• 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!

Perceptual Effects of Room Reflections

OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,388
Location
Seattle Area
Yes, unsophisticated listeners who aren't used to the clarity of reflection free listening might find the improvement unfamiliar and even jarring at first. But in the world of audio professionals - people who create and listen to music for a living - early reflections are not considered beneficial. I have visits from people who are both professional and lay, and everyone universally comments on the amazing clarity and overall excellent sound of my living room system which is very well treated.
Removing the reflections makes speakers like point sources. How on earth is that representative of a real thing? I have never heard that in an acoustic concert where sounds emanate from the two points that the stereo speakers would be at. The live venue is far more of a diffused experience.

Side reflections (not all reflections) are so beneficial that listeners can't get enough of them. It is for this reason that delayed mixing in surround provides so much more of a spacious experience.
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,388
Location
Seattle Area
If you read Ethan's whole post again, he clearly meant 'early reflections'. I have mixed feeling about this topic, but I think some kind of room treatment is always beneficial. With a dry room, you can always artificially add reverb.
It is not the same thing. I addressed that in Clark's paper in the OP.
 

ernestcarl

Major Contributor
Joined
Sep 4, 2019
Messages
3,106
Likes
2,313
Location
Canada
It is not the same thing. I addressed that in Clark's paper in the OP.

It's been a while since I read your original article, Amir. So I am not sure what you mean. Hmmm... okay, I will have to reread it again.
 

Ethan Winer

Active Member
Industry Insider
Joined
Feb 29, 2016
Messages
142
Likes
181
Location
New Milford, CT, USA
Absorption at low frequencies require very deep porous abosorbers due to large wavelengths that we are dealing with. Your graphs still show chewed up response even with the bass traps added. EQ is mandatory for smooth response. There is no getting around it. Here is a simulation of a 4 inch absorber (1.0 means 100% absorption): ... As we see, there is hardly any absorption at 100 Hz. It becomes most effective past 400 Hz and by then, most of your bass problems/resonances are behind you.

Amir, I hate to say this but you are so far behind the curve I don't even know where to begin. You need to read my Bass Trap Myths article and also my Final Dirac Report including the two earlier test articles linked within. So I'll hit the high points here, and you can read my articles and learn from them.

EQ improvements are highly positional, versus bass traps that improve the response everywhere. EQ can reduce peaks, but not raise nulls or reduce ringing which are even more damaging in most rooms. Good bass traps can easily target very low frequencies without being three feet thick. If you want a perfect response you need to create an anechoic chamber at bass frequencies. But that's not needed, and just reducing the typically 30-40 dB response span at low frequencies down to 10 dB is a huge improvement.
 

Ethan Winer

Active Member
Industry Insider
Joined
Feb 29, 2016
Messages
142
Likes
181
Location
New Milford, CT, USA
"A recent study reported that mixers preferred attenuated lateral reflections, but mastering engineers preferred a more normally reflective room. In both cases, results were different with different songs. (Tervq, et al. JAES, 2014)."

I'm quite familiar with Floyd's work. I know quite a few mastering engineers, and not one of them prefers untreated early reflections. And just because someone claims to be a mastering engineer doesn't mean they're sophisticated listeners.

I'll be glad to address your other points, but first let me ask you this: Where do you stand on this issue? What treatment do you have in your room? Do you have any bass traps? And what are the dimensions of your room? Can you post a photo of the front of your room taken from the back to show the speakers and walls etc? That will help me know how to answer your post #118.
 

MRC01

Major Contributor
Joined
Feb 5, 2019
Messages
3,424
Likes
4,030
Location
Pacific Northwest
Removing the reflections makes speakers like point sources. How on earth is that representative of a real thing? I have never heard that in an acoustic concert where sounds emanate from the two points that the stereo speakers would be at. ...
Stereo recordings don't have to sound like point sources. Recordings made with micing techniques like ORTF or Blumlein pairs create a convincing stereo image having both width and depth. The individual instruments don't come from the speakers, but are placed with precision anywhere between them, creating the illusion of the band in front of you. I find this this illusion works best in rooms that are well damped; overly reflective rooms tend to bury this imagery making every recording sound like a giant "wall of sound" spatially.

Of course, some recordings sound like a giant wall of sound anyway. That's just the way they're made. My point is that an overly reflective room makes every recording sound like that, obscuring whatever stereo effect (if any) was captured in the recording.

My experience with reflective vs. damped rooms are roughly consistent with Ethan's comments. Decades ago when I first started listening to music, I liked the diffuse sound of strong reflections. Over the years as I became a more experienced critical listener, my taste gradually shifted to a more damped room. To me, this sounds cleaner and seems to allow the sound of each recording to pass through umolested.
 

Thomas_A

Major Contributor
Forum Donor
Joined
Jun 20, 2019
Messages
3,422
Likes
2,407
Location
Sweden
"Early reflections" should be delayed enough and be low enough in level compared to the direct sound . Also, if you treat some reflections, you may end up on others being more pronounced. Reflections also mask some of the errors of the stereo system, a good thing. I prefer to treat the wall close to the speaker wall and leave the rest of the room as a conventional living room. After all, this is how I would listen if I open up the back wall and let the musicians be in a virtual room behind the wall.

If some prefer to make a studio of the living room, it's fine with me. My preferens is a cosy living room.
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,388
Location
Seattle Area
Amir, I hate to say this but you are so far behind the curve I don't even know where to begin. You need to read my Bass Trap Myths article and also my Final Dirac Report including the two earlier test articles linked within. So I'll hit the high points here, and you can read my articles and learn from them.
I read through them just now. Sadly I don't see a single reference to any research to back them. I don't know how you believe in audio science in one domain, but not acoustics. Anyway, here are some points:

1. In your EQ test of Dirac, you don't report on listening to music. Why? You should have had a loved one test you blind. This is a huge benefit of EQ systems: you can turn their effect on and off almost instantly allowing quick comparisons. No way can you do that with a dozen absorbers you have thrown around the room. By the time you install all of that, you likely will step back and admire your handiwork with the memory of how things sounded before long gone.

I have listened to Dirac especially when limited to below transition frequencies and it is excellent. It not only fixes bass problems but removal of time domain ringing allows much better clarity above bass frequencies. Equalization of levels and delay between channels has immediate benefit in soundstage as well. Again, this is something you can't fix with acoustic products.

2. Your main beef with Dirac or other EQ systems is that their effect varies with position. Well duh! They do. And there are other tools to deal with that (multiple subs).

What you miss unfortunately in your analysis is psychoacoustics. You can't trust your single microphone as a replacement for two microphones and a brain interpreting what it hears differentially between the two ears. This is why I post the research in the original article. You continue to trust faulty, single microphone measurements above transition frequencies. This is is just wrong.

While in electronics we have high trust in measurements, in acoustic domain, measurements can easily lie to you and lie big time. As wavelengths get smaller than the distance between your two ears, you cannot trust a single microphone measurements.

You want proof? Move your head 2 inches while listening to music. Do you hear a big change? Of course not. We all move our heads when listening to music without noticing a single thing changing. Yet you go on about how your measurements change when you move the microphone 2 inches! That should have been proof enough that your analysis is incorrect. You are trusting your lay intuition about acoustics rather than understanding the science of acoustics. Again, this is covered in the OP article.

This is why I asked in #1 if you used your ears to evaluate. Until you do, your reference-free articles don't have much value I am afraid.

3. If you read the OP and pay attention to ERB section, you will see why worrying about many of the things that worry your eyes in graphs, are not a worry for your ears. This is why it doesn't matter if you fix some trough if it is narrower than your ERB.

4. You fill a room with countless absorbers and then report this graph:

basstrap_waterfalls.gif


The graph in blue still sucks, pun intended. :) The notes around 40 Hz will be a ton louder than the note at 60 Hz. That differential is far bigger than the reduction in peak energy at 40 Hz that you achieved with so many products you hung on the walls.

For this reason, even if one believes in your school of thought of putting tons of absorbers in a room, you still need EQ to fix the remaining (major) problems.

I don't want to write a whole book here. Dr. Toole has done so. I suggest reading his book. If you have an issue with what he says, go to the references, many of which are peer reviewed and high authoritative (not just from him but countless other researchers). If you disagree with those research efforts, then duplicate them with all the rigor of their testing and then report back. Until then you are just repeating myths in room acoustics which have been proven time and time to be wrong. We are not living in 1970s anymore. We know a ton about psychoacoustics of sound in rooms. Try to benefit from that research rather than posting the opposite.
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,388
Location
Seattle Area
My point is that an overly reflective room makes every recording sound like that, obscuring whatever stereo effect (if any) was captured in the recording.
No one is advocating overly reflective rooms. This is again covered (briefly) in the OP. To wit, we are not saying put your stereo on a giant church and expect it to sound good on much else beyond organ music. Each reflection in the room has different psychoacostics effect due to how that direction of reflection hits your ear. Side walls tend to show benefits this way. Front, back ceiling and floor don't have such benefits. So if needed, and in an empty room they definitely are, you should and can treat those surfaces.

The point here is that if you attempt to deal with bass modes with absorbers, you will stick so many in your room that you will make your room way too dead. Physics mandate this as ironically Ethan shows in his graphs. Despite all that he put in there, he barely made a dent in the bass modes. By the time you get there, you have a dead room that is unpleasant to use for everyday use (try talking to someone in a big field a few feet away), and for a lot of music listening.

And nothing is obscured this way. Reflections increase sound energy that arrives at your ear and help with such things as intelligibility. A ton of research has been performed by Bech and others not for audiophiles but to increase comprehension for students in classrooms.

Here is how Dr. Tool summarizes his decades of research performed by himself and others in this area:
Floyd-Reflections.png


One has to put away all intuition and trust the science here. Your intuition that reflections are bad, evil, etc. as Ethan states are just wrong.
 

krabapple

Major Contributor
Forum Donor
Joined
Apr 15, 2016
Messages
3,168
Likes
3,715
I'm quite familiar with Floyd's work. I know quite a few mastering engineers, and not one of them prefers untreated early reflections. And just because someone claims to be a mastering engineer doesn't mean they're sophisticated listeners.

I'll be glad to address your other points, but first let me ask you this: Where do you stand on this issue? What treatment do you have in your room? Do you have any bass traps? And what are the dimensions of your room? Can you post a photo of the front of your room taken from the back to show the speakers and walls etc? That will help me know how to answer your post #118.


Nuh-uh, that's not how it works. This isn't a 'fill out your gear profile' test. I didn't say anything about what I do or don't do. I'm referring to scientific research, done by others, with other people as subjects, whom you keep conveniently dismissing as 'unsophisticated', and I'm correcting misrepresentations you made about Dr. Toole (he likes 'boomy bass'?? Feel free to check out *his* room and gear...he's posted it online). I'm also referring to principles of claim testing.

Address those points.
 

MRC01

Major Contributor
Joined
Feb 5, 2019
Messages
3,424
Likes
4,030
Location
Pacific Northwest
No one is advocating overly reflective rooms. ... One has to put away all intuition and trust the science here. Your intuition that reflections are bad, evil, etc. as Ethan states are just wrong.
I'm not saying all reflections are bad, or that I want my room to be a completely dead anechoic chamber (which would be impossible to achieve anyway). My experience is that most untreated rooms are too reverberant/reflecting, even with speakers ideally positioned. They sound much better with room treatment. How much is a matter of taste. I find that I prefer more rather than less. But I have heard the effect of "too much" room treatment and pulled back from that. So it's not completely dead, some reflection is beneficial.

One related point: EQ is useful, but the cure can be worse than the disease in some cases. In low frequencies, I believe from experience in room treatment first, then EQ. Room treatment weakens the modes while EQ only slaps a band-aid over the problem. For example, my room had a 70 Hz null that no amount of EQ could fix -- all EQ did was pump more energy into a black hole, making the bass sound sloppy with worse decay and high distortion. Room treatment weakened this null, flattening FR while also improving CSD.
 

krabapple

Major Contributor
Forum Donor
Joined
Apr 15, 2016
Messages
3,168
Likes
3,715
I'm not saying all reflections are bad, or that I want my room to be a completely dead anechoic chamber (which would be impossible to achieve anyway). My experience is that most untreated rooms are too reverberant/reflecting, even with speakers ideally positioned. They sound much better with room treatment. How much is a matter of taste. I find that I prefer more rather than less. But I have heard the effect of "too much" room treatment and pulled back from that. So it's not completely dead, some reflection is beneficial.

One related point: EQ is useful, but the cure can be worse than the disease in some cases. In low frequencies, I believe from experience in room treatment first, then EQ. Room treatment weakens the modes while EQ only slaps a band-aid over the problem. For example, my room had a 70 Hz null that no amount of EQ could fix -- all EQ did was pump more energy into a black hole, making the bass sound sloppy with worse decay and high distortion. Room treatment weakened this null, flattening FR while also improving CSD.


Reflection of frequencies above Schroeder , not modal resonances, are what the whole 'lateral reflections/early reflections' debate is about. They're two separate issues.

The modal room resonance 'problem' has two aspects: having *too much* (humps) and *too little* energy (nulls) at different bass frequencies. Room EQ can totally flatten modal *humps*, and I'd hardly call fixing one out of two aspects a 'band-aid'. No software EQ I know of claims to eliminate modal *nulls* but that's what careful subwoofer(s) placement can treat. Those two approaches take you a long, long way towards treating modal issues, without installing any bass traps. Things get more complex when you have to treat more than one listening position, but I suspect many of us here are mainly concerned with the prime spot.
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,388
Location
Seattle Area
One related point: EQ is useful, but the cure can be worse than the disease in some cases. In low frequencies, I believe from experience in room treatment first, then EQ. Room treatment weakens the modes while EQ only slaps a band-aid over the problem. For example, my room had a 70 Hz null that no amount of EQ could fix -- all EQ did was pump more energy into a black hole, making the bass sound sloppy with worse decay and high distortion. Room treatment weakened this null, flattening FR while also improving CSD.
If you pull the peaks down, the nulls may become at the desired levels by themselves. Let's use Ethan's measurement with Dirac:

studio_response-off-v-on.gif


Notice how the null in blue with Dirac is far less than it was in when we had the red at 110 Hz. If you allow a sloping down curve for the target, that would have been less consequential still.

Acoustic products didn't solve that problem any better anyway:

waterfalls.gif


The trough is still there at 110 Hz in Blue after all the treatments Ethan threw at it. And there is a worse one at 60 Hz.

I highly encourage using EQ even if you use a ton of absorbers in your room. The improvement is quite easy to assess both in measurements and listening. And measurements here do work because the wavelength of the sound waves is quite large relative to distance between our ears so psychoacoustics plays little role in this domain (bass).
 

MRC01

Major Contributor
Joined
Feb 5, 2019
Messages
3,424
Likes
4,030
Location
Pacific Northwest
Reflection of frequencies above Schroeder , not modal resonances, are what the whole 'lateral reflections/early reflections' debate is about. They're two separate issues.

The modal room resonance 'problem' has two aspects: having *too much* (humps) and *too little* energy (nulls) at different bass frequencies. Room EQ can totally flatten modal *humps*, and I'd hardly call fixing one out of two aspects a 'band-aid'. ...
Above Schroder, I find that some absorption & diffraction can be beneficial, for example carefully placed RPG acoustic foam 4" thick. But experimenting, I've found there's a sweet spot between "not enough" and "too much".

Regarding EQ for bass humps: it can flatten the humps but it has less impact on CSD. For flattening humps, room treatment is still my first choice because I find it also improves both FR and CSD. Generally speaking, I use room treatment to get it as good as I can, which is never perfect, then EQ for the rest.
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,388
Location
Seattle Area
Regarding EQ for bass humps: it can flatten the humps but it has less impact on CSD.
Definitely not true. See this extensive article I wrote on this topic: https://www.audiosciencereview.com/...urements-understanding-time-and-frequency.25/

And this graph with just one filter added to pull down a peak:

i-z7Cg55m-L.png


Decay time is most definitely reduced at the frequency of the filter (green with filter, yellow without).

Adding more filters showing the same results:

i-tfbz347-L.png


As you see, the decay times across the full spectrum is massively reduced and equalized.

This said, you do want to listen to addition of each filter to make sure it has an improved subjective effect. Sadly you can't do this easily with absorbers and sadly people accept their sighted evaluation of it as being positive. I have seen many people years later remove these acoustic products and notice the improvement in sound.
 

MRC01

Major Contributor
Joined
Feb 5, 2019
Messages
3,424
Likes
4,030
Location
Pacific Northwest
If you pull the peaks down, the nulls may become at the desired levels by themselves. ...
I've observed the same. Reducing the peaks doesn't fix the nulls, but it often makes them less steep. However, when it comes to nulls, I find that room treatment is more effective than EQ.

... Acoustic products didn't solve that problem any better anyway:
In this example, consider the 100 Hz hump, and compare an EQ, and a room treatment, each reducing it by the same amount, say -6 dB. IME, the room usually also has a CSD problem ringing at 100 Hz, and the room treatment damps this decay more than the EQ does.
... I highly encourage using EQ even if you use a ton of absorbers in your room. ...
I agree!
 

MRC01

Major Contributor
Joined
Feb 5, 2019
Messages
3,424
Likes
4,030
Location
Pacific Northwest
Definitely not true. See this extensive article I wrote on this topic: https://www.audiosciencereview.com/...urements-understanding-time-and-frequency.25/
...
I didn't say that EQ doesn't improve CSD, I said that it doesn't improve it as much as room treatment that has equal effect on FR. I don't have a theory or paper to prove it, it's just an observation made from experience tuning rooms.

PS: my observation that room treatment improves CSD more than EQ does, is plausible if you think about it like this. If you EQ to attenuate a bass hump at a room mode, you are reducing energy in that mode, so obviously it will now decay faster. But you haven't changed the room response, which is causing that mode to exist in the first place. The room is still resonating at that frequency, you're just giving it less energy there. If instead of EQ, you use room treatment to get the same attenuation at that mode, you are not just reducing energy at that mode, you are actually changing the room response that causes this mode to exist. The room isn't resonating at this frequency anymore (or it's resonating more weakly).
 
Last edited:
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,388
Location
Seattle Area
PS: my observation that room treatment improves CSD more than EQ does, is plausible if you think about it like this. If you EQ to attenuate a bass hump at a room mode, you are reducing energy in that mode, so obviously it will now decay faster. But you haven't changed the room response, which is causing that mode to exist in the first place. The room is still resonating at that frequency, you're just giving it less energy there. If instead of EQ, you use room treatment to get the same attenuation at that mode, you are not just reducing energy at that mode, you are actually changing the room response that causes this mode to exist. The room isn't resonating at this frequency anymore (or it's resonating more weakly).
The "room" hasn't changed in either case. Acoustic products are absorbing some of that energy and converting it to heat. They don't know the difference between a room mode and not. In contrast, with EQ, we identify the room mode using measurements and dial just that region down. We don't try to absorb all energy at all frequencies.

But again the issue is that common porous absorbers are simply ineffective at low frequencies where modal response is the worst. To get them to have real benefit would require a ton of them. By then, you have made the rest of the room too dead. And of course made your room too ugly to want to sit there and listen to anything. :)
 

MRC01

Major Contributor
Joined
Feb 5, 2019
Messages
3,424
Likes
4,030
Location
Pacific Northwest
The "room" hasn't changed in either case. Acoustic products are absorbing some of that energy and converting it to heat. They don't know the difference between a room mode and not. In contrast, with EQ, we identify the room mode using measurements and dial just that region down. We don't try to absorb all energy at all frequencies.
If all room treatments were broad spectrum affecting all frequencies equally, I'd agree. But some are, and others are not. For example tuned membranes and bass traps can be tuned to affect narrower frequency ranges than tube traps. By absorbing the energy at the bass mode frequency, they weaken the reflected wave, which weakens the room mode. The treatments don't "know" it's a room mode, but you size & shape them to tune them to the room mode frequency, so the effect is the same: it prevents the room from resonating at that frequency (or at least weakens the resonance). Intuitively, it's like removing the wall that was causing the resonance, or at least making that wall semi-permeable to weaken the resonance. In this sense, one can say that it effectively changes the room's response. This is different from EQ, where the room modes are all the same and just as strong, but you're just reducing energy there.

BTW, your point about overdamping has merit; it's why we have to use different kinds of treatments (broad or narrow frequency spectrum) to solve different kinds of problems. For example, the reason I have tuned membranes is because the tube traps affect a broad spectrum of frequencies, and if I use enough of them to level the response at my problematic frequencies (70 Hz), it has the side-effect of over-damping the room. So I used fewer tube traps, replacing some of them with more narrowly focused 70 Hz tuned membranes. This enabled me to kill the resonance without over-damping the room.
 
Last edited:
Top Bottom