How NOT to set up speakers and room treatment ( Goldensound)

[amirm]

I don't know where this claim of additional room reverb for classical recordings comes from, but it seems not to come from people who have heard classical surround recordings in professionally sounding (= quite dry) rooms.

You don't know because you have looked or you don't know because you haven't? Likely the later so let me show you the research on that: Loudspeakers and Rooms for Sound Reproduction—A Scientific Review*

3.4 Effects of Single Reflections on Listener Preferences of Sound Quality
It is accepted that a reflective sound field is flattering to the sound of music. We like to listen in a reverberant space, rather than outdoors. The question is at what point does this positive attribute begin? Ando has provided some answers. Fig. 10 shows levels for a single delayed sound that listeners reported as enhancing the sound of classical music. Since the early reflections in real rooms are so low in level, the result suggests that we really need multichannel audio to provide added stronger and later reflections for our listening pleasure [49].

Pretty clear that listeners crave side reflections that are strong and delayed as much as 40 milliseconds. This is far longer path length than in our rooms (solid dots based on Alan Devantier's research) lending hand to the comment that multi-channel is able to convey that so much better than stereo. And that we do not like dry rooms as you claim unless it is multichannel.

 

[TheZebraKilledDarwin]

Suddenly? Didn't you not read me repeatedly saying how much personal research I performed before accepting Dr. Toole's compelling view? And what is the alternative? Listening to Joe random online? Or reading and understanding this paper:

LOUDSPEAKERS AND ROOMS FOR STEREOPHONIC SOUND REPRODUCTION
FLOYD E. TOOLE National Research Council of Canada, Ottawa, Ont., KIA OR6,Canada

For which Dr. Toole received a Silver Medal from Audio Engineering Society? This paper was published back in 1990 or 23 years ago. In there, you see graphs like this:

View attachment 290184

Making a safe bet that you don't know how to read this graph, let me interpret it for you. It says that in the common domestic rooms where delays are below 10 milliseconds for a lateral reflections (10 foot longer path length), to hear what you think is bad, i.e. distinct reflections, the reflection needs to actually be louder than the direct signal!

Amir,
with all respect, but you are making assumptions, that are not correct. That's in general the problem of academization, if theory is everything and practice is being ignored.

Do you know, that the Haas-effect is used in audio production to spread out mono sources? For example widen (background) vocals? Widen guitars?
It is even done on hard L or R panned instruments: give the other channel a little Haas effect.
All these cases have one thing in common: you don't hear a reflection, but you hear a dramatic change in the perceived stereo image of the source.

And that effect "works" even at relative -15 dB levels to the original.

That's the reality and it's not in that graph.

If you take only the information in that graph, you are applying an information filter and that filters out huge parts of reality and you come to wrong conclusions.

So lets see: we listen to you or a man who has dedicated his 50+ year career to this topic. Thinking...thinking... I think I will go with Dr. Toole and all the other researchers instead of what you have to say.

The problem of humanity these days is academization: the arrogance of academics, who lack practical experience and believe, their models were a full representation of the world.
Until a few years later someone tells them, that the old model did not take something important into account.

I know it''s futile, but it would be good if academics would show some humility and non-academics would trust academics less, even if they wear white coats...

 

[thewas]

Toole argues that side reflections actually allow you to hear the original signal more accurately, as evidenced by speech intelligibility experiments (Toole hypothesizes that this is because reflections allow your auditory system to get a "second look" at the signal).

The important question though is is such really the reason, or just the increased SPL integral over some time window?

 

[amirm]

Amir, with all respect, but you are making assumptions, that are not correct. That's in general the problem of academization, if theory is everything and practice is being ignored.

Do you know, that the Haas-effect is used in audio production to spread out mono sources? For example widen (background) vocals? Widen guitars?

What assumption? I quoted a graph that summarized research by a number of luminaries in the field. One of those is Helmut Haas who has contributed to the very topic Dr. Toole was referencing. The paper's title is: The Influence of a Single Echo [reflection] on the Audibility of Speech. This is specific research into impact of speech intelligibility in the presence of single reflection -- precisely what Dr. Toole had quoted and summarized in that graph. Here is the link to the paper's translation published in Journal of AES: https://www.aes.org/e-lib/online/browse.cfm?elib=2093.

Seems like the only thing you know about Haas' work is precedence effect. I suggest reading the paper and not jumping the gun this way.

 

[caught gesture]

The problem of humanity these days is academization: the arrogance of academics, who lack practical experience and believe, their models were a full representation of the world.
Until a few years later someone tells them, that the old model did not take something important into account.

I know it''s futile, but it would be good if academics would show some humility and non-academics would trust academics less, even if they wear white coats...

I would have to take some exception to these statements. The academics I know are passionate, not arrogant. They are passionate about exploring their chosen fields of study and, the ones I know at least, never state that their models are a full representation. They are constantly looking for more data that challenges their understanding. In fact, the desire to find something new is what drives them.

 

[amirm]

The problem of humanity these days is academization: the arrogance of academics, who lack practical experience and believe, their models were a full representation of the world.
Until a few years later someone tells them, that the old model did not take something important into account.

Academics? Dr. Toole's career has been about how to get better sound out of speakers and rooms. In which way is this academic? His research has been standardized in ANSI and CEA/CTA. It is the bible for designing great speakers. And better understanding of room and acoustics.

And what are we supposed to replace it with? Our esteemed members who haven't bothered to read a couple of papers and understand the research at high level let alone detail?

If you want to rant about something, rant about the instant gratification we seem to demand these days. Folks can't be bothered to go properly back to school and learn the science. If it it is not in a youtube video or five liner quote, it is too much for them to go and read and understand. Well, I have bad news for you: this field requires extensive reading and understanding. This is not a field for the lazy.

 

[Thomas_A]

Suddenly? Didn't you not read me repeatedly saying how much personal research I performed before accepting Dr. Toole's compelling view? And what is the alternative? Listening to Joe random online? Or reading and understanding this paper:

LOUDSPEAKERS AND ROOMS FOR STEREOPHONIC SOUND REPRODUCTION
FLOYD E. TOOLE National Research Council of Canada, Ottawa, Ont., KIA OR6,Canada

For which Dr. Toole received a Silver Medal from Audio Engineering Society? This paper was published back in 1990 or 23 years ago. In there, you see graphs like this:

View attachment 290184

Making a safe bet that you don't know how to read this graph, let me interpret it for you. It says that in the common domestic rooms where delays are below 10 milliseconds for a lateral reflections (10 foot longer path length), to hear what you think is bad, i.e. distinct reflections, the reflection needs to actually be louder than the direct signal! I have marked the 0 dB line indicating this (and the dashed/solid lines based on Haas and Meyer/Schodder research). This is obviously is not going to happen in any room unless it is the shape of a circle or something. So whatever you gut tells you here is wrong.

At the other extreme, the threshold of hearing if there is a reflection at all, is -15 dB or so. In other words, it doesn't take much for the side reflection to become inaudible.

In the middle we have image shift or spreading of the sound away from the speaker -- the very thing many listeners like.

Notice how the graph is composition of research by others plus that of Dr. Toole's team. And all of it is before Dr. Toole arrived at Harman.

So lets see: we listen to you or a man who has dedicated his 50+ year career to this topic. Thinking...thinking... I think I will go with Dr. Toole and all the other researchers instead of what you have to say.

I agree with this. I keep side and back wall without damping although my sofa has soft neck rests that do something. Otherwise I have plants in the window and paintings on the wall. Although the frequency content is not revealed in the typical ETC, I think the first ms is important, keeping the speaker area "clean", including perhaps also speaker diffraction low. Peaks between -10 and -20 dB don't seem to bother me.

 

[Wesayso]

Making a safe bet that you don't know how to read this graph, let me interpret it for you. It says that in the common domestic rooms where delays are below 10 milliseconds for a lateral reflections (10 foot longer path length), to hear what you think is bad, i.e. distinct reflections, the reflection needs to actually be louder than the direct signal! I have marked the 0 dB line indicating this (and the dashed/solid lines based on Haas and Meyer/Schodder research). This is obviously is not going to happen in any room unless it is the shape of a circle or something. So whatever you gut tells you here is wrong.

Yet, even when that reflection is sufficiently below the main level, it can and will alter the perceived position of the audible queues if it's arriving within that first 10 ms. Giving you that fake widening of the stage. For all material that's being played back in that room. Like I have said earlier, the same sauce on everything you play. Let's not even start on the effect it can have on tonality, even when you have the perfect speaker with a great directivity plot. Your room probably does not have the reflective surfaces that would do it justice. It could wack up the results of that great speaker especially if the reflections arrive within the first ~7 ms.

If one would avoid those earliest of reflections and still keep the later diffused reflections, much like in the LEDE concept, the room won't sound even close to dead and imaging will follow the actual material as it was recorded and mixed. Big and large if the recording has it, small and intimate when it is recorded and mixed like that. Maybe not of interest to those that usually listen to large scale orchestra's but quite a bit of fun for those that appreciate just about every other type of music. It still works quite well with large scale orchestra's too by the way.
All that is needed is a bit of care when picking the speaker for your specific room, and maybe some selective treatment as no EQ can solve that "same sauce on every recording" syndrome that those early reflections cause.

The result of some careful setup/treatment goes a long way for listening pleasure even in smaller rooms. Keeping an open mind when interpreting the works of Toole and others goes a long way to be able to improve your own specific room. Trusting a guy that spend 2 years reading papers is an option of course, but it isn't that hard to try some things yourself and see if you agree with his findings. Especially when you have that smaller sized room! Pick the speakers that suit the room, that should be the first step. About the only thing I do agree with is that padding an entire room would be a huge mistake. But there's plenty of gray area in between that can result in far better sound than "an untreated room according to Amirm's standards". I think it's save to say that I've been digesting papers and journals for far more than 2 years, yet I still try (and often fail) to keep an open mind about it. We don't all have the same room or same funds or same taste. Figure it out for yourselves. It really isn't that hard to try a few things and experiment. Personally, I went overboard with the experiments, but I have had a lot of fun doing it! There's no need to go that far, but you can do better if you keep your mind open, that's for sure!

 

[Axo1989]

Amir,
with all respect, but you are making assumptions, that are not correct. That's in general the problem of academization, if theory is everything and practice is being ignored.

Do you know, that the Haas-effect is used in audio production to spread out mono sources? For example widen (background) vocals? Widen guitars?
It is even done on hard L or R panned instruments: give the other channel a little Haas effect.
All these cases have one thing in common: you don't hear a reflection, but you hear a dramatic change in the perceived stereo image of the source.

And that effect "works" even at relative -15 dB levels to the original.

That's the reality and it's not in that graph.

If you take only the information in that graph, you are applying an information filter and that filters out huge parts of reality and you come to wrong conclusions.


The problem of humanity these days is academization: the arrogance of academics, who lack practical experience and believe, their models were a full representation of the world.
Until a few years later someone tells them, that the old model did not take something important into account.

I know it''s futile, but it would be good if academics would show some humility and non-academics would trust academics less, even if they wear white coats...

It's quite likely a problem that researchers and followers are often ignorant of music production techniques. Like the one you just mentioned. You end up shaking your head.

As I noted upthread, Toole's description of popular music appears over-generalised and dated (for example at s.4.4):

In popular music and jazz stereo recordings, images are often hard-panned to the left or right loudspeakers, again monophonic sources, and the phantom center image is the product of double mono, identical sounds emerging from L and R loudspeakers.

There's a thread here a while back with a screenshot of a Billie Eilish track up on the DAW. We can see around 18 vocal layers active at the marker iirc. How do people think the sonics are placed and sculpted? Double mono my arse.

 

[edechamps]

The important question though is is such really the reason, or just the increased SPL integral over some time window?

That's a good point. I went back to the relevant sections in Toole's book and I stand corrected - the main study he cites basically states that early reflections improve speech intelligibility by the same amount that the equivalent additional energy in the direct sound would. So in retrospect, I think I was being misleading when I said that early reflections improve speech intelligibility in a context where the listener has access to a volume control - it would be more correct to say that they don't hurt speech intelligibility.

 

[TheZebraKilledDarwin]

And what are we supposed to replace it with?

With an open mind maybe?

Let me repeat:
The impact on the perceived location of a soundsource due to the Haas effect even at levels - according to your trusted sources where they cannot be heard anyway - is huge.
At audible threshold levels in the pictures you posted, it is totally messing up the whole stereo image.

Thanks to computers and free software easily verifiable and reproducable for everybody. Since we are interested in the empirical scientific method, right?

 

[Wesayso]

That's a good point. I went back to the relevant sections in Toole's book and I stand corrected - the main study he cites basically states that early reflections improve speech intelligibility by the same amount that the equivalent additional energy in the direct sound would. So in retrospect, I think I was being misleading when I said that early reflections improve speech intelligibility in a context where the listener has access to a volume control - it would be more correct to say that they don't hurt speech intelligibility.

A clear example of an open mind!


If you know what it does, you can play with it. Guess what the music producers do .

 

[Matt Bell]

Amir,
with all respect, but you are making assumptions, that are not correct. That's in general the problem of academization, if theory is everything and practice is being ignored.

Do you know, that the Haas-effect is used in audio production to spread out mono sources? For example widen (background) vocals? Widen guitars?
It is even done on hard L or R panned instruments: give the other channel a little Haas effect.
All these cases have one thing in common: you don't hear a reflection, but you hear a dramatic change in the perceived stereo image of the source.

And that effect "works" even at relative -15 dB levels to the original.

That's the reality and it's not in that graph.

If you take only the information in that graph, you are applying an information filter and that filters out huge parts of reality and you come to wrong conclusions.


The problem of humanity these days is academization

I would have to take some exception to these statements. The academics I know are passionate, not arrogant. They are passionate about exploring their chosen fields of study and, the ones I know at least, never state that their models are a full representation. They are constantly looking for more data that challenges their understanding. In fact, the desire to find something new is what drives them.

Well exactly, and very nicely put. (I would have responded in an altogether harsher tone than you did, but I’m trying to keep to a vow of not getting angry online.)

Another point that needs to be borne in mind is that “academization” is necessary and desirable.

These subjects can take years to understand, and if you want to make a contribution to moving the subject forward, that’ll take more years again.

And in order to do that, you’ll most likely have to do it as your main job, working for a university or research institute or a company that has the resources to fund research.

There’s no way we could reach the state of knowledge we now have across the sciences without “academization”.

 

[TheZebraKilledDarwin]

Well exactly, and very nicely put. (I would have responded in an altogether harsher tone than you did, but I’m trying to keep to a vow of not getting angry online.)

Another point that needs to be borne in mind is that “academization” is necessary and desirable.

These subjects can take years to understand, and if you want to make a contribution to moving the subject forward, that’ll take more years again.

And in order to do that, you’ll most likely have to do it as your main job, working for a university or research institute or a company that has the resources to fund research.

There’s no way we could reach the state of knowledge we now have across the sciences without “academization”.

You get angry but you do not even understand what "academization" means although I explained it...

 

[theREALdotnet]

For the purposes of this work, a consensus of the available information suggested that a target of −20 dB and 20 ms would be appropriate for the assessment of stereophony. That is, at the listening position, no reflection greater than −20 dB relative to the direct sound would occur in the first 20 ms.

This is precisely what I’m looking for in my own ETC measurements. It has served me well by guiding me towards better stereo imaging, and it has saved me countless hours of listening comparisons, which are tedious and hard.

 

[theREALdotnet]

View attachment 290185

Reflections at a relative level of 0dB and more? Relative to the direct sound? How does that work?

 

[Berlin]

The problem of humanity these days is academization: the arrogance of academics, who lack practical experience and believe, their models were a full representation of the world.
Until a few years later someone tells them, that the old model did not take something important into account.

The problem with humanity these days is that academics were not believed decades ago when they were already predicting global warming quite accurately....

Unfortunately, this lack of trust in science is not easily remedied by adding or omitting absorbers....

 

[caught gesture]

You get angry but you do not even understand what "academization" means although I explained it...

In fairness, academization when taken back to the adjective academic can have a variety of meanings from “belonging to the the school or philosophy of Plato” to “so scholarly as to be unaware of the outside world”. In between are various other definitions. Your original statement paints academics as all being in the latter camp and all in need of showing humility. That is a sweeping generalisation. Indeed some academics do demonstrate arrogance but that trait is not specific to academia. There are arrogant garbage-collectors and arrogant CEOs, and arrogance can be found in individuals in any profession you care to list.

 

[Wesayso]

The problem with humanity these days is that academics were not believed decades ago when they were already predicting global warming quite accurately....

Unfortunately, this lack of trust in science is not easily remedied by adding or omitting absorbers....

There simply wasn't enough money to be made by listening to them, that's the bigger issue.

 

[Somafunk]

Ando has provided some answers. Fig. 10 shows levels for a single delayed sound that listeners reported as enhancing the sound of classical music.

And that we do not like dry rooms as you claim unless it is multichannel.

Forgive me if I have misinterpreted your statement but id say that the majority of listeners (myself as one) do not/have not ever listened to classical music, I fully understand the data that shows classical music recorded by multiple mic placements would benefit from reflected sounds reaching the listening position but this does not apply to my choice of electronic/techno music with a BPM of anywhere from 100BPM to 130BPM+ consisting of entirely synthesised sounds created in the box so to speak - god bless the Roland 303/606/909, Juno, Moog's etc whether they be software or physical equipment.

Before room treatment the reflective sound in my small 4.6m x 3.8m x 2.4m room contributed massively to what I can only describe as a muddy/undefined mess with no clarity or definition/separation in the listening position which felt very tiring to listen to. A kick should have a visceral snap that grabs you by the lapels and slaps you round the face before disappearing which is how it was programmed by the artist, then we have a slow breakdown leaving a single kick for 8 bars building up tension before additional layering of sounds reaching a well defined crescendo that does not benefit from room reflections at all.

It'd be a sad forum experience if this thread was locked as I feel it has been informative and beneficial to realise the conclusion that not everyone prefers the auditory experience of reflected sound.

Just one example, I could have chosen a more stripped down techno track but I feel this is more accessible for non electronic/techno listeners, try and listen to "Breaks/Breakbeat" from 20 odd years ago in a reflective room and its an auditory mess.