• 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!

Bass arriving out of phase at the listening position

daftcombo

Major Contributor
Forum Donor
Joined
Feb 5, 2019
Messages
3,687
Likes
4,068
The only solution I could find was to work around the issue using EQ allpass filters to restore proper phase - so that, after EQ, the summed response is consistent with the response of left and right channels. That works, but obviously a lot of finesse is required in designing the EQ filters. Someone who only has access to "basic" PEQ or doesn't know how to manipulate phase using allpass filters would have no chance of getting this right. They would have to make a carnelian choice between having a correct response when individual channels are playing, or the correct response when both channels are playing. And that's assuming they even spot the problem in the first place. I really hope my case is some kind of isolated extreme scenario - if it's not, then I guess we'll have to revise standard calibration procedures!
Hi!

Do you hear an improvement with the allpass filters changing the phase only?
I've never been able to hear anything but worse sound when I was manipulating phase.

My experience in small rooms (I have my pair of 8030C in a 6m2 room) is that there is a lot of bass variation (I mean from 60Hz to 100Hz) if you stay still in your chair or bend 25cm forwards or backwards. Near the center of the room, you lose a lot of bass. Near the backwall, you have too much. My REW measurements support that.
 

SoundGuy

Active Member
Joined
Aug 13, 2022
Messages
115
Likes
120
@edechamps

Mucking around with phase is not advisable.

Are you trying to get a flat measurement at the exact microphone location OR are you trying to get the setup to sound good and balanced for listening ?

If the former then carry on with your crazy approach. If the later then you really need to take a completely different approach, as I described above - fix your setup before playing computer jockey with flattening the measurements artificially.
 

Juhazi

Major Contributor
Joined
Sep 15, 2018
Messages
1,717
Likes
2,897
Location
Finland
Sorry, I don't have patience to read through all posts, but two things come to my mind

- most likely it is vertical mode minimum. Try measuring at different heights
- use REW's Spectrogram subpage, here my room makes 50Hz max spl delayed 30ms!

Graphs below are from different speakers in different rooms!

ht awppslmini LR earbutt 500ms 112.jpg ainogneo83 2x4 conf1 R spotspectr.jpg
 

ernestcarl

Major Contributor
Joined
Sep 4, 2019
Messages
3,106
Likes
2,313
Location
Canada
I've never been able to hear anything but worse sound when I was manipulating phase.

"Manipulating" could mean anything so specifics would be illuminating... Though, I agree that it's very easy to make things worse by "mucking" about with phase EQ as SoundGuy already says -- I find it at least worth testing out just to hear the differences with my own two ears. Its interesting how different speakers react to phase "manipulation"... for instance, my ported speakers by themselves if solo'd / playing test tracks in mono sound somewhat worse when using a steep 96dB/oct linear phase HPF, whereas 24-48dB oct sounds fine. My sealed single driver Fostex 6301 monitor takes the steeper 96dB/oct filter perfectly okay, too. As for my previously mentioned 83Hz (2x stacked 2nd order) all pass filters applied to the front desktop mains, well, they do not sound worse to my ears at all, despite the visible increase in GD. I could also pull down excess phase around 50Hz on my (modded) Sceptres down to -90 degrees without any audible sound "worsening" -- a lot more and bass test tracks start to audibly pre-ring or sound a bit odd.
 
OP
edechamps

edechamps

Addicted to Fun and Learning
Forum Donor
Joined
Nov 21, 2018
Messages
910
Likes
3,620
Location
London, United Kingdom
I wonder how much DSP can “restore” each channel — even if mostly only at at the mic measuring point — as close as possible to what’s theoretically an ideal minimum response.
Oh, and I’m sure most people here do not think of EQ as an actual “fix” or cure for boundary interference issues at all, but it would still be interesting to know to what extent can DSP help “mitigate” frequency response deviations and improve overall coherence for better summation between all speaker channels in this kind situation if further repositioning changes were simply not an option. Maybe not everyone here thinks much of it, but I’m quite curious about this myself…

In general, EQ can fix anything as far as frequency response and summation are concerned, as long as (1) there is only one listening position and (2) one is not trying to pull deep nulls. The thing is, some problems have to do with spatial impressions of sound (e.g. stereo image), not just frequency response. I suspect that weird non-minimal-phase acoustic issues such as these tend to not only alter the frequency response, they also tend to mess up spatial attributes (presumably because the brain is confused by the direct sound vs. reflection behavior). EQ can fix the former, but not the latter. This is why I'm trying to find an acoustic solution to replace my EQ solution. (Also I'm hopeful acoustic solutions might also help with the large nulls in the same frequency range, which may have the same root cause.)

Do you hear an improvement with the allpass filters changing the phase only?

Good question. If I flip the allpass filter on and off, I do hear a difference. With the allpass off the bass becomes "thinner" and some bass notes seem to disappear. It's quite content dependent though. This is exactly what one would expect given the measurements. A -6 to -10 dB dip between 80-110 Hz on coherent bass signals is not something that will make a night and day difference on everything, but it's definitely worth fixing.

I've never been able to hear anything but worse sound when I was manipulating phase.

Well you need to know precisely what you're doing. Allpass filters are a highly specialized tool meant to solve very specific problems and you wouldn't reach for one unless you have to. This is the first time I had to reach for an allpass filter in ~13 years experience calibrating various systems.

Many people would argue that phase response by itself is not audible. I agree with this statement. Phase itself is not the problem… until you end up trying to sum signals that have different phase, in which case it does become a real problem. Not because of the phase mind you, but because of the resulting cancellation in the magnitude response.

If you apply the same allpass filter on all channels simultaneously, I'd expect you would be completely unable to hear any difference. If you apply an allpass filter on a single channel without a very good reason, you could indeed make it sound much worse. Basically you'd end up taking a system that was (presumably) already summing correctly and destroying its summed response, thus introducing problems where none existed. It's very much a "if it ain't broke, don't fix it" kind of thing.

My experience in small rooms (I have my pair of 8030C in a 6m2 room) is that there is a lot of bass variation (I mean from 60Hz to 100Hz) if you stay still in your chair or bend 25cm forwards or backwards. Near the center of the room, you lose a lot of bass. Near the backwall, you have too much. My REW measurements support that.

Yes, this is fully expected of small rooms. However in practice I don't move enough in my chair that this is a significant problem. And even if it was, I'd still want the best average response.

Some people have suggested listening to the system nearfield. In fact I was using these speakers nearfield for ~12 years before I decided to switch to a far-field (well, relatively speaking) system. Part of the reason was precisely to reduce location dependency. Nearfield systems have more consistent bass due to the sheer dominance of the direct sound. But the spatial experience is pretty bad because of the lack of reverb and, more importantly, even small movements of the head significantly affect listener-speaker distances, resulting in a collapse of the stereo image.

Obviously the best solution is to listen to the far field in a large room, which takes care of both problems. But it's a luxury most people don't have when living in central London. Sorry @restorer-john.

I don’t know how willing @edechamps is going to be as that large TV is going to take up a lot of real estate — most likely having to completely block the window.

Right. People in this thread might have noticed I quickly tried some of the proposed solutions and ignored others. I'm not picking the suggestions at random. It's a matter of (1) how much hassle/cost to even try the solution, (2) how practical it is as a long-term solution and (3) how likely it is to work.

For example, it was proposed several times to rotate the system by 180°. Rear speaker measurements show this is quite likely to work so it's fine as far as (3) is concerned, but rearranging the room would take me many hours so it's not great with (1) and more importantly it would block my view out the window so it goes against (2) as well. That doesn't mean I won't do it. It just means it's not high on my list right now.

This is why I want to do some simulation work. It would be a fun project and it would greatly help with (1) because it would allow me to try stuff without actually having to do it. For example it might make it possible to determine if tuned bass absorbers would help without having to actually order a whole bunch of them.

Thank you all for reminding me the i can use all pass filters to insert delay without having to use FIR. I realised i have multi subwoofer nulls that this can easily take care of.

If you have multiple subwoofers you should definitely take a look at MSO. It's a fantastic piece of software that makes it very easy to compute IIR filters for individual subs for optimal summation in a multi-sub setup. I use it even though I only have one sub just for the frequency response optimization algorithm because it's so damn good. It is this algorithm that automatically optimized my allpass parameters.

Mucking around with phase is not advisable.
Are you trying to get a flat measurement at the exact microphone location OR are you trying to get the setup to sound good and balanced for listening ?
If the former then carry on with your crazy approach. If the later then you really need to take a completely different approach, as I described above - fix your setup before playing computer jockey with flattening the measurements artificially.

This is precisely what this thread is for: to fix the acoustics so that I don't have to do play heroics with DSP which is a giant hack.

- most likely it is vertical mode minimum. Try measuring at different heights

The problem is still there with the speakers positioned on the floor. Also a purely vertical mode would affect both speakers equally and would therefore not alter their phase in relation to one another.
 
Last edited:

Chromatischism

Major Contributor
Forum Donor
Joined
Jun 5, 2020
Messages
4,765
Likes
3,703

abdo123

Master Contributor
Forum Donor
Joined
Nov 15, 2020
Messages
7,424
Likes
7,940
Location
Brussels, Belgium
In general, EQ can fix anything as far as frequency response and summation are concerned, as long as (1) there is only one listening position and (2) one is not trying to pull deep nulls. The thing is, some problems have to do with spatial impressions of sound (e.g. stereo image), not just frequency response. I suspect that weird non-minimal-phase acoustic issues such as these tend to not only alter the frequency response, they also tend to mess up spatial attributes (presumably because the brain is confused by the direct sound vs. reflection behavior). EQ can fix the former, but not the latter. This is why I'm trying to find an acoustic solution to replace my EQ solution. (Also I'm hopeful acoustic solutions might also help with the large nulls in the same frequency range, which may have the same root cause.)



Good question. If I flip the allpass filter on and off, I do hear a difference. With the allpass off the bass becomes "thinner" and some bass notes seem to disappear. It's quite content dependent though. This is exactly what one would expect given the measurements. A -6 to -10 dB dip between 80-110 Hz on coherent bass signals is not something that will make a night and day difference on everything, but it's definitely worth fixing.



Well you need to know precisely what you're doing. Allpass filters are a highly specialized tool meant to solve very specific problems and you wouldn't reach for one unless you have to. This is the first time I had to reach for an allpass filter in ~13 years experience calibrating various systems.

Many people would argue that phase response by itself is not audible. I agree with this statement. Phase itself is not the problem… until you end up trying to sum signals that have different phase, in which case it does become a real problem. Not because of the phase mind you, but because of the resulting cancellation in the magnitude response.

If you apply the same allpass filter on all channels simultaneously, I'd expect you would be completely unable to hear any difference. If you apply an allpass filter on a single channel without a very good reason, you could indeed make it sound much worse. Basically you'd end up taking a system that was (presumably) already summing correctly and destroying its summed response, thus introducing problems where none existed. It's very much a "if it ain't broke, don't fix it" kind of thing.



Yes, this is fully expected of small rooms. However in practice I don't move enough in my chair that this is a significant problem. And even if it was, I'd still want the best average response.

Some people have suggested listening to the system nearfield. In fact I was using these speakers nearfield for ~12 years before I decided to switch to a far-field (well, relatively speaking) system. Part of the reason was precisely to reduce location dependency. Nearfield systems have more consistent bass due to the sheer dominance of the direct sound. But the spatial experience is pretty bad because of the lack of reverb and, more importantly, even small movements of the head significantly affect listener-speaker distances, resulting in a collapse of the stereo image.

Obviously the best solution is to listen to the far field in a large room, which takes care of both problems. But it's a luxury most people don't have when living in central London. Sorry @restorer-john.



Right. People in this thread might have noticed I quickly tried some of the proposed solutions and ignored others. I'm not picking the suggestions at random. It's a matter of (1) how much hassle/cost to even try the solution, (2) how practical it is as a long-term solution and (3) how likely it is to work.

For example, it was proposed several times to rotate the system by 180°. Rear speaker measurements show this is quite likely to work so it's fine as far as (3) is concerned, but rearranging the room would take me many hours so it's not great with (1) and more importantly it would block my view out the window so it goes against (2) as well. That doesn't mean I won't do it. It just means it's not high on my list right now.

This is why I want to do some simulation work. It would be a fun project and it would greatly help with (1) because it would allow me to try stuff without actually having to do it. For example it might make it possible to determine if tuned bass absorbers would help without having to actually order a whole bunch of them.



If you have multiple subwoofers you should definitely take a look at MSO. It's a fantastic piece of software that makes it very easy to compute IIR filters for individual subs for optimal summation in a multi-sub setup. I use it even though I only have one sub just for the frequency response optimization algorithm because it's so damn good. It is this algorithm that automatically optimized my allpass parameters.



This is precisely what this thread is for: to fix the acoustics so that I don't have to do play heroics with DSP which is a giant hack.



The problem is still there with the speakers positioned on the floor. Also a purely vertical mode would affect both speakers equally and would therefore not alter their phase in relation to one another.

I didn’t know you can add all passes in MSO?
 

Axo1989

Major Contributor
Joined
Jan 9, 2022
Messages
2,811
Likes
2,808
Location
Sydney
... Right. People in this thread might have noticed I quickly tried some of the proposed solutions and ignored others. I'm not picking the suggestions at random. It's a matter of (1) how much hassle/cost to even try the solution, (2) how practical it is as a long-term solution and (3) how likely it is to work.

For example, it was proposed several times to rotate the system by 180°. Rear speaker measurements show this is quite likely to work so it's fine as far as (3) is concerned, but rearranging the room would take me many hours so it's not great with (1) and more importantly it would block my view out the window so it goes against (2) as well. That doesn't mean I won't do it. It just means it's not high on my list right now. ...

I'd assume if you rotate 90º in that direction your screen will block your window. Rotate 180º and the window is on your right instead of your left. But I may be missing something in your setup.

... This is why I want to do some simulation work. It would be a fun project and it would greatly help with (1) because it would allow me to try stuff without actually having to do it. For example it might make it possible to determine if tuned bass absorbers would help without having to actually order a whole bunch of them.

Depending how valuable the closet space is for storage, you could model converting it to a fairly deep perforated panel absorber or similar, whatever works for your target frequencies.
 
OP
edechamps

edechamps

Addicted to Fun and Learning
Forum Donor
Joined
Nov 21, 2018
Messages
910
Likes
3,620
Location
London, United Kingdom
I'd assume if you rotate 90º in that direction your screen will block your window. Rotate 180º and the window is on your right instead of your left. But I may be missing something in your setup.

What I should have said is, there is a nice view from that window that only works if I look at it from the current angle. If I rotate myself 180°, the view outside the window is blocked. Sunlight reflection off the TV will also increase. It's not the end of the world, but I'd like to exhaust other options first.

There were also suggestions to set up the system diagonally across the room. That might actually be fine with regard to the window, but I'd need to go back to my floorplan to figure out if I can realistically make it work with regard to room layout.

Depending how valuable the closet space is for storage, you could model converting it to a fairly deep perforated panel absorber or similar, whatever works for your target frequencies.

Right. This kind of stuff is precisely why I need a simulation. There's no way I'm going to embark on something crazy like this if I have zero confidence in the outcome.
 

daftcombo

Major Contributor
Forum Donor
Joined
Feb 5, 2019
Messages
3,687
Likes
4,068
"Manipulating" could mean anything so specifics would be illuminating... Though, I agree that it's very easy to make things worse by "mucking" about with phase EQ as SoundGuy already says -- I find it at least worth testing out just to hear the differences with my own two ears. Its interesting how different speakers react to phase "manipulation"... for instance, my ported speakers by themselves if solo'd / playing test tracks in mono sound somewhat worse when using a steep 96dB/oct linear phase HPF, whereas 24-48dB oct sounds fine. My sealed single driver Fostex 6301 monitor takes the steeper 96dB/oct filter perfectly okay, too. As for my previously mentioned 83Hz (2x stacked 2nd order) all pass filters applied to the front desktop mains, well, they do not sound worse to my ears at all, despite the visible increase in GD. I could also pull down excess phase around 50Hz on my (modded) Sceptres down to -90 degrees without any audible sound "worsening" -- a lot more and bass test tracks start to audibly pre-ring or sound a bit odd.

I've only played with phase (using RePhase) with both speakers at the same times (so not was @edechamps is doing) and have never heard any improvement, but I rather got the sound wobbling when I put several phase rotations.

What you say about steep HP filters is interesting: I put a 96dB/oct HP filter around 45 Hz on my Aria 906 (2-way ported) to get rid of a +14dB room mode, and it sounds a bit weird on some tracks. It is a linear phase filter, so I thought phase would be untouched?!

It is off-topic though.
 

Juhazi

Major Contributor
Joined
Sep 15, 2018
Messages
1,717
Likes
2,897
Location
Finland
Were you able to improve the upper pair of curves with the large hole in it? And how do their before & after spects compare?
Left curves come with changing mic position, from ear to butt on the sofa. It's my study/HT room. Sub is below the screen onthe floor, middle of short front wall. Room is rectangular and sofa is on the other end. There is nothing I can do and to be honest, it is difficut to hear the dip. Room has very low RT60 0,2. HT is used only couple of times per year by my adult kids. Ten years ago when I built the sub I tried many tricks like 1-2 extra subs, one near the ceiling, but with minimal returns.

Spectrogram is from my living room, a single 4-way speaker playing. Room is big and open to two directions, LCR speakers on the long wall. Positions are fixed and the sound is OK, I've had this setup for 9 years now.

Like Toole I have noticed, that listening expression is more important than measurements below say 200Hz, and EQ automatics even Dirac won't help much (except the lowest mode and done with ½-1/3 oct smoothing). Measuring speakers from several points in 3D space helps to understand this.

aw7+ppsl left varipos 500ms 16.jpg ainogneo83 2x4 conf1 spot MMM 500ms 112.jpg

More info about multipoint measurements and analysis by Jean-Luc Ohl https://www.ohl.to/
 
Last edited:

GalZohar

Senior Member
Joined
Aug 8, 2021
Messages
442
Likes
216
You can easily check if it is a room mode and what kind of room mode it is by playing a test tone at the relevant frequency (peak/dip of said speaker) and move (you or the mic) around to see where the peaks and dips are. At the higher bass frequencies you start getting multi-dimensional modes.
 

hmt

Senior Member
Joined
Apr 11, 2020
Messages
397
Likes
479
I've only played with phase (using RePhase) with both speakers at the same times (so not was @edechamps is doing) and have never heard any improvement, but I rather got the sound wobbling when I put several phase rotations.

What you say about steep HP filters is interesting: I put a 96dB/oct HP filter around 45 Hz on my Aria 906 (2-way ported) to get rid of a +14dB room mode, and it sounds a bit weird on some tracks. It is a linear phase filter, so I thought phase would be untouched?!

It is off-topic though.
A room mode is a minimum phase issue. Better would be to use an inverse minimum phase filter to address it. It also improves the time domain response.
 

SoundGuy

Active Member
Joined
Aug 13, 2022
Messages
115
Likes
120
Rotation of 45 degrees is what I highly recommend. You will still see out the window and only partially block the view.
 

daftcombo

Major Contributor
Forum Donor
Joined
Feb 5, 2019
Messages
3,687
Likes
4,068
A room mode is a minimum phase issue. Better would be to use an inverse minimum phase filter to address it. It also improves the time domain response.
You are right. Perhaps I didn't put it clearly: I didn't want to deal with the room mode, I wanted to cut everything below 45 Hz in order to avoid the peak but also to avoid resonances that occur at any volume a bit below that frequency and make the walls and floor of my flat vibrate as if a truck was passing by in my street. That is to remain at peace with my neighbors.

I thought a linear-phase filter would be better in that case?

A minimal-phase filter would rotate the phase a lot all over the spectrum:

1663845516966.png
 
Last edited:

ernestcarl

Major Contributor
Joined
Sep 4, 2019
Messages
3,106
Likes
2,313
Location
Canada
I thought a linear-phase filter would be better in that case?

It should be, but some speakers seem to react or sound "worse" with certain filters. Most if not all audible pre-ringing should be cancelled out when there's full complementary summing, but subs are usually designed to be separate/modular and placed away from the mains.

A minimal-phase filter would rotate the phase a lot all over the spectrum:

Yes, but usually, it's the room that exaggerates and contributes the most in any visible increase in GD and phase rotations:

1663860125670.png 1663860131307.png
Unequalized response at the MLP measured in-room by another member here recently,

*At least a lot of the room-induced anomalies below 500Hz in this example seem "symmetrical".
 

SoundGuy

Active Member
Joined
Aug 13, 2022
Messages
115
Likes
120
It should be, but some speakers seem to react or sound "worse" with certain filters. Most if not all audible pre-ringing should be cancelled out when there's full complementary summing, but subs are usually designed to be separate/modular and placed away from the mains.



Yes, but usually, it's the room that exaggerates and contributes the most in any visible increase in GD and phase rotations:

View attachment 232556 View attachment 232557
Unequalized response at the MLP measured in-room by another member here recently,

*At least a lot of the room-induced anomalies below 500Hz in this example seem "symmetrical".
The room is pretty much linear - thee are some non-linearities with adiabetic effects and non-elastic behaviour of surfaces. The room does not cause GD. Just reverb or reflections.
 
Top Bottom