Dirac and similar

[Thomas_A]

All of this is known and extensively covered in Dr. Toole's book and for that matter, articles I have written. The trick is using multiple subwoofers to get smooth coverage across multiple seats: https://www.audiosciencereview.com/forum/index.php?threads/subwoofer-low-frequency-optimization.15/

For 2-channel listening, that is not as important since it is mostly a solitary situation with a single listener. You can perform spatial averaging. The results won't be perfect for one listener but would be very good for a few. You can see that evaluated in Dr. Olive et. al paper, The Subjective and Objective Evaluation of Room Correction Products

View attachment 32776

RC1 and RC2 are both Harman EQ with the former being optimized for one location instead of multiple.

The paper shows the efficacy of DSP correction above No EQ so your general thesis that these systems don't work does not hold.

For clarification, are you referring to resonator modes only here? I don't think Tangband denies that these should be EQed, they should.

 

[amirm]

For clarification, are you referring to resonator modes only here? I don't think Tangband denies that these should be EQed, they should.

You can't EQ them without measurements. Yet he says you shouldn't measure.

What goes on above a few hundred hertz is quite a bit more complicated than what he is talking about as well.

 

[Tangband]

I think Tangband is talking about reflector vs resonator modes. For example, placing a loudspeaker at different distances from the back wall will cause destructive/constructive interference, and placing the speaker closer to wall will place these interferences higher in frequency (+ you will get wall reinforcement). Placing a panel on the wall behind the speaker will reduce these interferences. The fundamental room modes will still be there and will most easily be dealt with using parametric EQ.

View attachment 32775

Yes, Im talking about reflector vs resonator modes. Maybe I was lost in translation ? I hope everyone understands now.

 

[Tangband]

What "normal" room? Most people in this day and age have open floor plans. As such, none of the rules of thumbs for rectangular rooms apply.

Even in rectangular rooms, the acoustic dimensions are not the same as physical dimensions since the boundaries are not usually hard.

For these reasons, measurements are absolutely essential to know where the room modes and resonances are.

I live in sweden- we have much smaller rooms than you have in general. And the walls are often made of concrete. You are completely right about that rooms can be very different in shapes. And that you need to do measurements if the room isnt rectangular.

 

[Tangband]

You can't EQ them without measurements. Yet he says you shouldn't measure.

What goes on above a few hundred hertz is quite a bit more complicated than what he is talking about as well.

For the frequencys between about 80-800 Hz ( depending on the size of your loudspeakers ) , it is usually more useful to play with the bafflestep correction instead in a dsp, to get the proper balance from the listening position. Using two loudspeakers when you listen. And you can put damping material ( made of 10-15 cm rockwool ) behind the loudspeakers to get rid of destructive reflections coming from the loudspeaker and bouncing on the wall.

This was only a brief visit. Im leaving now, and I really enjoy reading about all your reviews. This has become one of the most interesting internet places about hifi . Keep up the good work.

 

[Thomas_A]

You can't EQ them without measurements. Yet he says you shouldn't measure.

What goes on above a few hundred hertz is quite a bit more complicated than what he is talking about as well.

Well,I agree that you need to measure to investigate the response. But I would not EQ reflector effects, only the room modes. It is quite clear in my measurements what is a room mode and what is reflector modes.

 

[Krunok]

For the frequencys between about 80-800 Hz ( depending on the size of your loudspeakers ) , it is usually more useful to play with the bafflestep correction instead in a dsp, to get the proper balance from the listening position.

You can do absolutely the same thing with DSP what you do with baffle step correction circuit and that is regulary done with active crossovers.
Another thing is that I simply can't see any of us correcting the work of professional speaker designer because we lack the knowledge. Unless of course you have the knowledge to correct Toole's work, like you do.

 

[Krunok]

For clarification, are you referring to resonator modes only here? I don't think Tangband denies that these should be EQed, they should.

What Tangband fails to understand is that resonator modes don't magically stop at 80Hz. If you have a mode at 73Hz you will also have a 2nd mode at 146Hz, 3rd mode at 219Hz and 4th mode at 292Hz. After Schroeder frequency they will start to fade as room response is no longer dominated by resonances.

What he also fails to understand is that Schroeder frequency of his small room is likely around 300Hz, not at 80Hz.

Instead of trying to correct Toole I suggest this reading from Wiki to him:

Zones

The way that sound behaves in a room can be broken up into roughly four different frequency zones:

• The first zone is below the frequency that has a wavelength of twice the longest length of the room. In this zone, sound behaves very much like changes in static air pressure.
• Above that zone, until wavelengths are comparable to the dimensions of the room, room resonances dominate. This transition frequency is popularly known as the Schroeder frequency, or the cross-over frequency and it differentiates the low frequencies which creates standing waves within small rooms from the mid and high frequencies.
• The third region which extends approximately 2 octaves is a transition to the fourth zone.
• In the fourth zone, sounds behave like rays of light bouncing around the room.

Natural modes

The sound wave has reflections at the walls, floor and ceiling of the room. The incident wave then has interference with the reflected one. This action creates standing waves that generate nodes and high pressure zones.

But I would not EQ reflector effects, only the room modes. It is quite clear in my measurements what is a room mode and what is reflector modes.

What phenomena do you call "reflector effect"?

 

[digicidal]

What phenomena do you call "reflector effect"?

Seriously though, I believe he was referring to what most people would call "boundry reinforcement" - except for those lucky enough to have speakers than can naturally dig low enough without subs. But then again... it's a little hard to figure out just what he's specifically describing much of the time.

My take as a secondary translation would be "I would not apply EQ to primary reflections, instead using a BSC network to attempt to fix them at the speaker, and leaving the DSP to handle only significant room modes" - I can see the argument to some extent... but it's still screwy to me. Your speaker likely has many elements that were a compromise by the designer... once you start messing with those, then you're gonna need to redesign the whole crossover network possibly... what's next, building a new cabinet entirely, replacing all the drivers?

It seems like this is more of a "in a world where unlimited time, resources, and knowledge exist... do this" - which is all well and good, but the real world is one where most of us have a crappy room and somewhat crappy speakers, a limited budget and time... and just want the music to sound a little better if possible! (Which is what room correction software is designed for).

 

[Krunok]

View attachment 32796

Seriously though, I believe he was referring to what most people would call "boundry reinforcement" - except for those lucky enough to have speakers than can naturally dig low enough without subs. But then again... it's a little hard to figure out just what he's specifically describing much of the time.

My take as a secondary translation would be "I would not apply EQ to primary reflections, instead using a BSC network to attempt to fix them at the speaker, and leaving the DSP to handle only significant room modes" - I can see the argument to some extent... but it's still screwy to me. Your speaker likely has many elements that were a compromise by the designer... once you start messing with those, then you're gonna need to redesign the whole crossover network possibly... what's next, building a new cabinet entirely, replacing all the drivers?

It seems like this is more of a "in a world where unlimited time, resources, and knowledge exist... do this" - which is all well and good, but the real world is one where most of us have a crappy room and somewhat crappy speakers, a limited budget and time... and just want the music to sound a little better if possible! (Which is what room correction software is designed for).

But again, whatever passive BSC network can do DSP can do even better. In both cases you only need measurements precise enough to know what you are doing. But frankly, he didn't demonstrate nearly enough competency to cover his arogance in attempt to correcting Toole's work. In fact, he didn't demonstrate any competence at all.

Btw, my Harlech's are transmission line design which has "infinite" baffle so nothing much to correct there baffle related.

 

[digicidal]

But again, whatever passive BSC network can do DSP can do even better. In both cases you only need measurements precise enough to know what you are doing. But frankly, he didn't demonstrate nearly enough competency to cover his arogance in attempt to correcting Toole's work. In fact, he didn't demonstrate any competence at all.

Btw, my Harlech's are transmission line design which has "infinite" baffle so nothing much to correct there baffle related.

Exactly. I could be wrong but I think in his ideal world you would actually do almost everything at the speaker and use DSP to only eliminate room modes which were impossible to overcome at the speaker.

That's great if you're an accomplished builder making a pair of DIY darlings for a single room - but if you ever want to sell them (to anyone other than someone buying your house that is) you'd better be asking $250K/pair... because you'll have to redesign the whole thing for every single customer - as well as sell them on the concept of "you can't ever move these out of this room... and moving your primary LP is also forbidden". Good luck with marketing that one... although since $20K cables exist... there is a market I suppose.

Probably a bit better to design a flat (anechoic at least) speaker and let DSP do what it does pretty well (at least for us normal humans).

 

[sfdoddsy]

I’m confused. BSC obviously helps correct a speaker’s response due to baffle loss. But that has nothing to do with the room.

BSC is (usually) a shelving filter.

Room resonse errors are (usually) attended by narrow EQ.

You can certainly use DSP to do BSC, but it has nothing to with room EQ.

Or am I completely lost about the discussion.?

 

[Krunok]

I’m confused. BSC obviously helps correct a speaker’s response due to baffle loss. But that has nothing to do with the room.

BSC is (usually) a shelving filter.

Room resonse errors are (usually) attended by narrow EQ.

You can certainly use DSP to do BSC, but it has nothing to with room EQ.

Or am I completely lost about the discussion.?

You are not lost, he is.

 

[Krunok]

Exactly. I could be wrong but I think in his ideal world you would actually do almost everything at the speaker and use DSP to only eliminate room modes which were impossible to overcome at the speaker.

And that is a legitimate thought, but he's wrong in thinking he can do that correcting only for room modes up to 80Hz. He also suggested to do that based on calculations which in most cases doesn't work.

 

[Thomas_A]

What Tangband fails to understand is that resonator modes don't magically stop at 80Hz. If you have a mode at 73Hz you will also have a 2nd mode at 146Hz, 3rd mode at 219Hz and 4th mode at 292Hz. After Schroeder frequency they will start to fade as room response is no longer dominated by resonances.

What he also fails to understand is that Schroeder frequency of his small room is likely around 300Hz, not at 80Hz.

Instead of trying to correct Toole I suggest this reading from Wiki to him:

Zones

The way that sound behaves in a room can be broken up into roughly four different frequency zones:

• The first zone is below the frequency that has a wavelength of twice the longest length of the room. In this zone, sound behaves very much like changes in static air pressure.
• Above that zone, until wavelengths are comparable to the dimensions of the room, room resonances dominate. This transition frequency is popularly known as the Schroeder frequency, or the cross-over frequency and it differentiates the low frequencies which creates standing waves within small rooms from the mid and high frequencies.
• The third region which extends approximately 2 octaves is a transition to the fourth zone.
• In the fourth zone, sounds behave like rays of light bouncing around the room.

Natural modes

The sound wave has reflections at the walls, floor and ceiling of the room. The incident wave then has interference with the reflected one. This action creates standing waves that generate nodes and high pressure zones.



What phenomena do you call "reflector effect"?

The Speaker Boundary Interference Response or SBIR as depicted in my previous post. Or also that interference introduced at listener seat (by boundary interaction). I am sure this is very trivial knowledge here. I don't adjust for that in EQ, it is already fixed in speaker design and setup.

 

[digicidal]

And that is a legitimate thought, but he's wrong in thinking he can do that correcting only for room modes up to 80Hz. He also suggested to do that based on calculations which in most cases doesn't work.

You can have (although I'm not sure that's case here in this thread) a solution that's rational, repeatable, and successful - but is hyper-specific to the point that it's application is virtually null.

It's like finding a cure for baldness that's only effective in blind, albino males that are exactly 5 feet tall... sure for those 3 guys you've got a real winner on your hands. You've solved a problem... but you're not any closer to solving the problem. Or a cure for cancer that immediately kills the host after eradicating the cancer... it's still a repeatable solution - but you're not gonna have many volunteers in your test panel.

 

[Thomas_A]

It is quite easy to both calculate and measure the room mode that I previously fixed with a Behringer parametric EQ. It is one mode at 47 Hz that needs fix. I don't really need to touch anything else.

 

[Krunok]

It is quite easy to both calculate and measure the room mode that I previously fixed with a Behringer parametric EQ. It is one mode at 47 Hz that needs fix. I don't really need to touch anything else.

View attachment 32797

Can you pls post that graph with 1/12 smoothing?

Btw, it seems that panels you put behind your speakers didn't really prevent that dip at 80Hz.

 

[digicidal]

So that massive void at 80Hz is phase cancellation or ? Hard to see with that scale but it looks like nothing at all at that frequency?

 

[Krunok]

So that massive void at 80Hz is phase cancellation or ? Hard to see with that scale but it looks like nothing at all at that frequency?

It's probably cancellation between direct wave and wave reflected from the rear wall (SBIR).