The Reverberant Field: Why it matters, where the goal posts are

Duke LeJeune

[Industry Expert]/Member Sponsor
Jul 22, 2013
747
1,200
435
Princeton, Texas
Yes, in my testing people always prefer my horns to my Pioneer S-1EX (Same drivers as TAD Evolution Tower) and modded Odyssey Stratos amp, and the reason for the preference is the greater sense of immersion from the horns. But, I also think it takes some time to acclimate to the horns greater ratio of direct to reflected sounds and the near absence of 1st reflections. I think preference can change with experience and hearing lots of different systems so you're not as acclimated to one particular system.

I was very impressed with the Pioneer S-1EX when I heard it several years ago. Your experience comparing two excellent examples of different schools of thought essentially side-by-side is very interesting and educational to me.
One other factor is my room is not a fully treated and dedicated room, it has some absorption to adjust decay times, but is more like a typical living room, so the conventional speakers are at a disadvantage vs a dedicated room. It has high ceilings which I never want to be without again, it's a big advantage over 8ft ceilings. Anyways, with a dedicated room I've heard conventional speakers do a great job with immersion as well, so I'm thinking those early 1st reflections can work if the acoustic space is more ideal, otherwise you're probably better off with a speaker that avoids producing them in the 1st place.

So for me, I see a big difference in design for a speaker that is intended to be used in a dedicated space vs a living room. It may well be in a dedicated room a big cone n dome with wide dispersion is just as good as any other solution, while a speaker with narrow dispersion is, at least to me, the best choice for a non-ideal space.

I think you are absolutely correct.

I had not thought about this until you mentioned it a day or two ago, and now I think you nailed it. I HAVE heard "conventional" speakers do the immersion thing, but it has been in a professionally-treated dedicated room. You connected the dots for me, and that helps me define where my niche (speakers for non-dedicated rooms) lies in relation to other high-end speakers.

Thank you Dave!!
 
  • Like
Reactions: DaveC

Blackmorec

Well-Known Member
Feb 1, 2019
747
1,271
213
Small rooms vs large rooms vs concert halls

The first thing to be aware of when it comes to treating reflections is the room vs recording venue reflections because they are 2 discreet, separate entities where 1 is highly desirable and the other not and essentially you need to get rid of one in order to clearly hear the other.

The second thing to be aware of is the precedence effect or law of the first wavefront, a binaural psychoacoustical effect. When a sound is followed by another spectrally highly related sound, separated by a sufficiently short time delay (below the listener's echo threshold), listeners perceive a single auditory event; its perceived spatial location is dominated by the location of the first-arriving sound (the first wave front). The lagging sound also affects the perceived location. However, its effect is suppressed by the first-arriving sound and the 2 sounds are summed together and heard as a single sound.

Bearing that in mind, you’ll realise that concert halls are fairly large so there are very few If any ‘early’ reflections that meet the precedence effect, so reflections are typically assigned their own location and are heard as separate sounds, which we call hall reverberation. When recordings are made in the hall, this hall ambience is recorded along with the direct sound.

Lets now move on to listening rooms. A small, highly reflective room will produce a lot of early reflections that will meet the precedence effect criteria and essentially only those sounds that bounce between 2 walls will register to the listener as separate sounds . Now bear in mind that we are playing a recording from the concert hall that already has a recorded acoustic with later arriving reflections, so ideally you don’t want your listening room to add yet more late reflections. Adding substantial diffusion to the back wall is a great way to prevent sounds bouncing between the 2 walls then back to the listener. The small room is therefore capable of reproducing the concert hall’s reverberation, with no masking reflections from the listening room itself, resulting in the small room sounding to the listener like the concert hall I.e direct energy + late arriving hall reflections.

In a large room, most of the reflections reaching the ears of the listener will again not meet the requirements of the precedence effect and will therefore register with the listener as separate sound sources with their own direction; partially masking the concert halls own reflections (ambience) and partially imposing its own listening room identity and dimensions on the sound. In this case its highly beneficial to treat and remove those reflections with a degree of absorption and rear wall diffusion. Once treated the listener again hears the original hall sounds plus reverberation with no masking from the listening room, resulting in the large room sounding like the original concert hall venue.

The final difference is in sound intensity. In the small, reflective room, direct sounds and early first reflections are summed, leading to an increase in sound intensity and saturation. In the larger room those reflections are absorbed so correspondingly greater SPLs are required to achieve the same intensity levels. It can be seen from the above that speaker matching is quite important as it is easy to overload a small room, or under power a large room. Also the type of speaker is important....a sealed, forward firing cabinet with little distance between drivers is going to work better in a smaller room, where the sound gets less space to integrate, while large ported and dipole speakers will require more space for the direct sound from the drivers and back radiation to integrate.
 
  • Like
Reactions: Duke LeJeune

DaveC

Industry Expert
Nov 16, 2014
3,899
2,141
495
Precedence effect is very interesting... it's VERY different wrt time threshold depending on the music, and it's far different for vocals and tracks with rhythmic percussion, with those time thresholds being much smaller while the brain tends to sum symphonic music at much longer time delays.
 
  • Like
Reactions: Duke LeJeune

DaveC

Industry Expert
Nov 16, 2014
3,899
2,141
495
I was very impressed with the Pioneer S-1EX when I heard it several years ago. Your experience comparing two excellent examples of different schools of thought essentially side-by-side is very interesting and educational to me.


I think you are absolutely correct.

I had not thought about this until you mentioned it a day or two ago, and now I think you nailed it. I HAVE heard "conventional" speakers do the immersion thing, but it has been in a professionally-treated dedicated room. You connected the dots for me, and that helps me define where my niche (speakers for non-dedicated rooms) lies in relation to other high-end speakers.

Thank you Dave!!


It's a good topic, one that determines design priorities for sure.

I included the Polk speakers because they seem to accomplish the goal of immersion and are room friendly in that the cancellation works in untreated rooms and the distance the speakers are placed apart is much less critical, so they are quite living room friendly too.

They don't sound refined, the tweeters are not pleasant to listen to, but the design concept works, at least in the context of a fairly short demo at RMAF this year.
 
  • Like
Reactions: Duke LeJeune

Duke LeJeune

[Industry Expert]/Member Sponsor
Jul 22, 2013
747
1,200
435
Princeton, Texas
In Post #2 I wrote:

"So, we want to minimize the early reflections but encourage the later ones. The time scale will be of course different in our smaller rooms, but the general principle seems to be valid (I'll explain later). "

I'll try to do some of that explaining now.

The audiophile world has access to abundant anecdotal evidence of the benefits of introducing a time gap in between the first-arrival sound and the heavy onset of reflections (corresponding in concept but not in scale to the third clip Griesinger plays in the video linked in that same post). This evidence comes from the multitude of dipole speaker owners, from baby Maggies to the biggest Apogees and SoundLabs and so forth.

Dipole speaker owners have virtually all experimented with distance from the wall, and noticed improvements in many areas as that distance is increased. These areas of improvement are included in my "Done right..." paragraph in the opening post.

I'm one of these experimentalist dipole owners.

One of the most interesting things that happens as dipole speakers are pulled further out into the room is, we seem to hear more of the venue and less of the room we are listening in. At around five feet with an otherwise good setup and a good recording, we start to get into that elusive, immersive "you are there" territory. Later when I began working with Earl Geddes and he mentioned 10 milliseconds, I made the connection: Ten milliseconds is about the time it takes for the backwave to make the round trip to the wall at five feet distance.

The increase in soundstage depth was far greater than can be accounted for by adding few additional milliseconds to the backwave's arrival time. So here is what I think is going on:

The ear/brain system judges room size by the early reflections, and by adding a major onset of reverberant energy arriving at an unusually late time, the net result is a significant weakening of the "small room signature" we would normally experience. And with less small room signature superimposed atop the recording, we hear more of the recording and less of the room; assuming a suitable recording, we start to become "immersed" in the soundfield on the recording.

(I think the extra reverberant energy from the backwave does some backwards masking of the early reflections, as when I absorbed most of the backwave the awareness of being in a small room seemed to return.)

In other words, I think that the combination of weak early reflections and shifting the major onset of reflections to perhaps ten milliseconds (or more if possible) pays disproportionate dividends in pursuit of a sense of immersion.

This configuration does not begin to replicate the timescales in a concert hall, but it is consistent with the general principle described by Griesinger:

"Envelopment is perceived when the ear and brain can detect TWO separate streams: A foreground stream of direct sound, and a background stream of reverberation. Both streams must be present if sound is perceived as enveloping."

Why might this matter? Griesinger again:

"Envelopment is the holy grail of concert hall design. When reproducing sound in small spaces [home listening rooms], envelopment is often absent."

And quoting our own DaveC on the subject:

"The most important driver of preference is the sense of immersion."
 
Last edited:

Blackmorec

Well-Known Member
Feb 1, 2019
747
1,271
213
Hi Duke,
surely we are depending on the concert hall recording to provide the envelopment, not the room, otherwise we have 2 sets of acoustics.....the recording venue’s, which we want to hear because the timing and decay are correct for the venue and the room’s, which we don’t want to hear because both timing and decay are related to the room rather than to the recording. We don’t want to hear every recording with the listening room’s distinct acoustic....we want to hear the individual acoustic from each recording.
 
  • Like
Reactions: Duke LeJeune

Duke LeJeune

[Industry Expert]/Member Sponsor
Jul 22, 2013
747
1,200
435
Princeton, Texas
surely we are depending on the concert hall recording to provide the envelopment, not the room, otherwise we have 2 sets of acoustics.....the recording venue’s, which we want to hear because the timing and decay are correct for the venue and the room’s, which we don’t want to hear because both timing and decay are related to the room rather than to the recording. We don’t want to hear every recording with the listening room’s distinct acoustic....we want to hear the individual acoustic from each recording.

I totally agree!!

I was planning to do a separate thread about the Two Venues - the First Venue being the acoustic space on the recording (whether real or synthesized), and the Second Venue being the playback room. And obviously we want to hear the First Venue, and not hear the Second Venue.

I gotta roll up my sleeves now and do some real work, but I plan to come back to your excellent posts later today.
 
  • Like
Reactions: Blackmorec

Blackmorec

Well-Known Member
Feb 1, 2019
747
1,271
213
Precedence effect is very interesting... it's VERY different wrt time threshold depending on the music, and it's far different for vocals and tracks with rhythmic percussion, with those time thresholds being much smaller while the brain tends to sum symphonic music at much longer time delays.
Hi Dave,
psychoacoustics/human hearing is part of our sensory nervous system and goes back to the origin of the species where it developed as one of our survival mechanisms.....the ability to sort out the exact position of a sound‘s origin without being confused by nearby reflections. A useful ability for both hunter and hunted. Stereo utilises that ability to fool us into hearing single sources of sounds in a ‘soundscape’ rather than what is in reality 2 discreet sources (the loudspeakers). i would imagine that the biggest difference between vocal, rhythmic percussion and symphonic music is in fact the recording venue and recording techniques. Most vocal and rhythmic percussion are recorded in studios. often in sound booths without ambience or reverb and the reverb is added artificially while symphonic music is by necessity recorded in large echoey, highly reverberant concert halls, so the reverberation contents are going to vary markedly.
 
Last edited:

DaveC

Industry Expert
Nov 16, 2014
3,899
2,141
495
Hi Dave,
psychoacoustics/human hearing is part of our sensory nervous system and goes back to the origin of the species where it developed as one of our survival mechanisms.....the ability to sort out the exact position of a sound‘s origin without being confused by nearby reflections. A useful ability for both hunter and hunted. Stereo utilises that ability to fool us into hearing single sources of sounds in a ‘soundscape’ rather than what is in reality 2 discreet sources (the loudspeakers). i would imagine that the biggest difference between vocal, rhythmic percussion and symphonic music is in fact the recording venue and recording techniques. Most vocal and rhythmic percussion are recorded in studios. often in sound booths without ambience or reverb and the reverb is added artificially while symphonic music is by necessity recorded in large echoey, highly reverberant concert halls, so the reverberation contents are going to vary markedly.

Possibly, but I think it's been shown that the brain processes different sounds in different parts of the brain, with music and vocals having different types of processing and rhythmic percussion and single drum strikes being in yet another. I think the rules of perception are different for each of these categories of sound.

You can experiment with this at home with latency, it's interesting to hear music perceived as a single sound then hear it devolve to 2 separate sources at times depending on the content.
 

Blackmorec

Well-Known Member
Feb 1, 2019
747
1,271
213
Possibly, but I think it's been shown that the brain processes different sounds in different parts of the brain, with music and vocals having different types of processing and rhythmic percussion and single drum strikes being in yet another. I think the rules of perception are different for each of these categories of sound.

You can experiment with this at home with latency, it's interesting to hear music perceived as a single sound then hear it devolve to 2 separate sources at times depending on the content.
Interesting. I have not read any scientific research about this but i certainly don’t claim to cover all the literature on the subject. If you have any references I’d be interested.
Such a difference would potentially have a major impact on how we perceive music in immersive soundscapes.
Take reverb as an example:
Lets say that someone is singing in a very large building, a church say. There are two elements that tell us about the venue.....the time delay on the echo and the change in amplitude between the direct and reflected sound. Quiet passages shouldn’t have audible reverb, because the loss of amplitude takes the reverb below our threshold of detection. Louder passages have sufficient energy to travel, bounce and return to our ears as audible reverb. For the reverb to sound right, it has to be consistent......same delay time and same amplitude loss ratio between direct and reflected sounds. Given the role of amplitude, dynamic wave shape will also play a roll. Notes have a particular shape....take a trumpet or a saxophone....the note shape is usually something like a brash, percussive initial sound followed by a beautiful and increasingly smooth decay. What you’ll hear as reverb is the initial percussive part of the note only, which itself decays as amplitude drops with each reflection. That means that reverb has a different tonal quality to the direct sound. The trumpet or sax’s decay is based on the instrument’s resonance gradually dying away while the buildings reverb decay is based on the initial percussive sound dying away as it reflects multiple times. That’s why buildings impart their own sound onto performances. It also means that any sound with very dynamic, percussive initiation is going to trigger reverb, whereas smoother tones....for example violin or guitar typically don’t trigger audible resonance. But a violin still sounds different when played in a large building than it does in a studio. It just wont have that percussive, direction giving reflection that more percussive instruments have
That may be one way to explain the difference between processing various instruments, although this has to do with acoustics rather than psychoacoustics.
 

DaveC

Industry Expert
Nov 16, 2014
3,899
2,141
495
Interesting. I have not read any scientific research about this but i certainly don’t claim to cover all the literature on the subject. If you have any references I’d be interested.
Such a difference would potentially have a major impact on how we perceive music in immersive soundscapes.
Take reverb as an example:
Lets say that someone is singing in a very large building, a church say. There are two elements that tell us about the venue.....the time delay on the echo and the change in amplitude between the direct and reflected sound. Quiet passages shouldn’t have audible reverb, because the loss of amplitude takes the reverb below our threshold of detection. Louder passages have sufficient energy to travel, bounce and return to our ears as audible reverb. For the reverb to sound right, it has to be consistent......same delay time and same amplitude loss ratio between direct and reflected sounds. Given the role of amplitude, dynamic wave shape will also play a roll. Notes have a particular shape....take a trumpet or a saxophone....the note shape is usually something like a brash, percussive initial sound followed by a beautiful and increasingly smooth decay. What you’ll hear as reverb is the initial percussive part of the note only, which itself decays as amplitude drops with each reflection. That means that reverb has a different tonal quality to the direct sound. The trumpet or sax’s decay is based on the instrument’s resonance gradually dying away while the buildings reverb decay is based on the initial percussive sound dying away as it reflects multiple times. That’s why buildings impart their own sound onto performances. It also means that any sound with very dynamic, percussive initiation is going to trigger reverb, whereas smoother tones....for example violin or guitar typically don’t trigger audible resonance. But a violin still sounds different when played in a large building than it does in a studio. It just wont have that percussive, direction giving reflection that more percussive instruments have
That may be one way to explain the difference between processing various instruments, although this has to do with acoustics rather than psychoacoustics.


Yup! I think it has something to do with how a system can be better for one type of music vs another... certainly not the only reason and maybe not even a main factor, but it points to the possibility that that how the brain interprets reflections and how the precedence effect works may be different depending on what the system is reproducing.

IME, flowing music without vocals and a defined rhythmic beat allows the brain to merge reflections + the direct sound into a single source with much longer delays vs vocals or impulses. Vocals seem to vary depending on whether it's spoken clearly vs being more part of the music.

https://www.forbes.com/sites/evaams...om-instrumental-music-or-speech/#5995f02e2a54
 

Hipper

Well-Known Member
Jun 12, 2011
68
11
83
Hertfordshire, UK
@KlausR. and @Duke LeJeune I'd be interested on your opinions on the new Polk speakers that cancel the output of the opposite speaker in it's half of the room...

It accomplishes something similar to adding a divider in the room going from the LP to between the speakers.

What I've noticed is folks used to reflective rooms tend to describe the result of too much direct vs reflected sound, or the Polk sound, as close to headphones and initially it's met with some resistance but over a short period of time most acclimate and "get" the whole point and their brain has a chance to process the different presentation.

On the Polk speakers, one advantage is the distance in between the speakers doesn't matter as much, and it'll produce a soundstage dictated by the recording with images well outside the boundaries of the speaker in a way only the best setup and performing systems can. Of course the downside is symmetry and sitting right in the sweetspot is more important, but I wouldn't say it's like listening to a system that really beams and is head in a vice, it's more forgving than that but maybe less than an average system.


Thanks for mentioning these. I'd not heard of them.

What they are trying to achieve is the same as Ambiophonics:

https://www.ambiophonics.org/

That is, stopping the interaural crosstalk . This is where the left ear hears not only the sound from the left speaker but also that from the right. The latter is slightly delayed and so causes comb filtering, notably around 2kHz, which can damage the stereo phantom image.

(It is the same sound you get with headphones. Ironically that is why some headphone amps have a crossfeed processor, to reintroduce the interaural crosstalk!)

According to Toole, early side wall reflections can add some of the dB loss caused by interaural crosstalk to the phantom image.

My experience in a small room (14' x 13') with plenty of bass traps is that removing side wall reflections is beneficial. I find the centre image stronger and sharper as a result. I do not get an immersive sound though, my mostly studio bound pop music never expanding beyond the speakers.
 

KlausR.

Well-Known Member
Dec 13, 2010
291
29
333
Hello Duke,

I agree that perception thresholds for first reflections are different for different delay times, but that does not change this basic principle articulated by David Griesinger:

“The earlier a reflection arrives the more it contributes to masking the direct sound.”

By definition, reflections have lower levels than the direct sound. How then is a reflection capable of masking the direct sound? The mere existence of perception thresholds would be evidence to the contrary?

I take David's recordings as illustrative of this general principle, rather than as being limited in applicability to large rooms only. While I wouldn't “transcribe” the results of experiments in large rooms to small rooms, in this case I certainly would not dismiss them (not saying that you are dismissing them, just advocating for a middle ground).

It is not clear (to me) what exactly the recording was: was it a Lokki recording with electronic orchestra? Did Griesinger make the recording himself, and where?

Further, this recording was manipulated by Matlab, but in what manner? Temporal gating, equalization? Then this recording was reproduced by a single loudspeaker (mono, right channel of the binaural recording) during Griesinger’s lecture and again recorded by one or more microphones, those mikes not knowing about binaural decoloration. Further, from what I read about concert hall acoustics, reflections up to 80 ms delay are considered as early. 80 ms is echo threshold (precedence effect) for slow music, and is above echo thresholds for all other types of signals, so I would not be surprised that a clip direct + 1st sounds bad if gated at 80 ms. However, does a concert hall, every concert hall, sound bad when you are attending a concert?

In the spirit of "a picture is worth a thousand words", I re-submit the sound clips from that Griesinger lecture, post #2 above. If those clips are not evidence of early reflections being detrimental, I don't know what is.

Both clips (direct + 1st, all no 1st) sound somewhat strange to me. If early reflections really were detrimental, then what about the last clip (all together)? This should sound strange as well with the early reflections being present. In addition, the 2nd clip is the one Griesinger uses for demonstrating the detrimental effects but he is talking at the same time, and he is biasing the audience by making comments about how this clip sounds in opinion! He later talks about his preference re: the clips! If I understood it correctly, the recording was equalized for a seat he didn’t like. Again, his preference. Further, the clips are played in mono, right channel of a binaural recording.

In view of the above I really don’t think that one can take this particular recording of a recording of concert hall sound as evidence for early reflections having detrimental effects of reproduce sound in a domestic setting.

I think Geddes would also disagree with your position. He suggests a target of avoiding reflections arriving within 10 milliseconds of the direct sound. In his room he breaks up the floor and ceiling reflections, as their path lengths are inevitably too short for his 10 ms target.

Earl advocates no early reflections, hence his constant directivity loudspeakers. Still, there is no evidence, as far as I’m aware, for detrimental effects of these reflections. The delay of a reflection is determined by the setup of the system in the room. Absorbers will absorb reflections regardless of delay, so why specifically aim at 10 ms? They [absorbers] do this [absorb regardless of delay] also on floor and ceiling so I don’t quite understand his approach of breaking up those reflections. Speaking of absorbing reflections: perception thresholds for lateral reflections with music are as a low as -25 dB for 10 ms delay, yet the recommendations require decreasing reflection level by 10 or 15 dB, which means that the lateral reflections are still audible. However, the studies are for a single reflection, what thresholds with music look like in a complete sound field, is not known. Extrapolation of known data would give a threshold of -19 dB.


I agree with Geddes based on my own experiences... not that I'm putting forward my own experiences as "primary evidence”, but rather as my “excuse” for agreeing with him.

Speakers with bad off-axis behaviour, acoustically asymmetric setup, presence of massive acoustic treatment, in these cases early/first order reflections may be detrimental. Does that mean there are detrimental in each and every case? I wonder what evidence Earl bases his suggestion on.

With all respect, as long as no appropriate controls have been included in listening tests, I would consider personal listening experience as anecdotal at best.

Another way that early reflections are detrimental in home audio is, they tell you that you're in a small room. From your paper:
" The delay time of the early reflections determines the impression of the size of the room, for natural hearing and also for two-channel sound transmission (Kuhl 1978).”

This relates to the size of the room in which the recording has been made, not the room in which the recording is listened to. I have modified the write-up accordingly. Thanks for the pointer. Kuhl's paper is from 1978, more recent research states that reverberation time is the most important parameter:

Hameed et al., “Psychoacoustic Cues in Room Size Perception”, AES paper 6084 (2004)
Yadav et al., “Investigating auditory room size perception with autophonic stimuli”, AES paper 8934 (2013)

Can you tell me what you mean by "the field is rather (highly) directional"? I have a vague idea of what that means but you probably have a very clear one.

As an answer I attach my write-up on sound fields.

I'm confused by what seems (to me) to be conflicting statements about Inter-Aural Cross Correlation (IACC).

Well spotted! I had a good look at those passages and the corresponding papers and have amended the write-up accordingly. It now reads:

However, acoustic treatment results in a substantial, frequency dependent change of the interaural cross-correlation coefficient ICCC (Tohyama 1989) and results generally in a increase of the coefficient relative to the reflective wall when absorbers or diffusers are added (King et al. 2011, Kishinaga 1979) with subsequent narrower image (Kishinaga 1979) and leads in the extreme case of the anechoic chamber to the perceived sound field differing greatly from the original field (Tohyama 1989).

- Adding acoustic treatment (absorbers, diffusers) to first reflection points results in a increase of the interaural cross-correlation coefficient ICCC, whereas low values of ICCC are preferred.

Is ICCC the same thing as IACC?

Yes.

Klaus
 

Attachments

  • Schallfeld_GB.pdf
    679.2 KB · Views: 5
  • Like
Reactions: Duke LeJeune

Blackmorec

Well-Known Member
Feb 1, 2019
747
1,271
213
Interesting subject worth a few notes:
Firstly you can’t equate early elections in a large hall with early reflections in a small room, because from a timing point of view, and timing is at the crux of this discussion, an early arriving reflection in a large hall would have the same timing as a late reflection in a small room.

If we look at the Law of First Wavefront, that too is all about the timing of the reflection specifically early reflections that arrive within a few ms of the original sound being heard. By definition, the Law of First Wavefront can only work on reflections that fall below the listener’s ‘echo threshold‘ and such reflections only exist in small rooms....based simply on dimensions and the time sound takes to travel. In larger halls the reflections simply take too long to cover the distances involved.
In small rooms, the early arriving reflections are summed along with the original sound and only one sound is heard, with the original sound’s direction component, so again it cannot by definition mask the original sound since, as far as the ear is concerned, there is only the original sound. In large halls, the early arriving reflections have a delay above the ‘echo threshhold’ and are heard as a separate sound source with unique directional component and they can indeed interfere with the original sound, masking its decay, which is obscured by the early arriving reflection.

Thus, you can’t take measurements and their conclusions made in a large hall and apply them to a small room as sounds and reflections in a small room are treated differently psychoacoustically.

This means that in small rooms, late arriving reflections, which would mask the original sound‘s decay and the recording venue’s reverberation should be treated by diffusing them, thereby preventing multiple reflections, while in large rooms, early arriving reflections should be absorbed, because they too will mask note decay and recording venue reverberation.
 
Last edited:
  • Like
Reactions: Duke LeJeune

KlausR.

Well-Known Member
Dec 13, 2010
291
29
333
@KlausR. and @Duke LeJeune I'd be interested on your opinions on the new Polk speakers that cancel the output of the opposite speaker in it's half of the room...

Interaural crosstalk, and methods for its cancellation, are known since the early 60ies, and I have some AES papers on that issue, but never had the time to read them thoroughly, so I don’t have an opinion on that topic.

Two questions come to mind:
1. Have the detrimental effects of that crosstalk ever been proven to exist in controlled listening tests with music?
2. If X-talk is a real problem, then why don't all those speakers costing $$$$$$$ include means for cancelling it? Why don't studio monitors take this into account?

In AES paper 7916 there are some interesting passages:
"Response notches are annoying if not filled in by reflections."
"Response notches are almost inaudible if the notches are filled in by reflections within 10 ms."
"Very dead rooms leave audible comb filtering."
"As mentioned, in most typical consumer listening environments, problems due to notches in the magnitude response are rarely obvious except during controlled intelligibility tests or in very dry environments, because of the diffusion caused by room reflections."

http://www.aes.org/e-lib/browse.cfm?elib=15111


Klaus
 

KlausR.

Well-Known Member
Dec 13, 2010
291
29
333
In the small, reflective room, direct sounds and early first reflections are summed, leading to an increase in sound intensity and saturation. In the larger room those reflections are absorbed so correspondingly greater SPLs are required to achieve the same intensity levels. It can be seen from the above that speaker matching is quite important as it is easy to overload a small room, or under power a large room.

This issue of speaker/amp matching the room or not regularly comes up so some time ago I did some number crunching which gives an idea of what happens in rooms of different sizes, see attached PDF. I had prepared the piece in German and then translated into English, that's why chart and tables are still in German.

Klaus
 

Attachments

  • SPL in 3 rooms.pdf
    174.1 KB · Views: 12

Duke LeJeune

[Industry Expert]/Member Sponsor
Jul 22, 2013
747
1,200
435
Princeton, Texas
Hello Duke, By definition, reflections have lower levels than the direct sound. How then is a reflection capable of masking the direct sound? The mere existence of perception thresholds would be evidence to the contrary?

Thank you for your in-depth reply, Klaus.

Perception thresholds do not preclude partial masking, or more precisely, partial backwards masking.

I hear a loss of clarity when the early reflections are included in those Griesinger clips, and consider that to be evidence of masking under those circumstances. Of course what I claim to hear is merely "anecdotal", so I cannot expect my claim of hearing anything on those clips to constitute evidence for anyone other than myself.

It is not clear (to me) what exactly the recording was: was it a Lokki recording with electronic orchestra? Did Griesinger make the recording himself, and where?

I don't know the details of how Griesinger made the recording but I think it was in Boston Symphony, from the seat he had a ticket for, and I speculate that he used a binaural microphone setup with tiny microphones in his own ears.

I really don’t think that one can take this particular recording of a recording of concert hall sound as evidence for early reflections having detrimental effects of reproduce sound in a domestic setting.

The basic principles in play are these, as described by Griesinger:

"Envelopment is perceived when the ear and brain can detect TWO separate streams: A foreground stream of direct sound, and a background stream of reverberation. Both streams must be present if sound is perceived as enveloping."

"The earlier a reflection arrives the more it contributes to masking the direct sound.”

Griesinger's clips are an illustration of these principles in action in a concert hall.

Because of the shorter timeframes, my guess is that a corresponding set of recordings in a home listening room would not show as dramatic a difference.

...there is no evidence, as far as I’m aware, for detrimental effects of these [early] reflections.

Have you ever owned fullrange dipole speakers, like Maggies? If so, did you notice an improvement in clarity from pulling the speakers farther out from the wall? If you did, I suggest that was evidence of the detrimental effects of early reflections, and of the benefits of later-onset reflections.

In one of his listening rooms Toole used a diagonal speaker placement, which geometrically minimizes early sidewall reflections. What he has to say about that configuration certainly implies a lack of early sidewall reflections is desirable:

“This was deliberate. I think many people are unaware of the advantages of a diagonal arrangement. There are essentially no sidewall reflections.”

Again quoting Toole, from the third edition of "Sound Reproduction", page 194:

"Go into almost any recording control room and there is a high probability that early lateral reflections have been eliminated or attenuated by appropriately angled reflecting surfaces or massive absorbers, or both.... Using only professional sound engineers as listeners, [researchers] found that narrow-dispersion loudspeakers were required for good reproduction of voices in radio dramas; dance and popular music were also desirably "aggressive" with "highly directed" loudspeakers. The majority of these same listeners, however, preferred wide-dispersion loudspeakers for the reproduction of symphonic music at home."

In other words, what Toole is saying is that recording engineers, who rely on clarity to do their jobs well, prefer to minimize early reflections when listening on the job. But for recreational listening at home, the majority prefer wide-dispersion speakers. Why might this be?

Here is what I think is going on: The early reflections at home degrade clarity just like they do at work, but a wider pattern results in more late-arriving reflections, and the net benefit of those additional late-arriving reflections outweighs any detriment from the early ones.

If this is indeed the case, then the "best of both worlds" might be the suppression of early reflections, followed by lot of late reflections. This is exactly what Griesinger's clips demonstrate, and imo this is consistent with the widely-reported experiences of dipole owners.

Reduced early reflections followed by increased late-onset reflections (with a corresponding increase in late reflections) is a goal of my loudspeaker designs, with the intensity of those late-onset reflections being user-adjustable.

Absorbers will absorb reflections regardless of delay, so why specifically aim at 10 ms? They [absorbers] do this [absorb regardless of delay] also on floor and ceiling so I don’t quite understand his approach of breaking up those reflections.

My understanding is that 10 ms is just a point on a multi-dimensional continuum wherein level and frequency and direction and properties of the reflecting surface and the presence of other reflections all play roles. In my opinion a 10 ms target seems to be both practical and beneficial. 12 ms is a little bit less practical and a little bit more beneficial. And so forth.

Seigfried Linkwitz came up with 6 ms as his recommended minimum:

"Reflections generated by the two loudspeakers should be delayed copies of the direct sound to the listener. The delay should be greater than 6 ms. The high frequency content of the reflections should not be intentionally attenuated."

One problem with aggressive use of absorption to kill the early reflections is, that energy is then no longer around to become beneficial late reflections. Another problem is, absorption selectively attenuates short wavelengths more than longer ones, altering the spectral balance of the reflections accordingly.

With all respect, as long as no appropriate controls have been included in listening tests, I would consider personal listening experience as anecdotal at best.

Of course my personal listening experience is anecdotal! Please note that I made a point of NOT claiming my own experiences constituted evidence! The only person my experiences can constitute evidence for is myself; for you and anyone else, any conclusions which I draw from my own experiences are opinions. My apologies for not making that clear.

This [the delay time of the early reflections determines the impression of the size of the room] relates to the size of the room in which the recording has been made, not the room in which the recording is listened to.

I think it also applies to the playback room. I think I hear a correlation between soundstage depth and distance between speakers and wall, whether the speakers be monopoles or dipoles.
 
Last edited:

Duke LeJeune

[Industry Expert]/Member Sponsor
Jul 22, 2013
747
1,200
435
Princeton, Texas
In small rooms, the early arriving reflections are summed along with the original sound and only one sound is heard, with the original sound’s direction component, so again it cannot by definition mask the original sound since, as far as the ear is concerned, there is only the original sound.

I agree with the first part but I'm not sure about your conclusion that "[the early arriving reflections] cannot by definition mask the original source."

The early arriving reflections are apparently doing something which is detrimental to the clarity of the original source, and that "something" may or may not be partial backwards masking. If our definitions preclude the possibility of early reflections degrading clarity, and if such degradation actually occurs, then our definitions are not completely accurate.

Thus, you can’t take measurements and their conclusions made in a large hall and apply them to a small room as sounds and reflections in a small room are treated differently psychoacoustically.

Griesinger's clips are evidence that the principles he describes are valid in a large hall. They may or may not be valid in a small room based on his clips alone, but the experiences of dipole loudspeaker owners and of recording engineers are arguably consistent with his principles being applicable in small rooms as well.

This means that in small rooms, late arriving reflections, which would mask the original sound‘s decay and the recording venue’s reverberation should be treated by diffusing them, thereby preventing multiple reflections, while in large rooms, early arriving reflections should be absorbed, because they too will mask note decay and recording venue reverberation.

This makes sense to me.
 

Duke LeJeune

[Industry Expert]/Member Sponsor
Jul 22, 2013
747
1,200
435
Princeton, Texas
Two questions come to mind:
1. Have the detrimental effects of that crosstalk ever been proven to exist in controlled listening tests with music?

There is a perceptual 2 kHz notch in the direct sound of a phantom center image, but reverberation usually fills it in. Imo eliminating this 2 kHz notch is neither the goal nor primary effect of Polk's crosstalk cancellation array.

2. If X-talk is a real problem, then why don't all those speakers costing $$$$$$$ include means for cancelling it? Why don't studio monitors take this into account?

Imo the more relevant question would be, does crosstalk cancellation make a worthwhile net improvement? And imo that is the sort of question best answered by actual listening.
 

KlausR.

Well-Known Member
Dec 13, 2010
291
29
333
Hallo Duke,

I hear a loss of clarity when the early reflections are included in those Griesinger clips, and consider that to be evidence of masking under those circumstances. Of course what I claim to hear is merely "anecdotal", so I cannot expect my claim of hearing anything on those clips to constitute evidence for anyone other than myself.]/quote]

As long as the delays of the reflections are not known one can only speculate about what psychoacoustic effect is at work.

Perception thresholds do not preclude partial masking, or more precisely, partial backwards masking.

I had a very quick look in JASA for backward masking and found Pickett, “Backward masking”, JASA 1959, p.1613. In general the masker (noise burst) has to be much louder than the signal (sine tone), how much louder depends on signal duration, duration of interval between signal and masker, frequency relation between masker and signal. Now, how can a reflection which by definition is lower in level than the signal, achieve backward masking?

The basic principles in play are these:
"Envelopment is perceived when the ear and brain can detect TWO separate streams: A foreground stream of direct sound, and a background stream of reverberation. Both streams must be present if sound is perceived as enveloping."
"The earlier a reflection arrives the more it contributes to masking the direct sound.”
Griesinger's clips are an illustration of these principles in action in a concert hall.
Because of the shorter timeframes, my guess is that a corresponding set of recordings in a home listening room would not show as dramatic a difference.

The only way to know is to make such recordings, and then listen with test controls implemented. I think that, just as Blackmorec has pointed out, because of these shorter timeframes, the situation is very different psychoacoustically so that the only was to know is to make a recording in a home and listen to it.

Have you ever owned fullrange dipole speakers, like Maggies? If so, did you notice an improvement in clarity from pulling the speakers farther out from the wall? If you did, I suggest that was evidence of the detrimental effects of early reflections, and of the benefits of later-onset reflections.

I don’t know enough about dipoles to comment on that. But what is true for Maggies must not necessarily be true for other speakers?

In one of his listening rooms Toole used a diagonal speaker placement, which geometrically minimizes early sidewall reflections. What he has to say about that configuration certainly implies a lack of early sidewall reflections is desirable:

“This was deliberate. I think many people are unaware of the advantages of a diagonal arrangement. There are essentially no sidewall reflections.”

It certainly makes sense for speakers with horrible off-axis response! Or for strong acoustic asymmetry.

Again quoting Toole, from the third edition of "Sound Reproduction", page 194:
"Go into almost any recording control room and there is a high probability that early lateral reflections have been eliminated or attenuated by appropriately angled reflecting surfaces or massive absorbers, or both....

Tool also says in his book (emphasis added): “Nevertheless, there are some who will insist on eliminating those pesky reflections as a matter of ritual. The ritual had its origins in recording control rooms—listening in stereo—justified by alarmist cautions of comb filtering (see Chapter 9) or degraded speech intelligibility (see Chapter 10). These are not problems. The real factor appears to be spaciousness and the possibility that recording engineers, like musicians, are many times more sensitive to it and the reflections causing it than ordinary folk (see Section 8.1). They appear to feel that their work is impeded by lateral reflections, but many (most?) of them prefer to have them in place for recreational listening.”

However, in section 8.1 Toole refers to Ando et al. (2000), “Formulae describing subjective attributes for sound fields based on a model of the auditory-brain system”. In this paper reference is made to two papers in the context of early reflection delays: Nakayama, “Preferred time delay of a single reflection for performers” and Sato, “Subjective preference of cellists for the delay time of a single reflection in a performance”.

Nakayama put the musicians in an anechoic room playing their instrument with the reflection coming from diagonally above. Sato put the musicians in an anechoic room playing their instrument with the reflection coming from behind. Based on these tests Ando and then Toole draw the conclusion that musicians, and perhaps also recording engineers, are more sensitive to reflections, which is why the latter need a reflection-free zone. Judge yourself!

Using only professional sound engineers as listeners, [researchers] found that narrow-dispersion loudspeakers were required for good reproduction of voices in radio dramas; dance and popular music were also desirably "aggressive" with "highly directed" loudspeakers. The majority of these same listeners, however, preferred wide-dispersion loudspeakers for the reproduction of symphonic music at home."

Here Toole refers to the 1978 Kuhl paper. However, this paper further states: “Half of the test subjects thinks they could use the wide-dispersion speaker also for recordings of symphonic music whereas the other half preferred the narrow-dispersion speaker.”

This paints a somewhat different picture, does it not?

FYI: in their paper Kuhl & Plantz compared two speakers with directivity index of 6.1 and 8.8.

Siegfried Linkwitz came up with 6 ms as his recommended minimum:

Linkwitz AES paper 7162 mentions these 6 ms, but I can’t see the reasoning this recommendation is based upon. It would be interesting to know where Geddes’ 10 ms come from.

One problem with aggressive use of absorption to kill the early reflections is, that energy is then no longer around to become beneficial late reflections. Another problem is, absorption selectively attenuates short wavelengths more than longer ones, altering the spectral balance of the reflections accordingly.

Agressive absorption further has the effect of rendering those reflections that have not been absorbed more audible. But as long as the result is pleasing to the ear ...

Re: Polk speakers. Their patents US 10,327,086 and 10,327,064 should be able to explain in detail what’s ging on.

Klaus
 

About us

  • What’s Best Forum is THE forum for high end audio, product reviews, advice and sharing experiences on the best of everything else. This is THE place where audiophiles and audio companies discuss vintage, contemporary and new audio products, music servers, music streamers, computer audio, digital-to-analog converters, turntables, phono stages, cartridges, reel-to-reel tape machines, speakers, headphones and tube and solid-state amplification. Founded in 2010 What’s Best Forum invites intelligent and courteous people of all interests and backgrounds to describe and discuss the best of everything. From beginners to life-long hobbyists to industry professionals, we enjoy learning about new things and meeting new people, and participating in spirited debates.

Quick Navigation

User Menu

Steve Williams
Site Founder | Site Owner | Administrator
Ron Resnick
Site Co-Owner | Administrator
Julian (The Fixer)
Website Build | Marketing Managersing