song #9.
View attachment 157772
I will say anything over 80hz will be low enough to be masked by the 30hz tones. Thus, harmonics should not give any clue of directions.

I must admit that, with these adjustments, even using 20db difference files, it is much harder to tell the directions. But, I can still hear the directions.

Still got 10/10 ABX.
View attachment 157773

I set up a couple of small subs in a rectangular room, where I was getting a pretty smooth and nearly identical L and R response between 30Hz and 40Hz. The subs were in corners 2 ft from side and back walls, about 8 feet between them. I made files LP filtered at 35Hz (48dB slope) so.

What I found is that when I was closer to the subs, or more or less inside an equilateral triangle with the subs, then I could clearly tell if L or R were playing. But when I sat about 10 feet away then my sense of directionality completely disappeared. I failed an ABX test (got 8 of 16, felt like guessing).

This is not really too out of line with your results since you said the 20dB difference files made a hard ABX test even though you got 10/10.

 

[Pdxwayne]

I set up a couple of small subs in a rectangular room, where I was getting a pretty smooth and nearly identical L and R response between 30Hz and 40Hz. The subs were in corners 2 ft from side and back walls, about 8 feet between them. I made files LP filtered at 35Hz (48dB slope) so.

What I found is that when I was closer to the subs, or more or less inside an equilateral triangle with the subs, then I could clearly tell if L or R were playing. But when I sat about 10 feet away then my sense of directionality completely disappeared. I failed an ABX test (got 8 of 16, felt like guessing).

This is not really too out of line with your results since you said the 20dB difference files made a hard ABX test even though you got 10/10.

Thanks for doing the tests!

I was about 12 ft away, subs about 10 ft aparts. Indeed, not easy as the low sub bass seems like coming from all around. But, I can still hear pressure differences.

Can I say you confirmed that hearing directivity of sub bass is possible, when the combinations of room and distance are right?

 

[pjug]

Thanks for doing the tests!

I was about 12 ft away, subs about 10 ft aparts. Indeed, not easy as the low sub bass seems like coming from all around. But, I can still hear pressure differences.

Can I say you confirmed that hearing directivity of sub bass is possible, when the combinations of room and distance are right?

I did not abx it since I wouldn't listen that way. But when I was closer, inside an equilateral triangle is where hearing the direction seemed obvious in the arrangement I set up. But this was with the 20dB L/R difference. I don't know that it would be noticeable with more subtle L/R difference. I would say certainly it is possible to hear a L/R difference. Really anyone can get close to one sub to the point where you can hear where the sound comes from. To me the question is would it ever matter in my own listening and I think the answer to that, for me, is no.

 

[Pdxwayne]

I did not abx it since I wouldn't listen that way. But when I was closer, inside an equilateral triangle is where hearing the direction seemed obvious in the arrangement I set up. But this was with the 20dB L/R difference. I don't know that it would be noticeable with more subtle L/R difference. I would say certainly it is possible to hear a L/R difference. Really anyone can get close to one sub to the point where you can hear where the sound comes from. To me the question is would it ever matter in my own listening and I think the answer to that, for me, is no.

Nice! I appreciate you taking time to tests it out and post your findings here!

BTW, did you have a chance to ABX #3 and #4 using headphones? I am curious how you would do. Thanks again!

 

[pjug]

Nice! I appreciate you taking time to tests it out and post your findings here!

BTW, did you have a chance to ABX #3 and #4 using headphones? I am curious how you would do. Thanks again!

Likewise I appreciate what you've done and also any comments for anyone else whether I agree or not. I will try 3 and 4 on headphones tomorrow. I am wondering if playing too loud was a problem the first time I tried it (and bailed).

 

[pjug]

Here is my ABX of 3 and 4 with headphones, FWIW.

 

[Pdxwayne]

Here is my ABX of 3 and 4 with headphones, FWIW.
View attachment 158347

Great! Thanks for sharing!

 

[Pdxwayne]

I did not abx it since I wouldn't listen that way. But when I was closer, inside an equilateral triangle is where hearing the direction seemed obvious in the arrangement I set up. But this was with the 20dB L/R difference. I don't know that it would be noticeable with more subtle L/R difference. I would say certainly it is possible to hear a L/R difference. Really anyone can get close to one sub to the point where you can hear where the sound comes from. To me the question is would it ever matter in my own listening and I think the answer to that, for me, is no.

BTW, regarding your room, should I assume it is an enclosed room with these dimensions?
~12ft wide
~15ft long
~9 ft high

If you sit within 8 ft of subs, then you can hear directivity of subs, when playing test files (low passed at 30 Hz with 48db slope).

If you sit about 10ft away from the subs, then you can't hear directivity.

Based on your headphones ABX results, I will assume we have similar sensitivity to sub bass dB changes. So, I would say, in your room, I would have failed too if sitting 10ft away.
: )

 

[pjug]

BTW, regarding your room, should I assume it is an enclosed room with these dimensions?
~12ft wide
~15ft long
~9 ft high

If you sit within 8 ft of subs, then you can hear directivity of subs, when playing test files (low passed at 30 Hz with 48db slope).

If you sit about 10ft away from the subs, then you can't hear directivity.

Based on your headphones ABX results, I will assume we have similar sensitivity to sub bass dB changes. So, I would say, in your room, I would have failed too if sitting 10ft away.
: )

It is 12x16x8high. Maybe you would do better on ABX where I was sitting, but my feeling is that if 20dB L/R difference is hard to distinguish then there is no point in having stereo subs in that kind of setup.

 

[Pdxwayne]

It is 12x16x8high. Maybe you would do better on ABX where I was sitting, but my feeling is that if 20dB L/R difference is hard to distinguish then there is no point in having stereo subs in that kind of setup.

Well, do note that I added 24db filter to the test files from 30hz on using the subwoofers lowest crossover.

You on the other hand did even more, 48db filter from 30hz.

Stereo bass should not just about 30hz, right? There are still notes from 30hz up to 80hz, right? So, using our results to make a statement of "not worth" is a bit strong, don't you think?

 

[pjug]

Well, do note that I added 24db filter to the test files from 30hz on using the subwoofers lowest crossover.

You on the other hand did even more, 48db filter from 30hz.

Stereo bass should not just about 30hz, right? There are still notes from 30hz up to 80hz, right? So, using our results to make a statement of "not worth" is a bit strong, don't you think?

True, I used 35Hz 48dB filter so I was just getting that narrow band peaking at 31Hz. I don't know how to make a good test without a narrow band track though. I thought my room would be fairly symmetrical but it isn't really. The L and R have several modes between 40Hz and 100Hz that are different in the L and R. The L and R curves were almost the same between 30Hz and 40Hz, though.

That's why I abandoned my idea of speeding up the track and using the main speakers that were already in the room.

 

[Chromatischism]

In summary of the last couple of pages:

You can detect the direction of the subs when:

• One is 20 dB louder than the other
• You sit close to it

Would this represent a real-world scenario where someone is trying to achieve "stereo bass"?

 

[Pdxwayne]

In summary of the last couple of pages:

You can detect the direction of the subs when:

• One is 20 dB louder than the other
• You sit close to it

Would this represent a real-world scenario where someone is trying to achieve "stereo bass"?

The tests were to confirm if low sub bass directions can be audible when there are no notes above 80hz that are loud enough to give direction clues.

In our tests, 30hz tones directions can be audible, even when low passed at either:
24db slope at 30hz (in addition to 48db at 80hz)
Or
48db slope at 35hz

At 24db slope at 30hz (in addition to 48db at 80hz), I can hear directions from 12ft away (my listening seat) and I did not check closer to subs. Note that my room is +5000 cubic ft and opens to all downstairs and upstairs. Also, for this particular scenario, I only checked 20db files.

At 48db slope at 35hz, in an enclosed small room, @pjug needs to get closer to subs.

Neither case would represent typical music. 30hz tones that steeply low passed do not really sound that good.
; )

Main purpose is to confirm if one can hear directivity of bass. Also note that our tests were focusing on ~30hz. There are still tones above 30hz, but below 80hz, that we haven't checked. So the two observations you mentioned, 20db difference and closer to subs might change with higher sub tones. For those higher bass notes, maybe even at much lower db difference and at normal sitting location, one can sense directions in certain type of rooms.