I cannot trust the Harman speaker preference score

[Andrej]

Agreed!

Imo one critical piece of information which is not conveyed by the decay time is, what's happening "early" as opposed to what's happening "late" in the decay process.

David Griesinger finds that the earliest reflections are the most detrimental:

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

"When presence is lacking the earliest reflections are the most responsible."

"Transients are not corrupted by reflections if the room [i.e. early-reflection-free time interval] is large enough - and 10ms of reflections free time is enough."

But what about later reflections - shouldn't they also be minimized? Not according to Griesinger (assuming these later reflections are spectrally correct):

"Envelopment is the Holy Grail... When reproducing sound in small spaces [home listening rooms], envelopment is often absent. 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."

"Where the background stream is easily separated from the foreground stream, envelopment is about 6 dB stronger for a given direct-to-reverberant ratio."

So imo it makes sense to pursue loudspeaker radiation patterns, loudspeaker and listener positioning in the room, and acoustic treatment strategies which minimize the early reflections but WITHOUT overly attenuating (or spectrally degrading) the energy which will arrive as beneficial later reflections.

I agree only in so far as it creates the desired "sound" which is a perfectly valid reason. It will certainly help many recordings, and of course it is the optimal way to design recording/listening venues where music is generated. However, for ultimate fidelity, whether you like the outcome or not, any reflection is noise, by definition.

I suspect that very few of us have listened to a simple recording of sounds not using amplification in a room approaching anechoic, when you are transported to the recording venue with it's acoustics. I love it, and will sacrifice other things to keep having such experiences.

 

[Duke]

I'm not disputing that spatial quality doesn't matter, and that it affects naturalness, pleasantness, and other attributes. But a lot of this is moot once you move beyond stereo to immersive audio where the spatial impressions are highly influenced by the number and locations of sources and how the sounds are mixed.

Stereo is 1950s, and the research focus now is immersive audio.

I agree that multichannel audio is the superior concept. Here are some issues I have encountered with the implementation of multichannel audio:

1. The music I most want to hear is only available in stereo.

2. My limited experience with upmixing has been that the level of the surround channels has to be correctly set by hand for each recording in order to avoid a potentially distracting (illusion-collapsing) result.

3. Purpose-built center channel speakers are never a timbral match with good left and right speakers.

4. With upmixing or music videos or multichannel music recordings, I have yet to hear a center channel speaker that conveyed the same sense of depth to the center of the soundstage that a good stereo setup could.

5. It seems to me that "bang for the buck" generally favors two-channel, but this may depend on the price range, and surely depends on personal priorities.

And I readily concede that I and my objections may well be in the minority.

 

[Andrej]

I am not quite sure how is what you just said a reply to my comment about 1st reflections and a wall of infinite mirrors being a poor analogy?
I have never been in a room where the sound even remotely paralleled the distinct impression of seeing multiple reflections like those mirrors.

Yes, you acclimate to a new room, you acclimate to new speakers, various different production qualities all sorts of acclimation. That is in a way part my original point. You can acclimate to the point you might enjoy sound in an anechoic chamber or from a bluetooth speaker in the bathroom. Who knows what the limits are.

But in any case reflections that are following the 1st arriving sounds closely in time will be perceived as a simultaneous event, not multiple images or even close. I feel this is well documented and frankly makes logical sense to me.

I apologize for the miscommunication. I was talking about my fondness for sound reproduction in an anechoic chamber, and that to fully appreciate it one has to take the time to adjust to such an unnatural acoustic environment. I might have responded to the wrong message.
You were asking somebody to go listen there and to report back. Since I had the experience, I was reporting my impressions.

 

[Duke]

any reflection is noise, by definition.

I disagree with this.

A reflection is a reflection, by definition. It CAN be effectively "noise" depending on the specifics, but reflections are far from necessarily interpreted by the ear/brain system as "noise".

 

[Andrej]

I disagree with this.

A reflection is a reflection, by definition. It CAN be effectively "noise" depending on the specifics, but reflections are far from necessarily interpreted by the ear/brain system as "noise".

In the signal sense - what comes in and what comes out, it is. Whether it helps perceptually, that is a different story. It is similar (as somebody already pointed out) to having some distortions generated by an amplifier which might be perceptually beneficial, but it is still distortion (or a synonym for noise). I also happen not to like it, but that is just a personal preference which might well be influenced by the fact that I like the intellectual purity of it

In support of that choice all I can say that I have always had several listening rooms with different systems and with varying levels of treatment. And all the systems were different (size, cost, complexity) although most of the time all could reproduce 20Hz-16kHz at reasonable loudness.
They were regularly rotated in different rooms, and the best system (as in how much I liked it) was the one in the room with most aggressive treatments. Every other metric paled in comparison. So for me, whether it is factually accurate or not, reflections are indeed the noise.

Here is what I ended up with, (at least for a while, and I miss it every day), while it was being constructed, some of the treatments had not yet been installed as you can see on the right.

 

[Andrej]

In the signal sense - what comes in and what comes out, it is. Whether it helps perceptually, that is a different story. It is similar (as somebody already pointed out) to having some distortions generated by an amplifier which might be perceptually beneficial, but it is still distortion (or a synonym for noise). I also happen not to like it, but that is just a personal preference which might well be influenced by the fact that I like the intellectual purity of it

In support of that choice all I can say that I have always had several listening rooms with different systems and with varying levels of treatment. And all the systems were different (size, cost, complexity) although most of the time all could reproduce 20Hz-16kHz at reasonable loudness.
They were regularly rotated in different rooms, and the best system (as in how much I liked it) was the one in the room with most aggressive treatments. Every other metric paled in comparison. So for me, whether it is factually accurate or not, reflections are indeed the noise.

Here is what I ended up with, (at least for a while, and I miss it every day), while it was being constructed, some of the treatments had not yet been installed as you can see on the right.

My final message on this topic, it just takes too much time to keep track of all the messages and responding to some (too many?-) of them.

Getting back to the topic of this exchange - perceptual evaluations of speakers. My concern is that the way it was done in the past combined 3 factors which can be significant biases in the outcome, making them worthless for me personally:

1) Frequency response (at least on-axis)
2) Room response
3) Single volume (due to Fletcher-Munson it favors some, and not others)

This is close to optimal for a typical home environment, for a determined audiophile, it probably is not.

My suggestion (as impractical as it might be) is to try something like this:

1) Dirac
2) As close to anechoic as possible, or leave as a typical room to measure the effects of off-axis contributions, or both
3) When doing 1) use an appropriate target curve for that volume.
And possibly:
4) Add high pass filtering to remove the effects of bandwidth (say at 80Hz) and speaker size.
Also adding a sub would solve bandwith deficiencies (post filtering) and one sub could be used with all the speakers
(since they are all calibrated to identical loudness)

All of this is relatively easy and manageable, and would make it more useful to many more. Especially given that something like Dirac or a digital parametric equalizer with guidance from REW is accessible to all, although not necessarily mastered by many.

Feel free to argue it to shreds - the beauty of this hobby is the journey at least as much as the destination. Cheers!

 

[Duke]

In the signal sense - what comes in and what comes out, it is. Whether it helps perceptually, that is a different story. It is similar (as somebody already pointed out) to having some distortions generated by an amplifier which might be perceptually beneficial, but it is still distortion (or a synonym for noise). I also happen not to like it, but that is just a personal preference which might well be influenced by the fact that I like the intellectual purity of it

Imo defining reflections as "distortion" for the sake of "intellectual purity" is not useful because it implies the incorrect conclusion that anechoic conditions are the ideal.

Floyd Toole on the subject:

"In-head localization seems like the logical opposite of an enveloping, external, and spacious auditory illusion. Perceptions of sounds originating inside the head, which routinely occur in headphone listening, can also occur in loudspeaker listening when the direct sound is not supported by the right amount and kind of reflected sound. The author and his colleagues have experienced the phenomenon many times when listening to stereo recordings in an anechoic chamber, usually with acoustically “dry” sounds hard panned to center or, less often, to the sides... In an anechoic chamber, it can occur when listening to a single loudspeaker, especially on the frontal axis, in which case front-back reversals are also frequent occurrences. This phenomenon is so strong that it need not be a “blind” situation. Interestingly, a demonstration of four-loudspeaker Ambisonic recordings played in an anechoic chamber yielded an auditory impression that was almost totally within the head. This was a great disappointment to the gathered enthusiasts, all of whom anticipated an approximation of perfection. It suggested that, psychoacoustically, something fundamentally important was not being captured or communicated to the ears. An identical setup in a normally reflective room sounded far more realistic.”

I think one issue which manifested in different forms in the various speaker systems you described in post #599 is, poor off-axis response. It is not apparent to me that any of the speaker systems you described would have an off-axis response that closely tracks the on-axis response. So in your situation approaching anechoic conditions was evidently beneficial, but that's hardly a general truth in the context of home audio.

Assuming that the reflections are spectrally correct and that their timing and level and diffusiveness and decay are good, they will enhance both timbre and spatial quality. In the context of a "they are here" presentation, reflections done right play a similar role to what they would play in a venue, within the obvious limitations. In the context of a "you are there" presentation, reflections done right are carriers for the venue information on the recording.

Note that multichannel sound done well is the deliberate injection of reflective energy via the surround channels, so multichannel is based on the premise that reflections done right are beneficial.

 

[Holmz]

Imo defining reflections as "distortion" for the sake of "intellectual purity" is not useful because it implies the incorrect conclusion that anechoic conditions are the ideal.

Floyd Toole on the subject:
…

Yep. His definition of intellectual purity is a bit odd.

In the signal sense - what comes in and what comes out, it is. Whether it helps perceptually, that is a different story. It is similar (as somebody already pointed out) to having some distortions generated by an amplifier which might be perceptually beneficial, but it is still distortion (or a synonym for noise). I also happen not to like it, but that is just a personal preference which might well be influenced by the fact that I like the intellectual purity of it
…

Whether it is a bat, a dolphin, or a radar system… there is a significant difference between broadband white noise, and a reflected signal.
These echo are more bat like, in that they inform us of the boundaries.

To equate them with noise is intellectually dishonest.


Further one could largely remove the echos with a comb filter, or create the echo, with a comb filter.
Similarly HD in an amplifier is a different distortion, than say, crossover distortion.
So all distortions and all noises are not created equal. If they were we would not have different descriptions of them.

Can we get back to whether the Olive preference is worthwhile or lacking?
And if it is lacking,.. then what needs to be done to improve it?

In the mean time I will continue to put some value in it, and not discount decades of work in the science of audio, based upon a few unpublished opinions in an audio forum.

 

[witwald]

My hypothetical example was a speaker that scores very well by Harman criteria in mono, perhaps by virtue of cabinet sounds compensating for peaks and dips.

That is an interesting concept: making the cabinet behave so badly while at the same time being able to control its vibration characteristics, such that the overarching peaks and dips in the loudspeaker's overall frequency response can somehow be ameliorated. Is this even feasible, without going down the cost-no-object route? An approach such as this appears to be contrary to what we know about cabinet vibrations and how they contribute to a loudspeaker's sonic signature.

 

[witwald]

Frequency response abberration in particular ranges will produce psychoacoustic effects which may sound nice (e.g. presence dip).

Won't the presence dip that is referred to above sound nice for only some types of listeners? For example, a sound engineer monitoring close to the speakers may prefer the presence dip, but an at-home listener using a more distant listening position may hear it is a lack of spectral energy that distorts the sound reproduction. So what sounds "nice" in one situation will sound "bad" in another situation (at least to an experienced/trained listener).

And then there is the issue directivity, which seems to be generating bias/favour in mono testing, whilst being a less accurate transduction of the signal due to increased early-reflection room interaction/interference.

It would seem that a viable approach to solving this problem is to choose a loudspeaker on the simple basis that it is known to be highly accurate at the transduction of the input signal (e.g., very flat, well-extended frequency response, with low distortion and low compression characteristics), and with a well-controlled directivity pattern. It then behooves the listener to adjust their listening room to suit their particular expectations of the sound reproduction quality. This is the classical speaker–room interaction problem, except that the major problems with the loudspeaker have simply been engineered out of the equation, which is what present knowledge and technology seem to be largely capable of achieving.

Would not such an approach help to solve the problem of choosing a loudspeaker from the plethora of products in the marketplace?

 

[CinDyment]

Being pedantic, noise is uncorrelated to the signal, so reflections are not noise because they are correlated.

Discussing purity to the original presentation in 2 channel audio is like discussing which shade of white is correct. All are correct and none are correct.

The reflections could accidentally on purpose create a sound field closer to the original, or not, at least w.r.t. interpretation of the presentation. So much information is missing in 2 channel audio that what you artificially create could be closer to the original, you really don't know.

 

[witwald]

Multi-channel may be better than 2-channel but it is still flawed.

Looking at it dispassionately, without the embedded marketing streak, it would seem that you have hit the nail on the head. One has to clearly identify what the aim is, and multi-channel audio seems to be largely aimed at providing a different "experience", not a high-fidelity recreation of the sound field as would be heard by a listener in a specific location at the original acoustic event. Much of multi-channel audio seems to be targeted more at creating entertaining "special effects" than high fidelity.

 

[CinDyment]

Looking at it dispassionately, without the embedded marketing streak, it would seem that you have hit the nail on the head. One has to clearly identify what the aim is, and multi-channel audio seems to be largely aimed at providing a different "experience", not a high-fidelity recreation of the sound field as would be heard by a listener in a specific location at the original acoustic event. Much of multi-channel audio seems to be targeted more at creating entertaining "special effects" than high fidelity.

I disagree w.r.t. music. How many multi-channel recordings have you listened to? It is flawed, but I don't think the emphasis is "special effects".

 

[witwald]

Every mainstream pop album is being released with Dolby Atmos on the major streaming services.

Be that as it may, is the end result of that something that will be taken up by the broad range of music consumers? I doubt it. It's a technology looking for a market, albeit a quite small one, but through its marketing of it, the music industry hopes to sell more music. It would be interesting to know how many listens are via a Dolby Atmos sound system. It seems unlikely that it would be more than 0.5%, if that.

There is definitely a paradigm shift happening because no one listens to stereo anymore.

That seems to be quite an inaccurate assessment...

It's either binaural, single speakers or multiple speakers / sound bars .etc Dolby Atmos caters to all these formats with a single master.

And the bulk of the market will continue to listen in stereo, as they don't really care one iota about the other options, which are simply not on their radar.

 

[witwald]

“Abbey Road Studios has software that allows you to essentially reverse-engineer an old track of audio with multiple instruments on it and separate it into individual tracks.”

It's all quite unknown from that simple statement as to how well those multiple instruments are separated into their individual tracks. I wonder if any of those tracks of the so-called isolated instruments are available for an assessment? So, now we have companies playing with the works that the artists originally recorded to simply create more product that they can market to consumers, while bypassing the original art. I suppose the next thing will be to post-process Jackson Pollock's Blue Poles to make it into a 3D installation, making it so much "better", as a result of applying all those special effects that are now possible to apply.

 

[Holmz]

Be that as it may, is the end result of that something that will be taken up by the broad range of music consumers? I doubt it. It's a technology looking for a market, albeit a quite small one, but through its marketing of it, the music industry hopes to sell more music. It would be interesting to know how many listens are via a Dolby Atmos sound system. It seems unlikely that it would be more than 0.5%, if that.

…

Or a way to peddle music to all the HT people that have Atmos systems.
That would be a brilliant marketing strategy.
New content, new service, mo money.

 

[Newman]

I think that this is the ultimate sticking point for folks and one of the main crux's of audiophilia that Toole touches on in his book when he laments the the lack of basic tone controls on so much hifi gear.
He is alluding to the fact that in his worldview the content as it is, does not have to be idolized or stomached at the expense of pure enjoyment.

This is why I disagree with him passionately (on this subject). Music is art, it should be respected. There used to be a different word for melodic sounds that allowed you to time pass by: Muzak. EQ that as you feel like it.

I respect the artist. If they wanted less bass or more treble they would put it on the record. By using tone control or EQ track by track I feel like I’m disrespecting the artist.

Naturally this what I think. It’s my opinion.

It is what you [sarumbear] think and it is more than just opinion: it is wrong. You completely misunderstood what Toole is advocating. (You seem to have an anti-researcher bias that leads you too hastily to “passionate disagreement” with them. Not sure why you are so inclined, but you are.)

Toole is 100% in favour of respecting the art of the music in its recording and playback. With no exceptions. You would also know this if you had read his material with an unbiased eye. Strange.

He advocates the use of speakers with smooth, flat and extended FR….because music is art, and should be respected.

He advocates the use of speakers with near-constant or smoothly varying beamwidth and DI….because music is art, and should be respected.

He advocates the use of room treatment that preserves the character of the incident sound….because music is art, and should be respected.

He advocates the limited use of room equalisation to counter the worst effects of small-room playback, particularly in the bass….because music is art, and should be respected.

He advocates the use of multichannel recording and playback technology because it is much better able than 2-channel to capture and replicate …and respect… the artist and the music as art, and more fully experience that art, which artists only put to 2-channel because it was the only widely available alternative in its day, but would always have used multichannel if it was a realistic option, because they would have been easily able to perceive its superior ability to represent the magic and art of the music itself.

And…he advocates the use of tone controls to adjust for many recordings having too much or too little bass or treble, which does not respect the sonic qualities that the music originally had, and failing to have decent tone controls that can make such corrections when necessary is a distortion of the art of the music….because music is art, and should be respected.

That’s where you come in and get it all back to front.

 

[Sean Olive]

Sorry, English is not my first language and I am not particularly eloquent.

What I am trying to say is that although mono testing is adequate for assessing speaker performance (indentifying issues in the transduction of the signal - distortions), it is unable to create the streophonic illusion.
And because most people will buy speakers to listen to music in stereo, for preference assessment stereo listening is in my view mandatory.
Also, mono testing doesn't take into account the positioning of the loudspeaker in relation to the room and its boundaries because it is placed in the centre of the testing room where early reflections from side-walls are minimal, far from what one will experience in a domestic setup.
Then there is the issue of room-interference below Schroeder: different speakers will require a different position to achieve the best balance in the bass and sub-bass, and mono testing speakers in the same position will not take that into account. Likewise, there is an optimal listener position so using a large audience of seats will not be optimal. And some speakers have different requirements regarding toe-in or distance to boundaries for correct performance.

I’m not claiming that mono tests recreate stereo imagery in stereo tests. I’m arguing that a speaker that tests well in mono will test well in stereo. Floyd’s paper shows that spatial ratings in the stereo tests are quite variable and highly dependent on the recording.

If you believe testing a mono speaker in the center of the room is invalid then set it up as a stereo left or right as we can do with our speaker mover. Then the reflections are identical to a left or right speaker in a stereo setup.

As far as how the speaker couples into the room modes the optimal position of the loudspeaker will depend on the acoustical properties and dimensions of the room and the locations of the listener(s). To say that the optimal position is different for every speaker is nonsense, especially for conventional speakers where below the room transition frequency the speaker is close to a monopole and will couple into the room modes much the same way.

As an aside if there were different optimal locations for different speakers I’ve never seen this specified in the loudspeaker setup manual. And wouldn’t they be different for different rooms? And there would no need for room correction and calibration.

A much larger effect/bias is comparing different speakers in different locations of the room. Not only will the reflection patterns be different but there will be differences in bass due to how they couple into the room modes. These positional biases have been well documented in our AES papers and are the reason we invested significant money into an automated speaker mover that places each speaker in the exact position.

Only a single listener in the optimal seat is used in our product benchmarking tests. For group or tour demonstrations we will use multiple seats. We analyze the data separately for each seat and can see how the ratings vary across seat. The consistency in ratings across seats is quite remarkable except when a speaker has a significantly different frequency response on and off axis producing a much different direct sound for listeners in different seats

 

[Newman]

This would be Reflection-Free Zone room layout:

 

[Voo]

its nice to own well measured/rated/respected gear but sometimes it just doesnt work/sound as good to one as another. enjoy the journey and try/buy as much as possible before you give your money away.