Perceptual Effects of Room Reflections

[josh358]

Just wanted to add that the main issue is that head tracking doesn’t track you while walking or even moving on the chair but it just tracks your head inclination. It’s a very weird effect !

That’s why IMO you need to render sound on the fly. To avoid this unpleasant effect

Good point! Tracking can be done with eye detection -- I suppose, these days, AI detection of the position of the head, as well.

 

[eliash]

Really, neither is right. The headphones put the music inside your heard (assuming it isn't a binaural recording), room reflections impose a "one size fits all" acoustic on everything and limit the width of stereo image because of interaural crosstalk. Really, to do better, you need something more sophisticated than plain old stereo. Crosstalk cancellation with head tracking and HRTF compensation may be the ultimate (wave field synthesis, too, but that's less practical). After that, object-oriented systems with multiple speakers, or multiple speakers with convolving reverb, and after that conventional stereo. Head tracking and HRTF compensation could be even better than the same with loudspeakers and crosstalk cancellation.

regarding "...room reflections impose a "one size fits all" acoustic on everything and limit the width of stereo image because of interaural crosstalk..." my impression is that this also only half true. Having installed some wall absorbers in the "mirror points" against first reflections (from both sides, ceiling and floor -thick carpet-) helps a lot to "understand" the recording room's acoustical setup. A good proof for its effectiveness is playing mono. As long as the complete "sound picture" stays in the the center, one can be sure to obtain a lot of the original sound structure of a stereo recording (cross-check with H/Ps plus eventually some cross-feed). In case when the mono sound picture "unravels", from the center then your ears won't be able obtain much of the original spatial information from a stereo recording. Besides, when listening "life", there is always interaural crosstalk present as well. When looking at the mashup of late listening room reflections and eventually also room echoes, this will of course add a "sound flavor" to the recording, but from my perspective will not necessarily change the perceived original spatial information.

A good example is listening to ECM´s Avishai Cohen & Yonathan Avishai "Playing The Room" were one can get the impression to "look through a door" into the recording room (which seems fairly big, looking at the album photos)...

 

[josh358]

regarding "...room reflections impose a "one size fits all" acoustic on everything and limit the width of stereo image because of interaural crosstalk..." my impression is that this also only half true. Having installed some wall absorbers in the "mirror points" against first reflections (from both sides, ceiling and floor -thick carpet-) helps a lot to "understand" the recording room's acoustical setup. A good proof for its effectiveness is playing mono. As long as the complete "sound picture" stays in the the center, one can be sure to obtain a lot of the original sound structure of a stereo recording (cross-check with H/Ps plus eventually some cross-feed). In case when the mono sound picture "unravels", from the center then your ears won't be able obtain much of the original spatial information from a stereo recording. Besides, when listening "life", there is always interaural crosstalk present as well. When looking at the mashup of late listening room reflections and eventually also room echoes, this will of course add a "sound flavor" to the recording, but from my perspective will not necessarily change the perceived original spatial information.

A good example is listening to ECM´s Avishai Cohen & Yonathan Avishai "Playing The Room" were one can get the impression to "look through a door" into the recording room (which seems fairly big, looking at the album photos)...

I agree that a solid mono image a prerequisite of good stereo. But I think lateral reflections are bit more complex. IIRC, in experiments, listeners prefer reflections, and that they prefer them from the side, which results in an increase in perceived image width. Personally, I don't listen that way -- I have dipole speakers, and first reflection points on the sidewall are in the null. But the research suggests that most people do prefer those reflections.

Again, personally, my image is superb, even stunning -- but it is largely confined to the space between the speakers, with perhaps a bit of help from the corners on the front wall, which are in the dipole maximum.

OK, so room acoustics -- we determine the size of venue largely by the timing of reflections compared to the original sound, which is to say that the ear will get a very different spatial impression in a small room than a cathedral. The amplitude of reflections is also important -- it's the difference between singing in a bathroom and outdoors. A listening room with a low RT60 (nominal, since it doesn't really apply to small spaces) will sound different from a room with a high one. To most closely emulate the original sound, the initial time delay and length and shape of the reverberant tail should be reproduced, but that can't be done in a small room unless it's in the recording, as it sometimes but not always is.
Unfortunately, even if the recording contains natural or artificial reverberation, the room can overwhelm the image on the recording, and so negatively impact our sense of space. That's why I say that one size fits all doesn't work.

There are some ways around this. You can absorb in the reflections from the speakers, for a very dry, pinpoint, headphone-like sound. The problem is, it doesn't sound very real or IMO satisfying. You can use tricks to delay the first reflection of the room by >20 ms -- the initial time gap of a good concert hall is 20-25 ms, so the room's reverb won't mask it. You can also use a compromise solution, to whit, diffusion, which can reduce the initial reflection below the level at which it colors the sound while preserving the reverberant energy so the room doesn't sound too dry.

The problem with absorbing reflections of the speakers is that the ear can differentiate between reverberation coming from the speakers and reverberation arriving at different angles.

Re interaural crosstalk -- in real life, the timing changes depending on angle, from zero directly in front to very roughly a millisecond at 180 degrees. And the brain uses that to help determine the angle from which a sound arrives. A loudspeaker at a 45 degree angle corresponds to a sound at a 45 degree angle, and two loudspeakers at a 45 degree angle corresponds to unnatural cues. The only thing that really works is a sound panned to 45 degrees. In practice, this sets an upper limit on maximum angle. In the case of the standard 45 degree angle, we can't hear an image beyond 45 degrees from the center, even though the stereo recording contains a 180 degree image. Crosstalk cancellation removes this limitation, and the image spreads out beyond the speakers to something approaching a full 180 degree soundstage.

 

[eliash]

I agree that a solid mono image a prerequisite of good stereo. But I think lateral reflections are bit more complex. IIRC, in experiments, listeners prefer reflections, and that they prefer them from the side, which results in an increase in perceived image width. Personally, I don't listen that way -- I have dipole speakers, and first reflection points on the sidewall are in the null. But the research suggests that most people do prefer those reflections.

Again, personally, my image is superb, even stunning -- but it is largely confined to the space between the speakers, with perhaps a bit of help from the corners on the front wall, which are in the dipole maximum.

OK, so room acoustics -- we determine the size of venue largely by the timing of reflections compared to the original sound, which is to say that the ear will get a very different spatial impression in a small room than a cathedral. The amplitude of reflections is also important -- it's the difference between singing in a bathroom and outdoors. A listening room with a low RT60 (nominal, since it doesn't really apply to small spaces) will sound different from a room with a high one. To most closely emulate the original sound, the initial time delay and length and shape of the reverberant tail should be reproduced, but that can't be done in a small room unless it's in the recording, as it sometimes but not always is.
Unfortunately, even if the recording contains natural or artificial reverberation, the room can overwhelm the image on the recording, and so negatively impact our sense of space. That's why I say that one size fits all doesn't work.

There are some ways around this. You can absorb in the reflections from the speakers, for a very dry, pinpoint, headphone-like sound. The problem is, it doesn't sound very real or IMO satisfying. You can use tricks to delay the first reflection of the room by >20 ms -- the initial time gap of a good concert hall is 20-25 ms, so the room's reverb won't mask it. You can also use a compromise solution, to whit, diffusion, which can reduce the initial reflection below the level at which it colors the sound while preserving the reverberant energy so the room doesn't sound too dry.

The problem with absorbing reflections of the speakers is that the ear can differentiate between reverberation coming from the speakers and reverberation arriving at different angles.

Re interaural crosstalk -- in real life, the timing changes depending on angle, from zero directly in front to very roughly a millisecond at 180 degrees. And the brain uses that to help determine the angle from which a sound arrives. A loudspeaker at a 45 degree angle corresponds to a sound at a 45 degree angle, and two loudspeakers at a 45 degree angle corresponds to unnatural cues. The only thing that really works is a sound panned to 45 degrees. In practice, this sets an upper limit on maximum angle. In the case of the standard 45 degree angle, we can't hear an image beyond 45 degrees from the center, even though the stereo recording contains a 180 degree image. Crosstalk cancellation removes this limitation, and the image spreads out beyond the speakers to something approaching a full 180 degree soundstage.

There is not necessarily a contradiction between our views.
The issue from my perspective is when virtually widening up the stereo base width by first side wall reflections there is indeed a first impression of a "nicer" sound. I am using that fact in our large living room (for european standards) which has lots of glass windows, hard walls, corners and furniture, so lots of irregular reflections. Those can´t be altered significantly due to WAF and actually by using transmission line type speakers (Rega Ela´s) the weak bass actually sounds acceptable because of the heavy room resonances and wall amplification. The down side is missing sound field precision due to asymmetrical room layout and listening position. This is the same situation in my "small" listening room (wall on one side, window on the other, protruding chimney, tilted ceilings, large room volume around the corner), so in both cases, even with symmetrical-to-corners speaker setup and precise stereo triangle listening position, without first reflection absorption, a stable mono sound field cannot be reached. In such a case one needs to make the decision to stay with a "nice" immersive sound field (like Toole et al explain), where incidental spatial impressions also come from points far outside of the stereo base or a more precise studio control room approach, where the stereo sound stays inside the stereo base and allows better spatial instrument recognition. In the small room there was also an effect of painful perceived distortion (call it roughness, even if the scientific definition is different) also from the very short wall delay path without absorption, which literally forced me to explore the issue...

 

[josh358]

There is not necessarily a contradiction between our views.
The issue from my perspaective is when virtually widening up the stereo base width by first side wall reflections there is indeed a first impression of a "nicer" sound. I am using that fact in our large living room (for european standards) which has lots of glass windows, hard walls, corners and furniture, so lots of irregular reflections. Those can´t be altered significantly due to WAF and actually by using transmission line type speakers (Rega Ela´s) the weak bass actually sounds acceptable because of the heavy room resonances and wall amplification. The down side is missing sound field precision due to asymmetrical room layout and listening position. This is the same situation in my "small" listening room (wall on one side, window on the other, protruding chimney, tilted ceilings, large room volume around the corner), so in both cases, even with symmetrical-to-corners speaker setup and precise stereo triangle listening position, without first reflection absorption, a stable mono sound field cannot be reached. In such a case one needs to make the decision to stay with a "nice" immersive sound field (like Toole et al explain), where incidental spatial impressions also come from points far outside of the stereo base or a more precise studio control room approach, where the stereo sound stays inside the stereo base and allows better spatial instrument recognition. In the small room there was also an effect of painful perceived distortion (call it roughness, even the scientific definition is different) also from the very short wall delay path without absorption, which literally forced me to explore the issue...

That's been my experience as well - that when you suppress reflections, you get stable, pinpoint imaging. The most spectacular example of that I've ever heard was the IRS V in a treated room at Lyric HiFi in New York -- it was literally holographic, you felt like you could walk between the instruments. But I've never been wild about the sound in recording studios. It could be astoundingly real on single instrument, noise, voice, to the extent that I somehow thought something was in the control room when it was in the studio. But it just didn't sound like the real thing when listening to large scale works.

That raises a conundrum -- not to put too fine a point on it, you have to compromise between spaciousness and localization. And pinpoint localization really isn't natural for acoustical music. But it's been said that you need that localization when you lack visual cues, and I think there's a something in that.

Fascinating comment on "roughness." I won't dismiss the terminology because I know exactly what you mean and have struggled with it myself on several occasions. You'd think diffusion would fix it, but in my experience, anyway, it doesn't.

 

[eliash]

That's been my experience as well - that when you suppress reflections, you get stable, pinpoint imaging. The most spectacular example of that I've ever heard was the IRS V in a treated room at Lyric HiFi in New York -- it was literally holographic, you felt like you could walk between the instruments. But I've never been wild about the sound in recording studios. It could be astoundingly real on single instrument, noise, voice, to the extent that I somehow thought something was in the control room when it was in the studio. But it just didn't sound like the real thing when listening to large scale works.

That raises a conundrum -- not to put too fine a point on it, you have to compromise between spaciousness and localization. And pinpoint localization really isn't natural for acoustical music. But it's been said that you need that localization when you lack visual cues, and I think there's a something in that.

Fascinating comment on "roughness." I won't dismiss the terminology because I know exactly what you mean and have struggled with it myself on several occasions. You'd think diffusion would fix it, but in my experience, anyway, it doesn't.

Actually, ideal conditions in one or the other extreme (precise immersive vs. studio monitoring sound) are hardly reached, so one can experiment and work towards a status when peace of mind or better said listening is reached.

Btw.: Regarding that strange "roughness" effect, I came to the conclusion that it is caused by speaker phase modulation of audible mid and higher frequencies due to bass driver induced enclosure vibrations (elastic enclosure tilting). In my case the dynamic driving forces of 2 bass/mid drivers in a tower enclosure (Dynaudio Focus 220) also easily exceeded the tilting force at the enclosure top. Phase modulation itself seems rather inaudible, but with strong side wall reflections wave superposition converts it into (rather audible) amplitude modulation. The solutions which I applied controlled this roughness effect to an acceptable degree:
- 1st is a stiff enclosure coupling to the plaster floor layer by through-holing the wobbly wooden floor (large steel pads underneath the enclosure spikes already showed a significant positive effect).
- 2nd was installing an additonal mass of 3kg in the rear top of the speaker enclosure (by "externalising" the crossover) to at least balance the top front and rear enclosure tilting force.
- 3rd were the a. m. first reflection absorption measures...

 

[Duke]

when you suppress reflections, you get stable, pinpoint imaging... But I've never been wild about the sound in recording studios... just didn't sound like the real thing when listening to large scale works.

That makes total sense to me.

That raises a conundrum -- not to put too fine a point on it, you have to compromise between spaciousness and localization. And pinpoint localization really isn't natural for acoustical music. But it's been said that you need that localization when you lack visual cues, and I think there's a something in that.

According to Griesinger, a good seat in a good concert hall has BOTH spaciousness and localization. He uses different wording but talks about there being a distance beyond which we can no longer perceptually separate two violins playing the same part because the reverberant field has become too strong relative to the direct sound. Obviously the acoustics in our playback rooms result in a much stronger direct-to-reverberant sound ratio, but also in much earlier arrival times for the reflections. Geddes has stated that the earlier a reflection arrives and the stronger it is, the more detrimental it is, BUT he also points out that reflections are needed to give liveliness to the sound, consistent with what you said about (presumably heavily damped) recording studios.

Imo one thing to keep in mind is that there are TWO sets of acoustics presented in the playback room: The playback room's acoustics, AND the acoustics of the recording venue (whether the latter be real or engineered or both). Our playback rooms will never be able to replicate the acoustics of a much larger venue, but if we can minimize the playback room's signature while effectively presenting the venue cues on the recording, the latter can become perceptually dominant, resulting in a "you are there" experience.

And I agree with you that precise localization is arguably a very enjoyable surrogate for visual cues.

The issue from my perspective is when virtually widening up the stereo base width by first side wall reflections there is indeed a first impression of a "nicer" sound.... The down side is missing sound field precision... In such a case one needs to make the decision to stay with a "nice" immersive sound field (like Toole et al explain), where incidental spatial impressions also come from points far outside of the stereo base OR a more precise studio control room approach, where the stereo sound stays inside the stereo base and allows better spatial instrument recognition.[emphasis Duke's]

Very well said.

When we minimize the early same-side-wall reflections, we DO lose the enjoyable increase in Apparent Source Width in order to get a more "precise" sound field. And if we use absorption to accomplish that, we have also arguably degraded the (imo generally beneficial) later reflections by weakening them and by making their tonal balance duller, the latter based on the assumption that absorption is more effective at short wavelengths than at longer ones. It's a juggling of tradeoffs... but the more educated the juggler, the better the outcome is likely to be.

Personally I advocate relatively narrow-pattern speakers, toed-in aggressively to avoid early same-side-wall reflections without reliance on sidewall absorption, and with the speakers spaced a bit wider than normal to get good soundstage width. BUT with such a setup, sound images do not appear outside of the width of the speakers UNLESS they have been processed by the recording engineer to do so.

 

[375HP2482]

@oivavoi @amirm I dare say that the right balance would be modern studio recordings with headphones and acoustic events with speakers!

It’s interesting how (a good chunk of) headphones are characterized by limited number of reflections compared to room listening through speakers.

Nice article anyway Amir. Just reading some of your old ones!

I have found headphones are better for mixing simply because you can hear more detail due to the absence of room reflections. However, though balancing instruments is easier on headphones, reverb is needed to fill out the dry sound. Then, what may sound like a proper amount of reverb in a headset will sound excessive on loudspeakers, as their reflective environment itself adds a lot of (undesirable) reverb of its own.

So: Balance (mix) with headphones, polish (effects) and master (eq) on speakers.

 

[eliash]

That makes total sense to me.




According to Griesinger, a good seat in a good concert hall has BOTH spaciousness and localization. He uses different wording but talks about there being a distance beyond which we can no longer perceptually separate two violins playing the same part because the reverberant field has become too strong relative to the direct sound. Obviously the acoustics in our playback rooms result in a much stronger direct-to-reverberant sound ratio, but also in much earlier arrival times for the reflections. Geddes has stated that the earlier a reflection arrives and the stronger it is, the more detrimental it is, BUT he also points out that reflections are needed to give liveliness to the sound, consistent with what you said about (presumably heavily damped) recording studios.

Imo one thing to keep in mind is that there are TWO sets of acoustics presented in the playback room: The playback room's acoustics, AND the acoustics of the recording venue (whether the latter be real or engineered or both). Our playback rooms will never be able to replicate the acoustics of a much larger venue, but if we can minimize the playback room's signature while effectively presenting the venue cues on the recording, the latter can become perceptually dominant, resulting in a "you are there" experience.

And I agree with you that precise localization is arguably a very enjoyable surrogate for visual cues.




Very well said.

When we minimize the early same-side-wall reflections, we DO lose the enjoyable increase in Apparent Source Width in order to get a more "precise" sound field. And if we use absorption to accomplish that, we have also arguably degraded the (imo generally beneficial) later reflections by weakening them and by making their tonal balance duller, the latter based on the assumption that absorption is more effective at short wavelengths than at longer ones. It's a juggling of tradeoffs... but the more educated the juggler, the better the outcome is likely to be.

Personally I advocate relatively narrow-pattern speakers, toed-in aggressively to avoid early same-side-wall reflections without reliance on sidewall absorption, and with the speakers spaced a bit wider than normal to get good soundstage width. BUT with such a setup, sound images do not appear outside of the width of the speakers UNLESS they have been processed by the recording engineer to do so.

I tried your toeing-in approach, when I set up the a. m. speaker in my listening room (actually they came from the a. m. living room where their bass was simply too strong for its strong multiple room resonances). Toeing-in and even tilting in the smaller listening room did not really help on that "roughness" perception, probably because their radiation angle is quite wide (as it should be, according to the standards Amir cites in his speaker tests). So I had wished for some narrower radiation pattern...

 

[Blaspheme]

I tried your toeing-in approach, when I set up the a. m. speaker in my listening room (actually they came from the a. m. living room where their bass was simply too strong for its strong multiple room resonances). Toeing-in and even tilting in the smaller listening room did not really help on that "roughness" perception, probably because their radiation angle is quite wide (as it should be, according to the standards Amir cites in his speaker tests). So I had wished for some narrower radiation pattern...

There's a reason they like to listen to those poorly designed speakers in mono ...

Apropos this—going way back in the thread—there was some discussion of speaker preference per the classic Toole visualisation. I've noticed people perceive and conceptualise the same information differently. Conventionally, we have visual, auditory and kinaesthetic learning styles; along with verbal-linguistic, logical-mathematical, musical, spatial-visual, body-kinesthetic, interpersonal, intrapersonal, naturalist and existentialist intelligence types. I have dominant visual-spatial, naturalistic and verbal-linguistic intelligence, whereas my musical and body-kinesthetic is well down and I'm usually lazy and unfocused with the mathematical side of logical-mathematical (you may want a paper with some discussion on the terminology, I'd post the schematic but I think they fucked it up).

... so when I see this, I don't think 1,2,3 but see spatial relationships. And I wonder why Toole truncated the vertical scale and didn't use error bars. Because the average points are in the same place for sound quality for Rega and KEF in moon vs stereo (the KEF has a bit less spread in the latter) and because I know a bit about the speakers the main message I see is that stereo is required for spatial for all the speakers, and the directional Quads need it for both sound quality and spatial. Which isn't "better discrimination in mono". There is likely more data on more speaker preferences to support that interpretation, otherwise we wouldn't be hearing it (I assume). And by all means go ahead and use mono as part of a test regime (just don't stop there).

My point though is just using this as a familiar example to consider possible relationship between intelligence types and perception, and speculatively whether this is a significant factor in how we perceive audio (and perceive it as differently as we do). Do logical-mathematical types not perceive spatial characteristics of audio as vividly as visual-spatial types? What about body-kinesthetic or musical types: are dynamics and rhythm more important? Some of you may have already thought about this more than I have, in which case I'd be interested in experience and research that you may have.

 

[josh358]

According to Griesinger, a good seat in a good concert hall has BOTH spaciousness and localization. He uses different wording but talks about there being a distance beyond which we can no longer perceptually separate two violins playing the same part because the reverberant field has become too strong relative to the direct sound. Obviously the acoustics in our playback rooms result in a much stronger direct-to-reverberant sound ratio, but also in much earlier arrival times for the reflections. Geddes has stated that the earlier a reflection arrives and the stronger it is, the more detrimental it is, BUT he also points out that reflections are needed to give liveliness to the sound, consistent with what you said about (presumably heavily damped) recording studios.

Imo one thing to keep in mind is that there are TWO sets of acoustics presented in the playback room: The playback room's acoustics, AND the acoustics of the recording venue (whether the latter be real or engineered or both). Our playback rooms will never be able to replicate the acoustics of a much larger venue, but if we can minimize the playback room's signature while effectively presenting the venue cues on the recording, the latter can become perceptually dominant, resulting in a "you are there" experience.

And I agree with you that precise localization is arguably a very enjoyable surrogate for visual cues.

In my experience, you have to be very close to the stage to get that kind of localization. IIRC, the far field (the point at which reverberant energy constitutes more than 50% of what we hear) is at something like Row H. And when you're that close to the orchestra, the presentation is substantially different from what a conventional two-channel system, with its limited width, can reproduce, unless you use crosstalk cancellation.

Since the best concert halls have an initial time delay gap of 20-25 ms, speakers should be at least 10 feet out from surfaces. Needless to say, this isn't achievable in most rooms. In my experience, if they're closer than that, the recorded acoustical space seems smaller. There can also be an uncomfortable "double image" effect in which both the original and listening rooms seem to be present at once. It's hard to describe what it sounds like -- "constipated" might be useful -- as you move the speakers out, the sound seems to bloom, open up, become more natural.

If you delay the reflections and the recording was made in a smaller room, the ear hears the ITDG of the venue before that of the listening room, and the venue seems smaller. It is changed some by the secondary arrival.

I think that the best treatment for a small room is diffusion. It suppresses the early reflections without removing energy from the reverberant field. With directional speakers, absorption tends to make the room too dead. Per Linkwitz, dipoles do best with the RT60 of a typical living room. With conventional omni/cardioid boxes, though, some absorption is needed for the best results, doubly so if the room is large, which increases the RT60 and at a certain point can cause discrete echoes.

Finally, one of the attributes of a good concert hall is low interaural cross correlation. Again, diffusion can help with that.

At the end of the day, you're shuffling deck chairs on the Titanic, but some results seem more pleasing than others!

 

[josh358]

There's a reason they like to listen to those poorly designed speakers in mono ...

Apropos this—going way back in the thread—there was some discussion of speaker preference per the classic Toole visualisation. I've noticed people perceive and conceptualise the same information differently. Conventionally, we have visual, auditory and kinaesthetic learning styles; along with verbal-linguistic, logical-mathematical, musical, spatial-visual, body-kinesthetic, interpersonal, intrapersonal, naturalist and existentialist intelligence types. I have dominant visual-spatial, naturalistic and verbal-linguistic intelligence, whereas my musical and body-kinesthetic is well down and I'm usually lazy and unfocused with the mathematical side of logical-mathematical (you may want a paper with some discussion on the terminology, I'd post the schematic but I think they fucked it up).

View attachment 139261

... so when I see this, I don't think 1,2,3 but see spatial relationships. And I wonder why Toole truncated the vertical scale and didn't use error bars. Because the average points are in the same place for sound quality for Rega and KEF in moon vs stereo (the KEF has a bit less spread in the latter) and because I know a bit about the speakers the main message I see is that stereo is required for spatial for all the speakers, and the directional Quads need it for both sound quality and spatial. Which isn't "better discrimination in mono". There is likely more data on more speaker preferences to support that interpretation, otherwise we wouldn't be hearing it (I assume). And by all means go ahead and use mono as part of a test regime (just don't stop there).

My point though is just using this as a familiar example to consider possible relationship between intelligence types and perception, and speculatively whether this is a significant factor in how we perceive audio (and perceive it as differently as we do). Do logical-mathematical types not perceive spatial characteristics of audio as vividly as visual-spatial types? What about body-kinesthetic or musical types: are dynamics and rhythm more important? Some of you may have already thought about this more than I have, in which case I'd be interested in experience and research that you may have.

Interesting. I'd never thought to see this in terms of cognitive styles. I have observed over the years that while most people perceive the same things, people differ in what they most value. Musical preferences matter here -- the criteria for studio rock are different than the criteria for acoustical music, say -- as does familiarity with live music (those who aren't familiar with live acoustical music prefer boosted highs, for example). That last I think is important. I can sit my late father's third wife down, for example, and play something, and she makes astute observations, despite the fact that she knows nothing about audio and isn't at all interested in it. That's because she regularly attends concerts and knows what they sound like.

 

[Spkrdctr]

I tried your toeing-in approach, when I set up the a. m. speaker in my listening room (actually they came from the a. m. living room where their bass was simply too strong for its strong multiple room resonances). Toeing-in and even tilting in the smaller listening room did not really help on that "roughness" perception, probably because their radiation angle is quite wide (as it should be, according to the standards Amir cites in his speaker tests). So I had wished for some narrower radiation pattern...

Could you just easily EQ the Bass hump or humps out of it and Voila, have a much cleaner, better sound? I would try that route.

 

[Spkrdctr]

There's a reason they like to listen to those poorly designed speakers in mono ...

Apropos this—going way back in the thread—there was some discussion of speaker preference per the classic Toole visualisation. I've noticed people perceive and conceptualise the same information differently. Conventionally, we have visual, auditory and kinaesthetic learning styles; along with verbal-linguistic, logical-mathematical, musical, spatial-visual, body-kinesthetic, interpersonal, intrapersonal, naturalist and existentialist intelligence types. I have dominant visual-spatial, naturalistic and verbal-linguistic intelligence, whereas my musical and body-kinesthetic is well down and I'm usually lazy and unfocused with the mathematical side of logical-mathematical (you may want a paper with some discussion on the terminology, I'd post the schematic but I think they fucked it up).

View attachment 139261

... so when I see this, I don't think 1,2,3 but see spatial relationships. And I wonder why Toole truncated the vertical scale and didn't use error bars. Because the average points are in the same place for sound quality for Rega and KEF in moon vs stereo (the KEF has a bit less spread in the latter) and because I know a bit about the speakers the main message I see is that stereo is required for spatial for all the speakers, and the directional Quads need it for both sound quality and spatial. Which isn't "better discrimination in mono". There is likely more data on more speaker preferences to support that interpretation, otherwise we wouldn't be hearing it (I assume). And by all means go ahead and use mono as part of a test regime (just don't stop there).

My point though is just using this as a familiar example to consider possible relationship between intelligence types and perception, and speculatively whether this is a significant factor in how we perceive audio (and perceive it as differently as we do). Do logical-mathematical types not perceive spatial characteristics of audio as vividly as visual-spatial types? What about body-kinesthetic or musical types: are dynamics and rhythm more important? Some of you may have already thought about this more than I have, in which case I'd be interested in experience and research that you may have.

After thinking about your post and then thinking on it again, I like the last sentence. After all my thinking and ruminating on it, I decided that it was easier to go read a book and take a nap. Such is the joy of retirement! Seriously, that was a read worthy of me being back at work. I think it restarted a bout of work induced PTSD! But, I still enjoyed the post.

 

[eliash]

There's a reason they like to listen to those poorly designed speakers in mono ...

Apropos this—going way back in the thread—there was some discussion of speaker preference per the classic Toole visualisation. I've noticed people perceive and conceptualise the same information differently. Conventionally, we have visual, auditory and kinaesthetic learning styles; along with verbal-linguistic, logical-mathematical, musical, spatial-visual, body-kinesthetic, interpersonal, intrapersonal, naturalist and existentialist intelligence types. I have dominant visual-spatial, naturalistic and verbal-linguistic intelligence, whereas my musical and body-kinesthetic is well down and I'm usually lazy and unfocused with the mathematical side of logical-mathematical (you may want a paper with some discussion on the terminology, I'd post the schematic but I think they fucked it up).

View attachment 139261

... so when I see this, I don't think 1,2,3 but see spatial relationships. And I wonder why Toole truncated the vertical scale and didn't use error bars. Because the average points are in the same place for sound quality for Rega and KEF in moon vs stereo (the KEF has a bit less spread in the latter) and because I know a bit about the speakers the main message I see is that stereo is required for spatial for all the speakers, and the directional Quads need it for both sound quality and spatial. Which isn't "better discrimination in mono". There is likely more data on more speaker preferences to support that interpretation, otherwise we wouldn't be hearing it (I assume). And by all means go ahead and use mono as part of a test regime (just don't stop there).

My point though is just using this as a familiar example to consider possible relationship between intelligence types and perception, and speculatively whether this is a significant factor in how we perceive audio (and perceive it as differently as we do). Do logical-mathematical types not perceive spatial characteristics of audio as vividly as visual-spatial types? What about body-kinesthetic or musical types: are dynamics and rhythm more important? Some of you may have already thought about this more than I have, in which case I'd be interested in experience and research that you may have.

...beyond science and the above classification, as s.o. who generally likes to get a glimpse on" the ghost in the machine" I think I also need to understand what´s going on on-stage when listening to music. Maybe this sets (or maybe also limits) my preference mainly to chamber type of music where I am personally able to follow the musicians in their play. I can enjoy live music best from both the acoustical and visual side in parallel. At home I need to "rebuild" the missing visual picture as if I had my eyes closed in a life presentation. Presumably this is why I prefer this kind of studio monitoring approach with best possible spatial resolution and accept less immersive sound (still being able to acknowledge immersive sound when listening in cathedrals, to large orchestras or rock/pop/electronic music just because of the associated perception). On the other hand I was astonished to experience how effect-rich music e.g. by Yello (specificly older productions) yields a very intense physical impression in such a listening environment, just because of its highly dynamic content, which is probably not so much limited or cut down by superposition of a higher amount of reflections...

 

[eliash]

After thinking about your post and then thinking on it again, I like the last sentence. After all my thinking and ruminating on it, I decided that it was easier to go read a book and take a nap. Such is the joy of retirement! Seriously, that was a read worthy of me being back at work. I think it restarted a bout of work induced PTSD! But, I still enjoyed the post.

Good and hopefully quick recovery!

 

[Spkrdctr]

Good and hopefully quick recovery!

Good as new! Nothing like being retired to take the edge off of things.

 

[eliash]

Could you just easily EQ the Bass hump or humps out of it and Voila, have a much cleaner, better sound? I would try that route.

Sorry I missed that.
Resonances were simply too overwhelming in the large room (including vibrating windows on top of the existing room resonances, in my eyes a simply burned lower frequency range not worth being emphasised under these conditions), so I tried stuffing the port as the manufacturer recommended but that resulted in a rather dull bas and mid sound, all of the anyway limited bass attack was gone with this approach and the resonances stayed. Unfortunately the amp in that room is of the puristic sort, without tone control, so I took them up into my listening room (to replace 35y old 3way Dynaudio speaker kits), where they of course excited heavy room resonances as well. Here I could manage them by careful setup with an eye on specificly unpleasantly exaggerated but common musical notes in the listening position. The fundamental room resonance around 33Hz matched the reflex port resonance almost by the Hz, so this sounded like stage noise all the time. Solution for that was a mechanical port extender which tuned the cabinet some 2Hz lower. I think the speaker benefitted in general from this measure (maybe that´s why it had a mk2 version?). Btw. bass attack was greatly improved by mounting the a. m. 3kg mass in the cabinet rear top, even though I had in mind to improve the mentioned "roughness", which it did, but only to a small amount.

 

[Duke]

In my experience, you have to be very close to the stage to get that kind of localization. IIRC, the far field (the point at which reverberant energy constitutes more than 50% of what we hear) is at something like Row H. And when you're that close to the orchestra, the presentation is substantially different from what a conventional two-channel system, with its limited width, can reproduce, unless you use crosstalk cancellation.

I used a poor example of what Griesinger had in mind, drawing on my faulty memory. So let me quote him:

"As the quartet plays tirelessly on stage we walk backward away from them into the hall... At first it is easy to localize each instrument, and tell which played each note. The sound has an exciting, attention-grabbing quality. As you walk back the sound remains close and exciting. But suddenly in distance range of one or two rows all the instruments blend together into a fuzzy ball of sound. The attention-grabbing effect is gone. We call the distance at which this happens the Limit of Localization Distance, or LLD." (From this paper, second page.)

He reports good localization at Row X, near the center, in Boston Symphony Hall.

Since the best concert halls have an initial time delay gap of 20-25 ms, speakers should be at least 10 feet out from surfaces. Needless to say, this isn't achievable in most rooms. In my experience, if they're closer than that, the recorded acoustical space seems smaller. There can also be an uncomfortable "double image" effect in which both the original and listening rooms seem to be present at once. It's hard to describe what it sounds like -- "constipated" might be useful -- as you move the speakers out, the sound seems to bloom, open up, become more natural.

I have not been aware of the "double image" you describe from a setup which minimizes the early reflections while cultivating later ones (later than 10 milliseconds). Reports I get indicate the spatial presentation changes significantly from one recording to the next, which implies that the recordings' varying spatial cues (rather than the playback room's consistent spatial cues) are dominating perception. Imo this would indicate these later-arriving reflections are indeed acting as effective "carriers" of the venue cues on the recording, probably primarily the reverberation tails. See the last four paragraphs of this post for an example.

I think that the best treatment for a small room is diffusion. It suppresses the early reflections without removing energy from the reverberant field. With directional speakers, absorption tends to make the room too dead. Per Linkwitz, dipoles do best with the RT60 of a typical living room. With conventional omni/cardioid boxes, though, some absorption is needed for the best results, doubly so if the room is large, which increases the RT60 and at a certain point can cause discrete echoes.

This all makes sense to me.

Finally, one of the attributes of a good concert hall is low interaural cross correlation. Again, diffusion can help with that.

Yes, as can cross-firing directional speakers, such that the first significant lateral reflection for the left speaker is off the right-side wall, and vice versa.

At the end of the day, you're shuffling deck chairs on the Titanic, but some results seem more pleasing than others!

Well, some deck chairs were closer to the lifeboats!!

Under what should have been very poor conditions - a small untreated hotel room at Axpona - one listener challenged my schpiel by asking us to play a recording he brought which was made in a concert hall he was familiar with. He later posted this on another forum:

"The recording was a FLAC rip of the CD layer of an RCO Live SACD: Shostakovich—Symphony No. 15; Concertgebouw Orchestra/Bernard Haitink conductor. It's a live recording from March of 2010 (Haitink made a much earlier recording of the same piece with the London Philharmonic; he was the first person to record all the Shostakovich symphonies.) For a couple of years, this has been my go-to symphonic recording when I have just a short time to get a sense of an unfamiliar system. It's an excellent performance, something I can listen to repeatedly without going nuts, which is important at a show. In terms of audiophilia, it's an extremely detailed yet atmospheric representation of an orchestra, with excellent dynamics and fully characterized instrumental colors (bells, solo turns by violin, flute, piccolo, string bass, trumpet, etc.) And—with the right audio gear—it successfully renders the essence of (IMO) one of the greatest 3 or 4 concert halls on earth, the Concertgebouw (thus the orchestra's name) in Amsterdam. I've heard music there, and there's truly a sense of sound being present in the air around you.

"The multichannel program on the RCO Live SACDs (there are dozens) get this last aspect right; so did the Bienville Suite, nearly to the same degree, despite the presence of only two channels. My concern when Duke told me about the rear-firing drivers was that this would impart some generic, Bose-like spaciousness to the recording, but that wasn't the case—what I heard was the unique acoustic signature of the Concertgebouw." [emphasis Duke's]

Here is the room, for context. The quasi-triangular speakers have rear-firing drivers whose responses were tailored to their task. You can also see two of the four small subs we used, peeking out from behind the main speakers. Photo by Scott Hull of Part-Time Audiophile.

 

[Blaspheme]

After thinking about your post and then thinking on it again, I like the last sentence. After all my thinking and ruminating on it, I decided that it was easier to go read a book and take a nap. Such is the joy of retirement! Seriously, that was a read worthy of me being back at work. I think it restarted a bout of work induced PTSD! But, I still enjoyed the post.

Haha pleased for enjoyment, sorry for any PTSD, glad it was short-lived.