Well, if only one side (L or R) for some reason contains low frequency information (most information below 100 Hz or so is usually paned dead center so mono i o w) it would be like having only one sub (or a normal full range speaker for that matter).

It makes sense to me to assume that the best possible full range stereo reproduction system consists of … two full range speakers … right? Using only one source for the lower frequency range and two for only the upper range is a compromise, but it makes sense if surround is needed since it can otherwise quickly become quite costly if all speakers need to be full range … But when I design home theaters (which happens some time if the project as potential); I always use two subs (L & R) for the reasons exemplified above.

On Madona's CDs (and other people) or much recordings in classical music reproduce the basse below 100 hz in stereo.

Well, ideally; mastering engineers (or mixing engineers for that matter) would sum low frequency content below at least 80 Hz or so to mono. It´s just a waste of cone travel to only let one side of a system reproduce low frequencies that can’t be localized anyway. As long as the harmonics of for example a floor tom is coming from one side; it will still be perceived as coming only from that side.

Attention mastering engineers: Sum low frequencies to mono! It will result in lover distortion and less modal issues when listening to your work.

for the anecdote : i know a label Passavant Music where the label manager records in stereo till 40 Hz.

For the vinyl two sub is a good option.

I can't agree with that recommendation Jens.
What Mix and Mastering Engineers do is a particular thing, it should not be diluted by speaker management issues.
Due to equipment restrictions many early stereo recordings have a Bass or whole Drum kit on on side. Kick drums in particular are quite directional.
Vinyl aside, since Digital audio became common, many of us were delighted with the freedom to place very substantial LF to one side or the other.
So to hear any of this as mixed one requires L and R speakers to be full range.
DD

I think you completely missed my point.

What I´m saying is that if you sum low frequencies to mono; it will sound better in most cases since you put less stress on woofers (since the energy is distributed to two sets of woofers instead of one, so half the cone travel is needed to displace the same volume of air = less distortion, regardless of speaker used) assuming the system contains of two subs (L&R) or full range speakers. If not; nothing will change, so there’s no downside to this. Again, since you cannot localize low frequencies anyway, there’s no drawback, only positives (again assuming two channel low frequency system) in terms of lower distortion and potentially better modal behavior.

The mastering engineer’s job is to make sure that the recording sounds as good as it possibly can in as many systems as possible. What I describe would result in just that. There´s a reason why some engineers actually do this already …

But you are as always naturally entitled to a personal opinion.

In an idealised hypothetical situation LF should be omni directional. But real situations are sometimes more complex. e.g. A Kick drum has directional aspects, even in the LF. It is not a symmetrical wavefront. I believe this helps make it obvious where it is at. Also a single LF source will excite a room in a very different way to more sources.
Furthermore much of our audience listens on headphones. LF is very definitely localised.
Mastering Engineers may well be turning some of our efforts into Mono. Indeed Vinyl demands such. But then Vinyl has always been simply awful at LF.
As a Mix engineer I would not be happy to find an ME interfering with my intentions. As an ME I would never do so at least without serious consultation.
I do find a bizarre world of opposites sometimes though.
I find myself using Widening techniques when Mastering. Almost invariably it turns out that the Mix Engineer has been relying extensively on headphones.
So how would one be faithful to his/her intentions there....LOL
DD

I'm not sure so someone please correct me if I'm wrong, but I suppose at one position it would be possible (SBIR that counteracts a modal bump will probably not be minimum phase, for example) but I don't think it's possible in many positions in the room at once.

So yeah, in any practical case it seems very unlikely.

This seems to be an ongoing debate although scientists/engineers have repeatedly asserted that reproduction of low frequencies in a room are "minimum phase" so that controlling the frequency response directly impacts decay time.

I know that our testing with multiple subs has verified this (see graphs above inthe thread) but it seems to still be debated.

The "debate" isn't so much about whether low frequency anomalies are minimum phase -- there's not much to debate, since you can use REW or any modern analyzer and see exactly which parts of the frequency response are minimum phase and which partsaren't. (You will find that pure room resonances are minimum phase -- but often times you aren't dealing with a pure room resonance)

The "debate" I believe you are referring to is mostly about whether or not you can successfully use EQ to correct non-minimum phase issues for more than a single point in the room. And as far as I can tell, all actual engineers agree that when using EQ to correct the non-minimum phase region you will degrade the frequency response at other points in the room.

So actually there is no scientific or engineering debate about this point -- only misleading advertising from people trying to sell you digital room correction.

You bring up kick drums. A low frequency is a low frequency. The directional perception of 50Hz doesn't change depending on whether it's part of a kick drum or part of a bass guitar. You may want to read a bit about the Fourier transform if you are unfamiliar with it. This link is a good introduction: Fourier Analysis and Synthesis.

As for non-symmetrical waves -- they necessarily contain even-order harmonics. This page mentions it in the section "Harmonic Distortion" but there are more full presentations of this elsewhere: Amplifiers.

Perception of 50Hz is a very idealised notion, which does not relate to the reality I speak of. A bass drum lights up the room, particularly modally, very differently depending on which direction it is pointing and where it is located. Furthermore a kick drum wave FRONT hitting the head and body, is clearly felt by the skin. If it coming from one side only, it is blatantly tactile.
Mike, I recommend you get a bass drum in a room and discover the disconnect between (mis) reading someone else's theory about sound, vs actually hearing and feeling it.
BOOM
DD

We are talking about how low frequency information is reproduced by speakers … and if you think that you can localize a 50 Hz tone produced by a speaker in a room (assuming it doesn’t also produce any perceivable energy higher up in the frequency range due to distortion, port noise etc); you´re wrong.

Drum Acoustics & Miking

"A drum's harmonic shell and non-harmonic drumhead are coupled into an instrument that radiates like an acoustic dipole."

"Low drum frequencies are the least directional, although experiments show them dipped by several dB at 90° to the main axis and also out the back (180°)."

Towards a generalized theory of low-frequency sound source localization

https://www.researchgate.net/publica...e_localization

Do note that I said: "if you think that you can localize a 50 Hz tone produced by a speaker in a room ..."

And I’m referring to a normal sized room, not large venues. It might be possible to localize low frequencies in free field (or rooms that are large compared to wavelength).


Also again; I must stress that when people say that they can easily localize low frequency sources; they most likely do so because the source is also producing harmonic content (due to distortion, even if low; it will still be enough to help localize the source), port noise or other kind of distortion (mechanical noise etc.) that will make it easy to locate the position of the source.

I forgot the +1 on my previous post.

It should be forbidden to speak about bass reproduction in a room and cable on an internet forum.

+1

Jens, I am stating that a Kick drum is directional. It activates a room differently depending on what direction it is pointing and it's location. The drum accepts feedback from the room, in terms of reflections and resonances. This greatly amplifies the stimulus effect.
Speakers and Headphones will represent this differently.
If panned to the Left, on Cans it will all be on the left. There is no central summation. If the ME Mono summed, the listener would not hear the Artists intent.

More importantly there is one very large element of the Kick drum sound which is VERY directional and will be entirely represented if the speaker is good enough. There is a very large unidirectional blast which goes one direction, away from the beater and skin. Not immediately in any case, depending on the whether the drummer 'buries' the beater or encourages it to bounce away immediately. This can seen in waveforms and is easily felt on the skin, or in the chest if facing it. One of the early Recording Engineers used to place the Mic where this particular wavefront was strong enough to blow out a candle.
If a speaker is full range and powerful enough it should reproduce this effect fully. Even if all HF is removed, the source of this WaveFront is obvious. The ear drum and skin will feel the blast. A candle will extinguish.

DD

And I´m simply wondering why this is relevant for this discussion about speakers and rooms ...

As soon as you record your drum; all directional information is lost. Even if you built a very strange speaker with the same frequency based polar pattern as the drum recorded; this very special speaker still can’t guess how this drum was radiating when recorded since the recorded signal only contains the information based on a single (or a very limited number of) position(s). So either way; how a kick drum radiates is completely uninteresting for this discussion.

I do not know what is possible in a room, especially a small room, though it is possible in electronics. There are often equivalents between electronic and mechanical systems, though it seems to require less brain sweat to engineer an electronic system than a mechanical/acoustic equivalent.

A proper allpass filter ought to be flat, but allpass filters can have very long decay time or ringing.

For instance one can make electronic reverb out of allpass filters, and make them ring until next tuesday if you want.

That is not true. The initial wavefront has a large forward only pressure. This travels outward in one direction. A blast. A good speaker will represent that. It is even visible in the cone movement. It is visible in the Recorded Waveform.
If this recording is panned to say the Left, it will be just that single speaker broadcasting this blast. This will be easily detected by both ear and skin at the left side of the body.

Jens you have advocated in this thread that ME's should Make Mono at LF. This would be contra the artists intentions, is audible and detectable by feel. Both the blast and panning are also audible and headphones. I am presenting the reasons why it is an erroneous suggestion. I would strongly suggest we do not do LF Management and burn it into the Mix or Master.
I agree it is not exactly on Topic, but as you introduced it, I thought it best to counter the notion.


DD

Right, but in the context of room acoustics in normal sized rooms and frequencies well below schröder, i don't think such extreme high order allpass behaviour can occur. Or can it?

Ehmm ... ok ...

So now you're talking about cone travel (positive or negative) as if it was the same as directivity ... ?

And no; unless a very special speaker (dipole or cardioid, not very common in a studio setting), a 50 Hz wave (half or full period doesn't matter) will not be directional at all.

If you want to continue to make such claims; please back it up by some valid papers.

Thanks Evosilica

Dunno. Digital allpass filters for EQ or crossover tasks can be made with very short delay times (one sample dalay) but with a low center frequency and "rather long" decay time. That would be a high Q allpass filter, a very resonant allpass.

For typical reverb usage one could use the same structure, but use a longer loop delay for longer decay time with "not as much" feedback. Lower Q (but still rather high).

I don't have deep understanding of the topic, but ringing allpass filters can satisfy the allpass requirement and still sound rather bad. Which is part of why it is easy to make an allpass filter but difficult to make a truly good sounding digital reverb out of that building block.

For instance a 15 foot room dimension would be a 13.3 ms delay element. A digital allpass has equal amounts of feedback and feedforward surrounding the delay. Similar situation with analog allpass so far as that goes. The feedback alone would make a comb filter, so far as I know the same mechanism which causes room modes.

If one could arrange (perhaps via dsp and/or multiple subs) a feedforward which can "undo" the comb filtering and restore the allpass characteristic, then perhaps one could devise an acoustic situation where the frequency response is fairly flat but the time response is maddeningly long.

That is why I earlier asked SteveGTR how much electronic tweaking he used along with the four subs. Apparently some tweaking was involved. If he had said "oh, we just hooked up the four subs and it was instantly flat" then it would have been discouraging that I hadn't tried that "the very first thing".

Steve said his results flattened the FR without increasing decay time. My crude 2 sub tests increased decay time along with un-doing my null. But my room would be fairly inconvenient to place four subs, even small ones. My room also is perhaps "too open" having two always open doorways in the rear wall. It is good for dumping some excess bass into the rest of the house, but OTOH results in rather long-delayed low level reflections re-entering the room from the rest of the house thru the same doorways.

It results in a "2 slope" impulse response with fairly fast initial decay, but a longer low-level impulse response tail. The room doesn't sound so bad for a low budget amateur effort, but the low level bass decay looks rather long on waterfalls. Maybe if the small room was "fully enclosed" the bass decay would look better and respond better to multiple subs. Dunno. In other words, in my room maybe there are not enough walls for the multiple subs to interact with.

I was wondering about this - how we can perceive direction of a soundfield at lower frequencies. Is it audible, is it perceivable as part of the tactile sensations.

So I checked.

What I found was that direction is not audible and does not affect tactile feel - it does not matter whether the wavefront hits you head-on or from the side.

This is a good thing to know, because it simplifies set-up of bass systems. If you have to get the direction of the soundfield right, in combination with the other properties of sound, it gets a lot more complicated.

What does matter, though, is the level of particle velocity and sound intensity relative to sound pressure.


Yes, the different instruments generate different soundfields, and direction is one of these properties of the soundfield.

But in a reproduction chain none of this directional information is present, it was all lost when the original instrument was recorded, as it was then transformed into a 2-dimensional signal with just amplitude changing over time.

The loudspeaker system will reproduce everything the same way - direction of the sound will be the same regardless of what instrument it is, a bass drum will be the same as a bass guitar.

Localisation of a bass sound source is determined by harmonics at higher frequencies.


Some will claim low bass has no directivity.

This is not true, but if the room is very small and the frequency is very low, and the room is perfectly sealed, the intensity and transfer of energy will be zero and there will be no direction as the particle velocity of the sound field will be zero.

Adding a bass system with multiple subwoofer units works very well for getting the bass right.

Examples with 4 subwoofer units is presented in this thread, with subjective impression and measurements to support the claims.

So, why is this so controversial, why do some say it does not work, while others report excellent results.

I believe the reason is problems with set-up and calibration.

There is focus on getting the frequency response right, but time is forgotten. The impulse response of the bass system in itself must be right, and then this must be integrated with the main speakers in a proper way.

The purpose of adding more units is to fill nulls. Then you can use eq to fix the frequency response.

It will not be perfect across the whole room, but often there will be acceptable response, and for a small studio where mostly one person is working I assume it is a significant improvement if the bass is smooth and impulse correct at this one location.


What can be achieved:

- Improved frequency response
- Improved impulse response
- Significant reduction of low frequency resonances
- Better response across the whole room

<~~~~~~THIS!!!!

Plus the fact that clarity and imaging in mids and highs are greatly improved once you deal with small room modal issues!

Hello,

Audiophile brand of speaker provide foams to sealing the port and convert the speaker or a sub in sealed version.
Is it as simple or are there others parameters to convert a ported speaker in a sealed speaker ?

Thank you

Not only foam, but hard non-porous material also.