The ASR "objectivists tribe" ;) has to read the Keynotes at the AES160th by Lars Risbo

[Suono]

@TimVG I don't understand why you're addressing me. I agree with everything, except that I can question preferences as such—preferences, that is. I've had my own experiences with various setups, and I'm convinced of what I write.

 

[TimVG]

@TimVG I don't understand why you're addressing me. I agree with everything, except that I can question preferences as such—preferences, that is. I've had my own experiences with various setups, and I'm convinced of what I write.

Hi. I did not address you personally, otherwise I would have quoted you. It was a general remark aimed at the last few pages, which has little to do anymore with the AES keynotes.

 

[Gorgonzola]

Excerpt from ChatGPT's response within Amir's post ...

THD as a “surrogate marker”​

The article compares audio metrics like THD to surrogate markers in medicine. THD is useful, measurable, and worth optimizing, but it is not the final endpoint. The final endpoint in audio is the listening experience. PURIFI’s warning is that the industry may be “treating the numbers” instead of the actual perceptual result. (PURIFI)

I hope I haven't misunderstood, but it's my impression that Amir himself has touted THD (or SINAD) as a "surrogate marker".

 

[MaxwellsEq]

Regarding the debate both for and against listening-room side-wall reflections: I believe these listening-room reflections are not on the master tape.

The master tape, if it's a relatively simply-mic-ed recording will have captured the side-wall reflections (and all other reflections) of the recording venue. Multi-microphone installation will have captured many different reflections because of their locations. Studio multi-take recordings will have some room-reflections but also many artificial reverb reflections.

 

[MattHooper]

What exactly is empirical about theorizing about something whose very existence has not been established? Where's the empiricism involved in that process? Genuine questions, because all I see is a priori reasoning.

Have you ever heard of a scientific hypothesis?


Empirical thinking is the motivation for scientific hypotheses.

The question “What would hysteresis-induced distortion sound like?” is no less legitimate than asking “What would a gravitational wave look like in a detector?” …before gravitational waves had been directly observed.

Our describing a proposed observation is part of formulating a testable hypothesis.

Of course a description is not proof, nor is a hypothesis. But…a hypothesis is often proposed precisely because the existence or nature of a phenomenon has not yet been established! If we only theorized about phenomena whose existence was already proven there would be little need for hypotheses in the first place.

As I pointed out: hypothesizing and theorizing about AI, including via analogy to human intelligence, is what led to the creation of AI.

Your idea that it’s illegitimate to theorize about things that have not yet been established (including characterizing their properties) would rule out a significant chunk of science.

(ETA change word)

 

[Buckchester]

Something I recall reading in Toole’s book was that if your speakers have good on and off axis response, people prefer the sound when not treating first reflection points. If your speakers have problems with their off axis response, treating the first reflection points could possibly improve the overall sound by absorbing some of those issues.

Also, if you are going to treat the first reflection points you have to do it properly to avoid messing up a speaker with good off axis response. For example, I believe any absorbers need to at least 4 inches thick.

 

[JPA]

Something I recall reading in Toole’s book was that if your speakers have good on and off axis response, people prefer the sound when not treating first reflection points. If you’re speakers have problems with their off axis response, treating the first reflection points could possibly improve the overall sound by absorbing some of those issues.

Also, if you are going to treat the first reflection points you have to do it properly to avoid messing up a speaker with good off axis response. For example, I believe any absorbers need to at least 4 inches thick.

Yes, and preferably, spaced away from the wall a few inches.

 

[MattHooper]

My understanding is that a "really good seat" in a concert hall is within the LLD distance from the stage AND has sufficient temporal separation between the direct and reverberant sound to enable the perception of envelopment. These tend to be the seats up close and near the center.

My favourite seats at the symphony! I know lots of people like to sit further away from the orchestra for a greater sense of blend and cohesion (as well as with the surrounding acoustics). But my favourite closer seats because I like the vividness of individual instrument timbers, along with the more cinematic sweep of sound, and I enjoy being able to easily concentrate on a separate sections.

Maybe this is just me, but being able to close my eyes and effortlessly visualize where the sound images are located functions sort of like a surrogate for being able to see the performers. I have no idea whether imaging plays a similar role for anyone else.

Yes, I think that’s about right. For me anyway.

I’m one of those who insists that imaging happens in real music (that is an amplified acoustic music anyway). No doubt because I enjoy closer listening distances. So when I close my eyes at a live Orchestra. I still perceive some level of imaging precision.

But of course, most of the time I’m watching live music with my eyes open. And that says I think you’re right about imaging being a surrogate. If you’re watching a band play on stage with all sorts of different instruments happening, being able to visually concentrate on a single performer also helps you pick their contribution out in the sound as well. We don’t get that additional visual information when listening to our stereos, so the performer separated by imaging can help.

Right now I’m thinking about having listened a while back to one of Rush’s live concert albums. At the concert the sound would’ve come out as perceptually a giant wave of mono for the most part. No real imaging.

Put on the album, Alex Lifeson is mixed somewhat to the left of the stage, Neil Peart’s drums in the middle, and Geddy Lee’s bass just to the right of centre. Basically mimicking their typical orientation on stage.
That imaging helps me visualize the concert more believably in a way that it wouldn’t from a mono recording.

One of the big lessons from my own work in sound Post Production is that artificial embellishments often aid “ believability.”

 

[MattHooper]

It *is* bad science because it relies on lay intuition that reflections are echoes and hence, they must be eliminated. This invariably leads to rooms that are super dead.... There is no defensible position to say that sidewall reflections should be absorbed by default.

Agreed! THAT approach (absorption) to minimizing early same-side-wall reflections IS (imo) bad science.

And THAT has never been what I have advocated (nor Earl). From a post earlier today:

With regard to doing some absorption of side walll reflections, though my room has some helpful acoustic treatment, I still have to deal with early sidewall reflections.

It might not be optimal to the degree you are seeking, but in the spirit of not letting the perfect become the enemy of the good…. I have found that modulating sidewall reflections with simple, thick curtains to work very well for my purposes.

On the right side of my room, I have a wall that has thick velvet curtains that I can used to cut down certain reflection points mainly for the upper frequencies.

There’s also the thick velvet curtain you can see behind the speaker to the other side of the room entrance. I can pull that across the room opening to modulate the reflections on that side. I haven’t figured out whether they are reflections coming from sound going out into the hall and back.
But in any case, it helps modulate the sound as well.

The left side of my room has my left speaker right near reflective, vertical tiles from the fireplace.

If I don’t treat that area The left signal takes on something of a harsher timber in the upper frequencies. All it takes is covering that fireplace with a fitted cover - the same curtain material as the other side of the room - to completely “ fix” that problem.

I haven’t measured it, but in subjective terms I can’t notice any timber difference between the left and right speaker once I have adjusted reflections on both sides using some absorption.

I also feel that just the use of curtains allows me to modulate the top end timbre of the sound while also benefitting from enough reflection that adds some more spaciousness (as I do not toe in my speakers heavily).

The compromise as usual being between image focus and that spaciousness.

 

[Duke]

Why? Being an audiophile is like being fooled by the presentation being more than it is. If side reflections make your brain enjoy the experience more, then doing otherwise is the anthesis of being an audiophile.

I am all in favor of enhancing the experience! Imo the goal is to approximate the experience of hearing a live performance, moreso than to replicate "what the recording engineer heard".

But it sounds like you are saying that choosing a package of attributes which does not include apparent-source-width enhancing sidewall reflections would be "the antithesis" of what we should be doing. Am I misunderstanding you?

Or, are you including the assumption that "being an audiophile" is a bad thing? That word is used as a pejorative around here so pardon my confusion.

Do you recall the photo of @Floyd Toole's room from his home in California, the one where his Revel Salon 2's are standing on their heads atop a piece of furniture? On the left-hand side he has thick curtains where the first same-side-wall reflection would have happened, and on the right side he has an open doorway where the first same-side-wall reflection would have happened. So the ASW-broadening sidewall reflections were greatly diminished or eliminated. I hope that's not the antithesis of something desirable.

Note also that Floyd Toole used "tasteful upmixing" for stereo listening, and the contribution of the surround speakers was not part of "what the recording engineer heard", so I speculate that his goal was also something along the lines of enhancing the playback experience rather than replicating the recording engineer's experience.

What is the frequency response of the phantom center? You know, where the signer would be or your single violin.

If "audibility" matters more than what the frequency response measurement looks like, you might find this interesting -

Francis Rumsey cites a study which shows a correlation between the quality of a phantom image and the perceived tonal quality. In particular, the perception of coloration from stereo's 2 kHz comb filter dip goes away when you have a good enough phantom center image! And in general as the ability to detect a spatial image improves, the ability to detect a timbral coloration decreases. He calls this phenomenon "decoloration".

Here is his lecture, cued up to that part:

* * * *

In this kind of discussion, advocating for a particular approach can come across as saying or implying that everything else is wrong. That's not what I intend; I'm trying to say that a particular approach has legitimacy, rather than that it's the one and only correct way.

 

[amirm]

But it sounds like you are saying that choosing a package of attributes which does not include apparent-source-width enhancing sidewall reflections would be "the antithesis" of what we should be doing. Am I misunderstanding you?

A bit. I am saying if sidewall reflections make a listener believe more in what they are hearing, you don't want to eliminate them for the sake of it. The reflections are making up for the shortfalls of the system+recording to the mind.

Or, are you including the assumption that "being an audiophile" is a bad thing?

Not at all. I meant the opposite. The pursuit of what gear and setup to enjoy music to the max is what being an audiophile is about. To then change a setup as to make it "pure" and with it, reduce that enjoyment would be the antithesis of being an audiophile.

 

[amirm]

Do you recall the photo of @Floyd Toole's room from his home in California, the one where his Revel Salon 2's are standing on their heads atop a piece of furniture? On the left-hand side he has thick curtains where the first same-side-wall reflection would have happened, and on the right side he has an open doorway where the first same-side-wall reflection would have happened. So the ASW-broadening sidewall reflections are greatly diminished or eliminated. I hope that's not the antithesis of something desirable.

My setup is the same. I explained earlier that this is a subtle effect. If your home situation doesn't allow it, then that is that. What you don't want to do is go slap absorbers everywhere there is a reflection as some people do. Get superbly designed speakers, apply EQ for bass, and enjoy the music. Don't listen to people who go by lay intuition or folklore rooted in Pro space that reflections are bad and you have to get rid of all of them. The whole purpose of a speaker with good off axis response is to reduce such anxiety.

 

[amirm]

I hope I haven't misunderstood, but it's my impression that Amir himself has touted THD (or SINAD) as a "surrogate marker".

I use it as a predictor for the rest of the measurements in electronics at the two ends of the spectrum. A very high SINAD shows extreme care in design and will almost always result in the rest of the measurements being superb as well. Conversely, a poor SINAD almost guarantees the rest of the measurements will be poor as well.

With speakers, distortion is orthogonal to frequency response correctness so it doesn't play that role. With respect to THD alone, sometimes it plays a direct role in predicting audibility issues. In bass, when it reaches 50% or more, you are going to hear severe distortion. The challenge becomes quantifying things when it is 0.5%.

 

[Duke]

A bit. I am saying if sidewall reflections make a listener believe more in what they are hearing, you don't want to eliminate them for the sake of it. The reflections are making up for the shortfalls of the system+recording to the mind.

That makes total sense and I absolutely agree.

Not at all. I meant the opposite. The pursuit of what gear and setup to enjoy music to the max is what being an audiophile is about. To then change a setup as to make it "pure" and with it, reduce that enjoyment would be the antithesis of being an audiophile.

Again that makes total sense and I absolutely agree.

I explained earlier that this is a subtle effect. If your home situation doesn't allow it, then that is that. What you don't want to do is go slap absorbers everywhere there is a reflection as some people do. Get superbly designed speakers, apply EQ for bass, and enjoy the music. Don't listen to people who go by lay intuition or folklore rooted in Pro space that reflects are bad and you have to get rid of all of them. The whole purpose of a speaker with good off axis response is to reduce such anxiety.

Well now my comments are becoming redundant; once again that makes total sense and I absolutely agree. It looks like I may have been seeing through a lens that emphasized small areas of difference while overlooking broad areas of common ground.

I think we're both advocating speakers + setup that works "with the room" rather than "against the room" to make up for some of the perceptual shortfalls inherent to two-channel playback.

 

[olieb]

I take a somewhat different approach to "making up for the shortfalls of the system+recording", but I think we're both advocating speakers + setup that works "with the room" rather than "against the room" to make up for some of the shortfalls inherent to two-channel playback.

Given the very big impact that the room has on SQ, however, it is also possible to work “on the room”. It should be no dogma to use the room “as is“ and that includes to implement reflection, diffusion and absorption to some (reasonable) degree. It can make quite a difference (Toole has a full chapter on this).

My approach is very different anyways. I came to prefer headphone listening via binauralisation in an emulated room as my premier system. Room characteristics that are normally very hard to change (RT60, size, bass absorption) become quite “fluid“ in this approach. I arrived at a preference for rather dry acoustic in a large virtual room most of the time with the music I listen to. For material with recorded ambience on the low side I prefer to tune up the reflections somewhat.

 

[MattHooper]

I am all in favor of enhancing the experience! Imo the goal is to approximate the experience of hearing a live performance, moreso than to replicate "what the recording engineer heard".

But it sounds like you are saying that choosing a package of attributes which does not include apparent-source-width enhancing sidewall reflections would be "the antithesis" of what we should be doing. Am I misunderstanding you?

Or, are you including the assumption that "being an audiophile" is a bad thing? That word is used as a pejorative around here so pardon my confusion.

Do you recall the photo of @Floyd Toole's room from his home in California, the one where his Revel Salon 2's are standing on their heads atop a piece of furniture? On the left-hand side he has thick curtains where the first same-side-wall reflection would have happened, and on the right side he has an open doorway where the first same-side-wall reflection would have happened. So the ASW-broadening sidewall reflections were greatly diminished or eliminated. I hope that's not the antithesis of something desirable.

Note also that Floyd Toole used "tasteful upmixing" for stereo listening, and the contribution of the surround speakers was not part of "what the recording engineer heard", so I speculate that his goal was also something along the lines of enhancing the playback experience rather than replicating the recording engineer's experience.



If "audibility" matters more than what the frequency response measurement looks like, you might find this interesting.

Francis Rumsey cites a study which shows a correlation between the quality of a phantom image and the perceived tonal quality. In particular, the perception of coloration from stereo's 2 kHz comb filter dip goes away when you have a good enough phantom center image! And in general as the ability to detect a spatial image improves, the ability to detect a timbral coloration decreases. He calls this phenomenon "decoloration".

Here is his lecture, cued up to that part:

* * * *

In this kind of discussion, advocating for a particular approach can come across as saying or implying that everything else is wrong. That's not what I intend; I'm trying to say that a particular approach has legitimacy, rather than that it's the one and only correct way.

Great post!

First of all, I nominate Duke’s posts as a wonderful example of civil, good faith interactions.

Secondly, that’s a really interesting lecture. I am only partway through but I also found his comments about using a surround sound system to study which aspects were most important to experienced and naïve listeners.
It turned out that good timbral quality was significantly more important than any other factors such as spaciousness or front stage imaging (and also interesting that naïve listeners apparently did not rate imaging much at all, versus experience listeners who cared about it).

Yet one thing I always noticed in the more Home theatre oriented forums is that there is far less talk about timbre - Timbral correctness, timbral beauty, timbral accuracy - than in two channel audiophile. Talk was much more oriented towards scale of sound, dynamics, frequency extension.

And a majority of surround systems I would experience at stores and at home theater set ups struck me as timbrally bland, even if immersive. The sound “ surrounded” but it did not reward fine attention to timbral qualities of the sound.

I’m as big a home theatre nut as I am an audiophile, but my audiophile tendencies carried over strongly to selecting my surrounding system. That’s why I selected Hales Transcendence speakers as the L/C/R workhorse of the surrounding system, simply because having lived previously with examples of that speaker line, I found them among the most pure sounding and timberally colourful and beautiful speakers I had owned.

It worked out really well because not only do I thoroughly enjoy music through that system, but I find the timbral beauty and realism for movie soundtracks and sound effects to be really rewarding. Even just the sound of water trickling, or wind or rain has this slightly hair raising recognition of authenticity in my brain.

I think the study cited in that lecture was probably playing strictly music though.
If so, I wonder if listeners become even less discerning about timbre when it comes to movie soundtracks. (aside from perhaps timber matching mostly in terms of coherence between the speakers).

 

[Geert]

Francis Rumsey cites a study which shows a correlation between the quality of a phantom image and the perceived tonal quality

The study ’ON THE LOCALISATION IN THE SUPERIMPOSED SOUNDFIELD by Dipl.-Ing. Günther Theile, Berlin 1980’: https://hauptmikrofon.de/theile/1980-2_Diss._Theile_englisch.pdf

 

[Duke]

It turned out that good timbral quality was significantly more important than any other factors such as spaciousness or front stage imaging (and also interesting that naïve listeners apparently did not rate imaging much at all, versus experience listeners who cared about it).

Yes! That part starts around 25:00 in the video.

Arguably, to the extent that timbral quality is enhanced by having a high quality phantom source, front stage imaging does play a role therein, albeit a somewhat under-the-radar one.

... a majority of surround systems I would experience at stores and at home theater set ups struck me as timbrally bland, even if immersive. The sound “ surrounded” but it did not reward fine attention to timbral qualities of the sound. [emphasis Duke's]

Yes, he goes into that somewhat. In investigating wavefield synthesis, wherein many channels and many speakers in an array seek to replicate the wavefield, it was found that the more channels used, the more coloration resulted! This part starts around 51:10; this is not the exact same situation as multichannel audio, but it's an interesting finding anyway, imo, and suggests that a multiplicity of channels may not be the ideal solution after all, as least not in the configurations tested.

My take-away is that, in practice, there may be some fine-tuning involved to get a really good balance of desirable attributes without having distracting colorations.

 

[MattHooper]

Yes! That part starts around 25:00 in the video.



Yes, he goes into that somewhat. In investigating wavefield synthesis, wherein many channels and many speakers in an array seek to replicate the wavefield, it was found that the more channels used, the more coloration resulted! This part starts around 51:10; this is not the exact same situation as multichannel audio, but it's an interesting finding anyway, imo, and suggests that a multiplicity of channels may not be the ideal solution after all, as least not in the configurations tested.

My take-away is that, in practice, there may be some fine-tuning involved to get a really good balance of desirable attributes without having distracting colorations.

I ended up choosing my side and back surround speakers simply by ear. I searched for speakers that in my perception sounded timbrally like the Hales. That worked out to be speakers from the Monitor Audio Bronze line. Not the more expensive platinum series that I could’ve bought instead - they struck me as sounding a bit too tipped up and bleached sounding. But the cheaper Bronze line had a slightly darker warmer tone, and those ended up matching with the Hales beautifully. And then I’m careful not to play the surrounds too loud for I think some of the reasons you are getting it.

 

[Duke]

I ended up choosing my side and back surround speakers simply by ear. I searched for speakers that in my perception sounded timbrally like the Hales. That worked out to be speakers from the Monitor Audio Bronze line. Not the more expensive platinum series that I could’ve bought instead - they struck me as sounding a bit too tipped up and bleached sounding. But the cheaper Bronze line had a slightly darker warmer tone, and those ended up matching with the Hales beautifully. And then I’m careful not to play the surrounds too loud for I think some of the reasons you are getting it.

Yes, I like to add some late-onset reflection energy and ime both of the things you mention are important: Getting a good timbral match with the first-arrival sound; and setting the volume level correctly (by ear), as there seems to be a threshold above which clarity is degraded.