Bass and subwoofers

[olds1959special]

I like having multiple subwoofers that are in stereo and near my main speakers. Also adding a center subwoofer is not a bad idea.

 

[KenMasters]

I revisited stereo vs mono bass using ART. I used only the two subs placed to the outside of my fronts, I sit just over 2m from the front plane, so the front subs are spaced quite far to the sides of me. Two presets, one with subs and supporting speakers combined, one with left sub/left speakers and right sub/right speakers only - all else equal.

Unlike when testing using manual EQ and Denon's directional bass, I could not tell a difference. They sounded pretty much identical to me - only here or there a little less kick to the bass in stereo, but the difference of the "closed/pressured" feeling of the mono bass I experienced previously when experimenting with Denon's Directional bass option was not present.

I know, not exactly the pinnacle of scientific rigor, but enough for me to feel comfortable with leaving things as they are in my room with my setup.

 

[3ll3d00d]

Correct me if I'm wrong, but the prevailing way of correcting, is using L+R signal together down low.

I think correction methods don't care about what goes in, only what comes out, i.e. they equalise each physical output channel individually to a given target response (some minor exceptions to this are schemes which seek to make pairs of channels alike) irrespective of what content is played and how it's distributed across channels (whether coherently or not)

 

[Flaesh]

I'll publish it someday.. just 2+2 subwoofers in a living room with 4 windows, 3 doors, and some furniture.

So far, only an unfiltered teaser in one point.

On the left windows and doors are visible, but the table is almost invisible somewere at 40-70. Above 100 Hz it doesn't matter.

ps:

Unfortunately, the numbering got mixed up for some unknown reason.

 

[Keith_W]

I would say that's pretty remarkable given where you have placed the microphone. Any other subwoofer configuration would look umpteen times worse than that.

 

[Flaesh]

It's designed to allow listening in different places in the room. In a smoother room, like a dedicated cinema, the deviations would be smaller.
And here it is, two inches from the cone:

I'll write more in my thread later.

 

[bmc0]

I decided to experiment with creating the above signals in a different way: directly in the frequency domain.

These signals have the same problem as the organ-esque signal, so I modified the program I wrote to apply a correction based on the interchannel correlation, plus a small "fudge" factor to account for imperfect acoustic summing. Here are the magnitudes (single channel) of the new uncorrelated and mono'd LF signals, according to REW's RTA:

And this is the measured difference (offset +80dB) between the two when played back over my loudspeaker system, shown with 1/6th octave smoothing:

L and R are fairly well matched in both magnitude and phase in my setup, so the error is relatively small (note the scale). On a setup where the LF sources don't sum quite so coherently, a different correction factor would be required.

 

[levimax]

These signals have the same problem as the organ-esque signal, so I modified the program I wrote to apply a correction based on the interchannel correlation, plus a small "fudge" factor to account for imperfect acoustic summing. Here are the magnitudes (single channel) of the new uncorrelated and mono'd LF signals, according to REW's RTA:
View attachment 525246
And this is the measured difference (offset +80dB) between the two when played back over my loudspeaker system, shown with 1/6th octave smoothing:
View attachment 525249
L and R are fairly well matched in both magnitude and phase in my setup, so the error is relatively small (note the scale). On a setup where the LF sources don't sum quite so coherently, a different correction factor would be required.

For what it's worth I actually find these 2 files easier to tell apart, probably because no one is home and I can turn it up louder. The mono track definitely has a louder deep bass part. I am not sure what exactly I am listening to but it almost seems like some type of "beat" frequency that is very LF and it comes and goes but when it "comes" it is louder on the mono track and it is low enough to vibrate my desk and it "feels" louder as well. Interestingly I have summed the Billie Eilish track to mono below 100 Hz and can not reliably ABX it against the original. It seems like there is too much else going on in the song to be able to consistently hear the LF part. With this track the LF part comes through loud and clear which makes it a lot easier. Not sure what this means exactly.

 

[bmc0]

For what it's worth I actually find these 2 files easier to tell apart, probably because no one is home and I can turn it up louder. The mono track definitely has a louder deep bass part.

Interesting; you might try quantifying this with measurements. If you have a mic with analog output and can do playback and capture from the same device, set REW's RTA to 1/48th octave, 1M FFT length, rectangular window, and 0% overlap, then run two or three averages with each file playing (looped without a gap or crossfade) and compute A/B of the two results.

 

[levimax]

Interesting; you might try quantifying this with measurements. If you have a mic with analog output and can do playback and capture from the same device, set REW's RTA to 1/48th octave, 1M FFT length, rectangular window, and 0% overlap, then run two or three averages with each file playing (looped without a gap or crossfade) and compute A/B of the two results.

I have analog MIC and 2 interfaces so can do this unless the capture and playback have to be from same device? MMM or stationary MIC?

 

[bmc0]

The capture and playback clocks must be synced if you use the settings I posted. If that is not possible, you should use a different window (Hann is fine) and increase overlap. The mic should probably be stationary since the difference between the two cases is the important bit.

Interestingly I have summed the Billie Eilish track to mono below 100 Hz and can not reliably ABX it against the original.

I just listened to a section of this track ("I Didn't Change My Number") again after reading your comment. Personally, on my system, I find the difference in spatial quality with bass mono'd below 80Hz to be significantly more obvious than the noise signals I posted. Do note that the bass in that track is mono in some parts and not in others, which is why Thomas Lund has mentioned it as a good example of AEV contrasts. Here's a plot of a crude "LF envelopment" metric I was playing with a while ago (see post #212 in this thread):

Referring to the gray/black lines in the bottom plot: The group of spikes from ~28%-36% correspond to the section starting with "Don't take it out on me". The next group beginning at ~45% corresponds to the section starting with "I gotta work, I go to work". The value is low during the sections with mono bass.

 

[levimax]

The capture and playback clocks must be synced if you use the settings I posted. If that is not possible, you should use a different window (Hann is fine) and increase overlap. The mic should probably be stationary since the difference between the two cases is the important bit.


I just listened to a section of this track ("I Didn't Change My Number") again after reading your comment. Personally, on my system, I find the difference in spatial quality with bass mono'd below 80Hz to be significantly more obvious than the noise signals I posted. Do note that the bass in that track is mono in some parts and not in others, which is why Thomas Lund has mentioned it as a good example of AEV contrasts. Here's a plot of a crude "LF envelopment" metric I was playing with a while ago (see post #212 in this thread):
View attachment 525304
Referring to the gray/black lines in the bottom plot: The group of spikes from ~28%-36% correspond to the section starting with "Don't take it out on me". The next group beginning at ~45% corresponds to the section starting with "I gotta work, I go to work". The value is low during the sections with mono bass.

I am not 100% sure I did the sum to mono right on Billie Eilish, I will try again during the correct times in the song when I get a chance as well as double check my sum to mono work. I will also do the RTA measurement one way or another.

 

[levimax]

The capture and playback clocks must be synced if you use the settings I posted. If that is not possible, you should use a different window (Hann is fine) and increase overlap. The mic should probably be stationary since the difference between the two cases is the important bit.


I just listened to a section of this track ("I Didn't Change My Number") again after reading your comment. Personally, on my system, I find the difference in spatial quality with bass mono'd below 80Hz to be significantly more obvious than the noise signals I posted. Do note that the bass in that track is mono in some parts and not in others, which is why Thomas Lund has mentioned it as a good example of AEV contrasts. Here's a plot of a crude "LF envelopment" metric I was playing with a while ago (see post #212 in this thread):
View attachment 525304
Referring to the gray/black lines in the bottom plot: The group of spikes from ~28%-36% correspond to the section starting with "Don't take it out on me". The next group beginning at ~45% corresponds to the section starting with "I gotta work, I go to work". The value is low during the sections with mono bass.

I'm sorry, I made a careless mistake. The tracks that I compared were the previous "organ like" tracks with the boosted base. I thought you had rebalanced them but now I realize the tracks that are rebalanced are noise track which are not the tracks I could tell a difference on. Sorry about that. At least I learned how to loop tracks in Audacity and it looks like I can playback and record on the same device now.

 

[JRowland]

I just listened to a section of this track ("I Didn't Change My Number") again after reading your comment. Personally, on my system, I find the difference in spatial quality with bass mono'd below 80Hz to be significantly more obvious than the noise signals I posted. Do note that the bass in that track is mono in some parts and not in others, which is why Thomas Lund has mentioned it as a good example of AEV contrasts.

That’s what sucked me in, as a listener having that mono bass punching you in the chest while the other bass “explodes around the room”. It’s even better in surround using the Atmos mix, crazy from the rear speakers. Sorry for the tangent, glad someone understands.

 

[Thomas Lund]

It's as simple as thinking what the artist heard and bought at the control room with the flush mounted mains.
And as shown at the many songs analyzed at this thread, they absolutely heard stereo bass, no matter is the room spoiled it or not.

Yes, recreational reproduction vs. monitoring is key to this discussion. Considering the former, whatever gets you through the night can't just be dismissed. I am merely objecting to making pragmatism the ideal, for instance leading to somebody here describing recording professionals being half deaf. Music tracks can more likely sound off because of a systemic pro ability to recognise dimensions and details recreationals don’t.

One such quality is LF inter-aural fluctuation (AEV), latent in a recording. Judged by Bruce Swedien’s early productions, and by comments here, AEV must have been conveyed more reliably using flush-mount Urei monitors in Universal studios around 1960, than it is today, in living rooms around the world. Monitors have since improved but flush-mount still has advantages. A shorter listening distance also has, if the room is small and/or acoustically compromised.

Because art is all about touching fellow human beings, it shouldn’t come as a surprise when artists keep using that potent auditory stimulus, modulation of LF inter-aural movement/fluctuation.

Had perception been prioritised when the misleading home reproduction LF studies were conducted decades ago, they had focused more on change/movement than on measurable, absolute snap-shot metrics [previous posts]. Sometimes the real question is not: “Is this correct?” It is: “What made it look correct?” In case anybody is going to the AES convention in Copenhagen at the end of May, Geoff Martin will be giving a keynote wonderfully titled "The Perceptual Irrelevance of Physical Measurements”. Provoking probably everybody - including me

highly complex (AEV dimension) and not at all black and white, as in either this or that (stereo/mono).

There should be more results about AEV as a dimension by the end of the year. For now, it would be a pity if excessive spatial reproduction-distortion just continues, even more so beyond stereo. Because we once forgot to listen relevantly, normalised malpractice and created a long list of ripple effects, affecting also production and distribution.

 

[Andysu]

i do not care to sweep at loud SPL levels , unlike some that are hard of hearing

 

[MayI]

In case anybody is going to the AES convention in Copenhagen at the end of May, Geoff Martin will be giving a keynote wonderfully titled "The Perceptual Irrelevance of Physical Measurements”. Provoking probably everybody - including me

That should go well here . Hopefully it will be recorded for future viewing by those unable to attend.

 

[Andysu]

sw sweep

 

[Andysu]

two sweeps here and ooooohhh only i know what is going on , ooohhh heard that jet airliner outside wow it was in mono due to window being slightly open , it sounded better than dolby's own atmos

but the sw ooohhh two seeps here which is what most many out of i say 99% wouldn't even check , ooohhh there is something going on here , yes i know what it is , errr , sigh but that's what you get , oooohhhh

 

[audiofooled]

Geoff Martin will be giving a keynote wonderfully titled "The Perceptual Irrelevance of Physical Measurements”. Provoking probably everybody - including me

Looking forward to this, and as @MayI says, hopefully it gets recorded.

There should be more results about AEV as a dimension by the end of the year

Again, looking forward to that and I truly hope you'll keep us posted.