• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

An Enticing Marketing Story, Theory Without Measurement?

UliBru

Active Member
Technical Expert
Joined
Jul 10, 2019
Messages
123
Likes
337
UliBru, did you really try to fix a 3-way loudspeaker's step response by FIR'ing the input signal? What is the point there? The step response you showed is typical for a 3-way.
I do not only try. I do it. :)
In this thread room correction is always complaint about by discussing frequency responses and targets.
But room correction is not just creating a straight frequency response from a measured room frequency response.
The correction works in time domain. And it can also correct the time behaviour. Many room correction systems apply a minimumphase correction only. But it can also correct the excessphase. A typical example for the latter is the excessphase introduced by speaker crossovers.
 
  • Like
Reactions: j_j

Juhazi

Major Contributor
Joined
Sep 15, 2018
Messages
1,717
Likes
2,897
Location
Finland
OK, how do you describe it sounds corrected vs. without?
 

UliBru

Active Member
Technical Expert
Joined
Jul 10, 2019
Messages
123
Likes
337
OK, how do you describe it sounds corrected vs. without?
1. different
2. definitely better :)
3. more details, more clarity, more transparency, more naturalness, more depth. Less boom, less annoyance, less fatigue. Of course a highly subjective feeling
 

Snarfie

Major Contributor
Forum Donor
Joined
Apr 30, 2018
Messages
1,169
Likes
926
Location
Netherlands
It looks like software REW, Mathaudio an porbaly others have different approches to calculate inefficiensies or how to correct them. Who can tell us what the best approche/software or specific differences are. Is their an objective methode to test this software an their theories.
 
Last edited:

Snarfie

Major Contributor
Forum Donor
Joined
Apr 30, 2018
Messages
1,169
Likes
926
Location
Netherlands
1. different
2. definitely better :)
3. more details, more clarity, more transparency, more naturalness, more depth. Less boom, less annoyance, less fatigue. Of course a highly subjective feeling
Did you ever used mathaudio room eq. I'm curious what your findings will be when you compare your current roomcorrection software with mathaudio or for that matter with other roomcorrection software. Mabey you already did. ;)
 
Last edited:

UliBru

Active Member
Technical Expert
Joined
Jul 10, 2019
Messages
123
Likes
337
It looks like software like REW, Mathaudio an porbaly others have different approches to calculate inefficiensies or how to correct them. Who can tell us what the best approche/software or specific differences are. Is their an objective methode to test this software an their theories.
1. yes, the approaches are different
2. only independent tests can tell about quality, vendors are biased :)
3. specific approaches or differences are usually not published
4. are there objective methods for comparing amplifiers, speakers, power supplis, cables, ... ?
 

Snarfie

Major Contributor
Forum Donor
Joined
Apr 30, 2018
Messages
1,169
Likes
926
Location
Netherlands
1. yes, the approaches are different
2. only independent tests can tell about quality, vendors are biased :)
3. specific approaches or differences are usually not published
4. are there objective methods for comparing amplifiers, speakers, power supplis, cables, ... ?
I'm more interested in asr forum members findings because they have more experience than avarage and probably less biassed as vendors. Just as Klaus Heinz from HEDD is testing his prototype loudspeakers to a knolegeble group of listners. He stated if the results are positive in this group i'm at a prommising direction.
 
Last edited:
OP
svart-hvitt

svart-hvitt

Major Contributor
Joined
Aug 31, 2017
Messages
2,375
Likes
1,253
1. yes, the approaches are different
2. only independent tests can tell about quality, vendors are biased :)
3. specific approaches or differences are usually not published
4. are there objective methods for comparing amplifiers, speakers, power supplis, cables, ... ?

@UliBru , you seem to be a man of resources (intellectual, know-how). Wouldn’t it be in your interest to team up with some of the universites in Germany to arrange tests of correction software packages?

Maybe start with MSc student dissertations, then follow up with PhD level articles on correction software.

Remember, the only study on this seems to be the Olive et al. study and that makes this area an unexplored one
 

Juhazi

Major Contributor
Joined
Sep 15, 2018
Messages
1,717
Likes
2,897
Location
Finland
1. different
2. definitely better :)
3. more details, more clarity, more transparency, more naturalness, more depth. Less boom, less annoyance, less fatigue. Of course a highly subjective feeling

OK good. I haven't tried timing correction. I can accept it as used based on direct response correction like this, not as room correction at listening spot. I have minidsp 4x10HD and it can't do FIR (rePhase). I tried it shortly with a 2x4HD unit, but didn't learn how to . This is something I plan to learn when I have more spare time.

A multiway speaker has naturally a step response with multiple knees, depending on it's xo topology. This is because the xo-slope creates delay and crossing drivers must have matching timing if we use symmetric LR xo topology (like me LR2 and LR4 acoustically). Another way to look at it is groupd delay. So basically higher frequencies are delayed to compensate the "slow" bass driver, to achieve "ideal step" and zero group delay.

The KiiAudio Three has option to use FIR-corrected step or not - correction adds more overall delay to signal https://www.stereophile.com/content/kii-audio-three-loudspeaker-measurements
 
Last edited:

Krunok

Major Contributor
Joined
Mar 25, 2018
Messages
4,600
Likes
3,065
Location
Zg, Cro
I do not only try. I do it. :)
In this thread room correction is always complaint about by discussing frequency responses and targets.
But room correction is not just creating a straight frequency response from a measured room frequency response.
The correction works in time domain. And it can also correct the time behaviour. Many room correction systems apply a minimumphase correction only. But it can also correct the excessphase. A typical example for the latter is the excessphase introduced by speaker crossovers.

I did correction mannually with rePhase based on measurements I made with REW. I made a correction in time domain including the correction for passive XO. In rePhase it is very easy to generate filters that will have only correction in time domain vs only correction in frequency domain. When I tried to listen with correction only in time domain I wasn't able to hear any obvious differences. When I tried to listen with correction only in frequency domain with time domain excluded the difference was obvious.

I feel good (and kind of proud) when I look at my phase and step response graphs which look way better then they were before correction but in reality it is the frequency response correction that mattered.
 

Michael Anderson

New Member
Joined
Jun 27, 2016
Messages
2
Likes
1
Question to Doctor Toole

From the top of my memory, I read somewhere perhaps from Earl Geddes that it takes up to 50 ms to properly perceive a 50 Hz tone. It this is the case that woudl suggest that in the bass say below 100 Hz we are dealing with a quasi-steady state phenomenon, thus Linearization in the Frequency domain with steady state approximation should work..
Right? Wrong?

It seems generally agreed that minimum phase anomalies below the transition zone can be successfully equalised .

Above the transition zone equalisation is it said to be inappropriate because , unlike a microphone, we can hear that the total sound is made up off sound arriving at different times and from different directions including directly from the loudspeaker . That is we can distinguish between the direct sound and reflected sounds .
Equalising the room response in total means we mess up the direct sound which is not what we want to do .

When we equalise below transition we alter the sound coming from the loudspeaker but it seems this isnt a problem because at these frequencies we can't separate the speakers direct sound from the speaker plus room sound.
If this is correct why is it so ?
I recall the same explanation from Geddes as FranzM posted above but it seems this isnt the correct explanation ?
Does this mean in effect that below transition there is no direct sound ? ( Or there is only direct sound ).

A reason some people give for not liking any room equalisation is that ,even at bass frequencies, it messes up the direct sound which they believe they can hear as seperate from the room sound . Does this view have any validity ?

Dr Tooles book seems to say that above transition we can listen through the room and hear the direct sound of the loudspeaker but below transition the speaker cant be separated from the room. I am trying to understand why would be so ?
If @Floyd Toole or anyone else can shed light on this that would be good.
Thanks
 

Cosmik

Major Contributor
Joined
Apr 24, 2016
Messages
3,075
Likes
2,180
Location
UK
It seems generally agreed that minimum phase anomalies below the transition zone can be successfully equalised .

Above the transition zone equalisation is it said to be inappropriate because , unlike a microphone, we can hear that the total sound is made up off sound arriving at different times and from different directions including directly from the loudspeaker . That is we can distinguish between the direct sound and reflected sounds .
Equalising the room response in total means we mess up the direct sound which is not what we want to do .

When we equalise below transition we alter the sound coming from the loudspeaker but it seems this isnt a problem because at these frequencies we can't separate the speakers direct sound from the speaker plus room sound.
If this is correct why is it so ?
I recall the same explanation from Geddes as FranzM posted above but it seems this isnt the correct explanation ?
Does this mean in effect that below transition there is no direct sound ? ( Or there is only direct sound ).

A reason some people give for not liking any room equalisation is that ,even at bass frequencies, it messes up the direct sound which they believe they can hear as seperate from the room sound . Does this view have any validity ?

Dr Tooles book seems to say that above transition we can listen through the room and hear the direct sound of the loudspeaker but below transition the speaker cant be separated from the room. I am trying to understand why would be so ?
If @Floyd Toole or anyone else can shed light on this that would be good.
Thanks
I'm one of those people who doesn't do any kind of room EQ except for the equivalent of baffle step compensation (because my speakers don't have neutral dispersion) and this is to some extent room-dependent.

It's philosophy again, I think:

If you take the view that the room destroys the fidelity of the signal then it makes sense that the effects of the room should be 'corrected' if possible. If it can be shown that at bass frequencies this correction unambiguously works (in the sense that the listener's ears receive something the same as the recorded signal) then that seems to be a case for correction. But if it's not so unambiguous at higher frequencies then maybe the would-be room corrector backs off a little from that position. This is not the same as saying that correction at higher frequencies wouldn't be desirable; rather it's resignation that unfortunately it's not possible.

I can see other 'philosophical' positions that might be equally valid. Maybe the room isn't destroying the fidelity of the signal at all; it's merely adding some 'ambience' and the listener is quite capable of distinguishing between direct and ambient sound. But what about the bass? Well, the room's ambience is also reflected in the bass: it is a complete system where the size and shape of the room affects the timing and frequency magnitudes at all frequencies. If the human listener has evolved to interpret the sound of an approaching woolly mammoth in a forest - or simply learned to interpret sounds and physical sensations through 50 years of living in rooms - then 'correcting' the bass separately may signal that something is 'odd'. The idea of two separate bands processed differently immediately looks dodgy to me. So I live with supposedly uncorrected bass. Somehow I manage to struggle through.
 

Juhazi

Major Contributor
Joined
Sep 15, 2018
Messages
1,717
Likes
2,897
Location
Finland
Group delay of xo filter, measured in nearfield (direct sound) is minimal in magnitude (few ms for LR4 at 80Hz) and very smooth compared to GD whoopies that we see at spot measurements! If one wants to smooth the spot timing (excess GD), you must do tens of ms changes and this for sure makes direct sound or sound at another spot strange. Usually one takes several measurements and uses averaging and smoothing for FIR timing, whic alleviates this problem.
https://www.roomeqwizard.com/help/help_en-GB/html/graph_groupdelay.html
https://www.gearslutz.com/board/stu...velets-amp-group-delay-sub-alignment-rew.html

As said, I am happy without FIR and I find LR2 slopes better (more natural sound of piano, voice, etc.) than LR4 with my project speakers. Still, LR4 slopes give smoother off-axis response!
 
Last edited:

Floyd Toole

Senior Member
Audio Luminary
Technical Expert
Industry Insider
Forum Donor
Joined
Mar 12, 2018
Messages
367
Likes
3,893
It seems generally agreed that minimum phase anomalies below the transition zone can be successfully equalised .

Above the transition zone equalisation is it said to be inappropriate because , unlike a microphone, we can hear that the total sound is made up off sound arriving at different times and from different directions including directly from the loudspeaker . That is we can distinguish between the direct sound and reflected sounds .
Equalising the room response in total means we mess up the direct sound which is not what we want to do .

When we equalise below transition we alter the sound coming from the loudspeaker but it seems this isnt a problem because at these frequencies we can't separate the speakers direct sound from the speaker plus room sound.
If this is correct why is it so ?
I recall the same explanation from Geddes as FranzM posted above but it seems this isnt the correct explanation ?
Does this mean in effect that below transition there is no direct sound ? ( Or there is only direct sound ).

A reason some people give for not liking any room equalisation is that ,even at bass frequencies, it messes up the direct sound which they believe they can hear as seperate from the room sound . Does this view have any validity ?

Dr Tooles book seems to say that above transition we can listen through the room and hear the direct sound of the loudspeaker but below transition the speaker cant be separated from the room. I am trying to understand why would be so ?
If @Floyd Toole or anyone else can shed light on this that would be good.
Thanks

I'll keep this short because you have read my book. Todd Welti's papers have additional details if you want to dig deeper. (Sub)woofers are omnidirectional, wavelengths are long, and domestic rooms are small, so it is not surprising that the direct sound from woofers is swamped by the reflected sound. At specific frequencies the reflected sound is dominated by standing waves - resonances - so at those frequencies the sound is essentially communicated through resonances. These exhibit both peaks in the frequency response and in the time domain, the anticipated build-up and decay depending on the Q. Attenuating the amplitude peaks with parametric filters and/or mode-cancelling with multiple subs also diminishes the time-domain issues because the resonances behave as minimum-phase phenomena. This is straightforward physics.

How does it sound? Chapter 8 lays out a sequence of methods to address room modes, including examples from my own systems. Modes needed to be addressed because bass booms are intolerable. Each method, for certain situations, provided relief. My present system uses four subs in a sound-field-managed configuration. There are no "booms". Bass is exceptionally "tight", and there is no audible evidence of being in a small room in what is heard at low frequencies - no measurable or audible resonances.

Section 8.3 explains the audibility of spectral bumps vs. temporal ringing. There is much evidence that humans are very tolerant of time-domain misbehavior, but pay close attention to spectral effects.
 

Hipper

Addicted to Fun and Learning
Joined
Jun 16, 2019
Messages
753
Likes
625
Location
Herts., England
A reason some people give for not liking any room equalisation is that ,even at bass frequencies, it messes up the direct sound which they believe they can hear as separate from the room sound . Does this view have any validity ?

I use EQ and have not noticed this effect. I've always found the benefits far outway any possible defects, and in fact I don't hear any defects (I understand there may be issues at higher frequencies of course but we're not discussing that here).

I've never heard of this as a reason for not using EQ. Most talk about additional 'noise', not wanting to mess with the artist's intentions, or more technically, non minimum phase issues.
 

Cosmik

Major Contributor
Joined
Apr 24, 2016
Messages
3,075
Likes
2,180
Location
UK
"Bass booms"

I don't have any noticeable horribleness from the bass ends of my systems (stereo 8" sealed 3-way in fairly large monkey coffins, and 12" 3-way in very large boxes). Not huge rooms, although the 12"-ers' room has a high ceiling. Both systems are linear phase and time-aligned etc.

One thing that occurs to me is that a sharp cutoff at any audible frequency might begin to sound like a one-note boom - as the brain scrabbles for the frequency content it knows should be there. Bass reflex speakers that cut off at an audible frequency probably resemble that description. Sealed boxes, on the other hand, don't cut off as quickly, so possibly avoiding that problem.
 

Floyd Toole

Senior Member
Audio Luminary
Technical Expert
Industry Insider
Forum Donor
Joined
Mar 12, 2018
Messages
367
Likes
3,893
"Bass booms"

I don't have any noticeable horribleness from the bass ends of my systems (stereo 8" sealed 3-way in fairly large monkey coffins, and 12" 3-way in very large boxes). Not huge rooms, although the 12"-ers' room has a high ceiling. Both systems are linear phase and time-aligned etc.

One thing that occurs to me is that a sharp cutoff at any audible frequency might begin to sound like a one-note boom - as the brain scrabbles for the frequency content it knows should be there. Bass reflex speakers that cut off at an audible frequency probably resemble that description. Sealed boxes, on the other hand, don't cut off as quickly, so possibly avoiding that problem.

There is truth in what you say. Our attention is drawn to any discontinuity in frequency response, bumps, dips (to a lesser extent), and abruptly initiated tilts and bandwidth limitations. My system extends below 20 Hz and I do believe that is a noticeable reality, even though there is precious little down there.
 

Juhazi

Major Contributor
Joined
Sep 15, 2018
Messages
1,717
Likes
2,897
Location
Finland
Sounds very familiar to me. In '70s I listened closed box 3-ways in a wooden house at my parents, with great satisfaction. When I moved to own apartment in '80s I had 3-w br and later 2-w and br sub(s), but was never happy. My subs were smallish and when I measured them in -14 I realized that they couldn't go much below 40Hz. So, 35 years of my life was ruined with inadequate bass!

Then I decided to build only sealed woofers and to use Minidsp and now I can enjoy music at home again! Really low tuned br or TL woofers and subs might be ok too, but they are surprisingly rare - and big. My guess is that this might be one reason for the popularity of B&W floorstanders. It is not about SPL but the lowest notes and ambience effect.
 

Michael Anderson

New Member
Joined
Jun 27, 2016
Messages
2
Likes
1
.....(Sub)woofers are omnidirectional, wavelengths are long, and domestic rooms are small, so it is not surprising that the direct sound from woofers is swamped by the reflected sound. At specific frequencies the reflected sound is dominated by standing waves - resonances - so at those frequencies the sound is essentially communicated through resonances.....

How does it sound?

Dr Toole, Thanks for your reply. I currently use two subs and EQ to improve bass response. To me it sounds much better than without these measures.
What I am trying to understand is why it is OK to EQ below transition and not above. The argument about why not to EQ above transition is clear and credible but it seems the argument thats its OK to EQ below transition rests on two things :
1. At bass frequencies although like at higher frequencies there is direct sound and room sound the direct sound is overwhelmed by room sound due to the nearly all of the sound energy arriving after reflection from room boundaries due to omni directional radiation at these frequencies. This effect is multiplied at frequencies emphasised by room resonances .
We dont notice that EQ has modified , possibly quite a lot, the direct sound .
2. Although EQ and mutltiple subs isnt perfect in a practical sense its the best weve got.

I have no trouble accepting 2 but I think there is likely a lot more to 1 than I am understanding.
 
Top Bottom