isn't near corner placement best in that regard? of exciting as many modes as possible?
ps.
here are two room-speaker interaction frequency curves (room augmentation curves) according to Roy Allison's spreadsheet
- the first is for typical audiophile standmount speaker placement ca 1m away from both walls with acoustic center at the height of ca 75 cm
- the second is for the flooder in near corner placement
of course an ideal omnidirectional point source behaviour is assumed which is perhaps reality up to ca 500 Hz?
here are two room-speaker interaction frequency curves (room augmentation curves) according to Roy Allison's spreadsheet
- the first is for typical audiophile standmount speaker placement ca 1m away from both walls with acoustic center at the height of ca 75 cm
- the second is for the flooder in near corner placement
of course an ideal omnidirectional point source behaviour is assumed which is perhaps reality up to ca 500 Hz?
Attachments
If there was a single source only, the corner would be an
interesting place ...
There are some people who prefer corner placement for
subwoofers.
I cannot share that preference from my own experience.
Would not be the first reflections rather strong and
symmetrical ?
Maybe not preferable to listen on a diagonal of the room.
The notch in the first plot seems typical for comb filtering,
it can be reduced significantly with different distances to
the rear and side walls.
Playing with incommensurable distances to the walls helps ...
interesting place ...
There are some people who prefer corner placement for
subwoofers.
I cannot share that preference from my own experience.
Would not be the first reflections rather strong and
symmetrical ?
Maybe not preferable to listen on a diagonal of the room.
The notch in the first plot seems typical for comb filtering,
it can be reduced significantly with different distances to
the rear and side walls.
Playing with incommensurable distances to the walls helps ...
Last edited:
Corner placement excits all modes.
Directivity has no meaning or validity in the modal region.
The Schroeder frequency, Fs, tends to be fairly low when compared to the hearing bandwidth. Since the modal density increases with frequency, "modal density" is only relavent at and below Fs.
One cannot talk about sound in a small room below Fs in the same way as above Fs. Directivity has no relavence below Fs, transients have no relavenance below Fs, nor do reflections. But these things are all extremely important above Fs. We tend to hear sounds below Fs in steady state and above Fs in gated transient. You simply cannot talk about or even use the same concepts above and below Fs.
Directivity has no meaning or validity in the modal region.
The Schroeder frequency, Fs, tends to be fairly low when compared to the hearing bandwidth. Since the modal density increases with frequency, "modal density" is only relavent at and below Fs.
One cannot talk about sound in a small room below Fs in the same way as above Fs. Directivity has no relavence below Fs, transients have no relavenance below Fs, nor do reflections. But these things are all extremely important above Fs. We tend to hear sounds below Fs in steady state and above Fs in gated transient. You simply cannot talk about or even use the same concepts above and below Fs.
yes but it quickly appears that genuinely significant reduction of this effect demands really significantly different distances - like 100 to 160 cm
and when we want to keep the speaker at least 100 cm away from the nearest wall then the speaker lands quickly in the middle of the room - not much place for any normal stereo basis, not to mention WAF and other real life concerns
in this traditional away-from-the-walls placement better results gives 3-way speaker with woofer just above the floor, even when equidistant to both walls like 100 cm - 100 cm
A small amount of Basotect or felt like in the Carlsson will solve the problem. The bigger problem is that the corner loading only works up to a certain frequency. Simulate the corner in Hornresp as a conical horn plus eigth-space option and you will see. The solution I proposed earlier in the thread would be a very small bass driver and a very low-crossed horn tweeter.
Does that mean the Q of a closed box e.g. beeing
irrelevant for behaviour in time domain, when placed
in an acoustical small room ?
Do we all waste our time in finding alignments for
BR speakers e.g. which seem to be good compromise
between time - and frequency domain behaviour ?
Last edited:
Hi,
This spreadsheet forms ONE of the inputs into my own "XLBox" spreadsheet.
It is sadly by itself next to useless.
Further it only accounts for three walls in open space, it does not account for ceiling and other walls, meaning it actually undercounts the amplification of low frequencies.
Further it omits to account for the standing wave patterns and the speakers directivity.
In it's heyday it was a very useful tool, especially if one understands the implications from it's output AND it's limitations.
For example the floor reflection dip often usefully meshes with the floor/ceiling resonance, careful hight adjustment can play both off against each other.
And unless the speaker dumped on the floor in the corner was severely deficient in the bass response (like many of the Carlson/Sonab designs using the Philips 9710) it would be unlikely to produce a balanced sound.
As for corner placement, using pressure mode transducers (that is boxes that radiate low Frequencies omnidirectional) will excite room modes the least if they are placed in a pressure minimum (read room center) and most if they are placed in a pressure maxima (read walls).
In fact at least in part the old audiophile adage that "Speakers reliably sound best where they do the most visual damage to a room" comes from that effect.
By comparison velocity transducers (dipoles) excite room modes least near velocity minima, that is near wall and gradient transducers (cardiodes) fall in-between.
As for a tool that is at least reasonably useful for estimating speaker/room interactions, placement and accounts to a degree for off axis response is here:
hunecke.de | Loudspeakers Calculator
I tend to use this web-page every single time I am doing acoustic trouble shooting or am looking at ways to set up unknown acoustic spaces with traditional speakers. Again, it has imitations but helps me normally into the ballpark quite rapidly.
Ciao T
This spreadsheet forms ONE of the inputs into my own "XLBox" spreadsheet.
It is sadly by itself next to useless.
Further it only accounts for three walls in open space, it does not account for ceiling and other walls, meaning it actually undercounts the amplification of low frequencies.
Further it omits to account for the standing wave patterns and the speakers directivity.
In it's heyday it was a very useful tool, especially if one understands the implications from it's output AND it's limitations.
For example the floor reflection dip often usefully meshes with the floor/ceiling resonance, careful hight adjustment can play both off against each other.
And unless the speaker dumped on the floor in the corner was severely deficient in the bass response (like many of the Carlson/Sonab designs using the Philips 9710) it would be unlikely to produce a balanced sound.
As for corner placement, using pressure mode transducers (that is boxes that radiate low Frequencies omnidirectional) will excite room modes the least if they are placed in a pressure minimum (read room center) and most if they are placed in a pressure maxima (read walls).
In fact at least in part the old audiophile adage that "Speakers reliably sound best where they do the most visual damage to a room" comes from that effect.
By comparison velocity transducers (dipoles) excite room modes least near velocity minima, that is near wall and gradient transducers (cardiodes) fall in-between.
As for a tool that is at least reasonably useful for estimating speaker/room interactions, placement and accounts to a degree for off axis response is here:
hunecke.de | Loudspeakers Calculator
I tend to use this web-page every single time I am doing acoustic trouble shooting or am looking at ways to set up unknown acoustic spaces with traditional speakers. Again, it has imitations but helps me normally into the ballpark quite rapidly.
Ciao T
Dear Mr. Geddes,
Directivity as such does not, that is true. What does however is the nature of the transducer with respect to how it exited the sound waves.
Pressure mode transducers (sealed/vented boxes) indeed behave as you suggest.
Yet gradient (cardiode) or velocity (figure 8/dipole) transducers are a different story.
So while directivity itself has no meaning in the modal region, the way it is attained most certainly does.
Ciao T
Directivity as such does not, that is true. What does however is the nature of the transducer with respect to how it exited the sound waves.
Pressure mode transducers (sealed/vented boxes) indeed behave as you suggest.
Yet gradient (cardiode) or velocity (figure 8/dipole) transducers are a different story.
So while directivity itself has no meaning in the modal region, the way it is attained most certainly does.
Ciao T
Hi,
Pretty much so, yes. BUT the Q has impact in the amplitude domain and as a result very much impact the overall percieved sound. And a Qt = 0.5 closed box is still a good choice in an acoustically small room and tends to provide a bass that is perceived as more accurate.
Not really.
Because while intending to work on a problem that you cannot fix because it does not actually exist in the way it is stated, you inadvertently address a different problem that is very real and is improved with such "compromises" over so-called "maximum flat" alignments.
Ciao T
Pretty much so, yes. BUT the Q has impact in the amplitude domain and as a result very much impact the overall percieved sound. And a Qt = 0.5 closed box is still a good choice in an acoustically small room and tends to provide a bass that is perceived as more accurate.
Not really.
Because while intending to work on a problem that you cannot fix because it does not actually exist in the way it is stated, you inadvertently address a different problem that is very real and is improved with such "compromises" over so-called "maximum flat" alignments.
Ciao T
or we can choose suitable driver - woofer, midwoofer or full range - that gently rolls off below 200 Hz ca 6 dB per octave - just to match the room loading
best,
graaf
So you think it is all about compensation of room gain and
flat in room response ...
That would be an explanation why with open baffles a high Q system
can be tolerated very well, as long as open baffle rolloff and room
gain sum up to a flat frequency response of the
"speaker and room" system as a whole.
So we do not hear the bass "hanging over" in time domain but
in frequency domain solely if the in room response is bass heavy with
a high Q system ...
I will have to think about.
yes indeed, and these are actually good news
thank You so much for the link
best regards,
graaf
I don't see you disagree with anything that I said.
There are only two basic types of excitation, pressure and pressure gradient, or velocity. All sources are composed of a various mixture of these two fundamental types. But if you think of a velocity type (pressure gardient) as simply being two out of phase pressure sources seperated by some distance in space then there really is only the one type. In modeling rooms we only ever use the one type of source and create everything out of those.
How each source then excites each mode is the issue and finally we sum over all the sources. In doing this we never see a "directional" pattern only a modal pattern. Since any discrete mode does not allow an arbitray wave direction, (it is fixed in space) as long as there are only a finite number of modes being excited then the wave propagation direction is not arbitrary either, but fixed by the modes being excited. Hence, there is no directivity until the modal density is sufficient to allow for an arbitary wave direction in space.
Of course, I presume that this is what you meant .
The speakers resonances is just another resonance in the room and is no more nor no less important than any other resonance. Bass "hanging over" is probably a good thing if it is broadband, but certainly a bad thing if it is modal. Large rooms have very large "bass hangover" and the bass is usually quite good - but its broadband since the bass is well above the modal range in a large room.
It would be impossible to hear bass in the time domain in the same way that we hear HF. A 100 Hz. tone take several ms. for the brain to even register it - basicaly a several cycles. The ear is steady state by this time. Clearly below 100 Hz and everything is steady state.
Flat in room LF response spatially averaged is the key, and only multiple LF sources can do this in a small room. The Q, etc. of any individual source then becomes totally insignificant.
Hi,
Not quite as simple as this. The whole way speakers and room interact may be described as 'slightly tricky and difficult' if one has lived a long time in England, where extreme understatement is raised to an artform...
To a degree at least.
If the open baffle (dipole at low frequencies) is placed fairly close near walls (take care to absorb the midrange with some 4" thick basotect beheind though), it will not excite room modes very well. If we then couple the listner well to the air velocity (so don't sit him near walls) things come together nicely.
One may notice that the old Audio Physics setup conventional speakers is the exact opposite of this and tends to work about equally well...
In room we need to consider in the modal region how the transducers are coupled to the modes AND how the listener is coupled to them.
This certainly does not do any harm... ;-)
It is a major contributing factor, IMNSHO.
Just measure a "maximum flat" aligned bass reflex speaker (preferably also with full baffle step compensation) in a normal room with a "common" near wall placement from the listening position. Take care NOT to use pseudo-anechoic measurements (you can always window afterwards).
While you are at it, also measure RT20/40/60.
Then measure the Audio Physics setup of the same speaker.
Then place a "maximum flat" open baffle near side and rear walls (absorb the rear wave higher up) and place the listening position almost directly in the room center. Now measure again. What I find usually most interesting is the set of curves or RT20/40/60 against frequency.
In some ways I find these tell more about what a speaker sounds like (in the specific room) than even the on/off axis FR Curves and the Step/Impulse response.
In the end what matters is to enjoy the music though. And I have friends who enjoy theirs using a pair of Bose 901's.
My votes for speakers I enjoyed most in living (acoustically small) rooms go to corner loaded Tannoy Horns (which are actually pretty flat when placed thusly) and the JBL Everest and the relatives with the asymmetric horns and the MEG RL-901K and of course my own designs... ;-)
Ciao T
Not quite as simple as this. The whole way speakers and room interact may be described as 'slightly tricky and difficult' if one has lived a long time in England, where extreme understatement is raised to an artform...
To a degree at least.
If the open baffle (dipole at low frequencies) is placed fairly close near walls (take care to absorb the midrange with some 4" thick basotect beheind though), it will not excite room modes very well. If we then couple the listner well to the air velocity (so don't sit him near walls) things come together nicely.
One may notice that the old Audio Physics setup conventional speakers is the exact opposite of this and tends to work about equally well...
In room we need to consider in the modal region how the transducers are coupled to the modes AND how the listener is coupled to them.
This certainly does not do any harm... ;-)
It is a major contributing factor, IMNSHO.
Just measure a "maximum flat" aligned bass reflex speaker (preferably also with full baffle step compensation) in a normal room with a "common" near wall placement from the listening position. Take care NOT to use pseudo-anechoic measurements (you can always window afterwards).
While you are at it, also measure RT20/40/60.
Then measure the Audio Physics setup of the same speaker.
Then place a "maximum flat" open baffle near side and rear walls (absorb the rear wave higher up) and place the listening position almost directly in the room center. Now measure again. What I find usually most interesting is the set of curves or RT20/40/60 against frequency.
In some ways I find these tell more about what a speaker sounds like (in the specific room) than even the on/off axis FR Curves and the Step/Impulse response.
In the end what matters is to enjoy the music though. And I have friends who enjoy theirs using a pair of Bose 901's.
My votes for speakers I enjoyed most in living (acoustically small) rooms go to corner loaded Tannoy Horns (which are actually pretty flat when placed thusly) and the JBL Everest and the relatives with the asymmetric horns and the MEG RL-901K and of course my own designs... ;-)
Ciao T
I mean the case of near the corner placement which is illustrated in my previous post with a graph