Lets combine "What do we hear?" with "how do we hear" and then put blurred lines to distinguish that "how do we hear" between no-room situation and rooms of different size and reflectionness and appropriate EQ presets. This area deserves more research than everything else. And its rather psychoacoustics then physics. So far it been thought of as "equalizable conditions". While it is possible to a degree, it's not possible to precisely mimic the same musical source to be equal in different rooms. So it's rather non-equalizable conditions. For example, industry can't even agree on HF attenuation depending on room size (2 decades long X-curve debate).
Then add "where do we listen". If that's what you meant by "other way around", then I would very much agree.
Then add "how many of us listen simultaneously" (version for DYI - how big is my sofa?). So we came to Pro design approach mentioned earlier in this thread. I'd rephrase it as "What is the audience size and seat placement geometry?" Which leads to answer to "what is desired directivity?" or rather "polar response".
Now when polar response is constant regardless of frequency (hard goal) and matched to audience size (more or less of it will do harm for obvious reasons) one may find that it sounds very different from another polar response also constant and matched to audience size but differrent (size) one. And circle starts again as there are no strict baselines on what's optimal in-room polar response against "how do humans hear" and given room size. EQing may help to lessen the difference. But EQing against what exactly? There is no good baseline known to me.
We know that straight EQing doesn't sound good in reflective rooms. Still it sounds great outdoors and to some degree in treated rooms. Is this knowledge sufficient for building great stuff or being good as sound engineer? It's not, at least from my experience being part of the audience in many concerts in different conditions. HF/MF in-room envelopment is strongly bound to polar response. It depends on room size and treatment. It gives great sensation when balanced the right way. And it's not the straight line EQing. And this is only one aspect of sound reproduction, still perhaps, the most important one. We've all been there and heard it (different ways), and, probably that's why this thread is still alive. Thank you for that.
Dr. Uli Brueggemann's Acourate software uses an objective measure called Interaural Coherence Coefficient (IACC), which is a measure of channel and room reflection equality for the first 80 milliseconds of sound travel.
IACC is a common standard especially for describing the spaciousness of sound in concert halls (Griesinger et al). The smaller the IACC value, the more spacious the sound. Acourate uses IACC in the opposite direction, the higher the measured IACC value the more that the two channels/room reflections are approaching equality.
A proposed approach would be to objectively measure the IACC of different speakers, with different directivities, in the same room and location. The speaker that measures the higher IACC would be the winner and perhaps answer the question what is the ideal directivity pattern for stereo speakers.
Anecdotally speaking for myself, increasing IACC correlates to improved stereo imaging and spaciousness or put subjectively, the speakers ability to *disappear* in ones listening environment.
IACC is a common standard especially for describing the spaciousness of sound in concert halls (Griesinger et al). The smaller the IACC value, the more spacious the sound. Acourate uses IACC in the opposite direction, the higher the measured IACC value the more that the two channels/room reflections are approaching equality.
A proposed approach would be to objectively measure the IACC of different speakers, with different directivities, in the same room and location. The speaker that measures the higher IACC would be the winner and perhaps answer the question what is the ideal directivity pattern for stereo speakers.
Anecdotally speaking for myself, increasing IACC correlates to improved stereo imaging and spaciousness or put subjectively, the speakers ability to *disappear* in ones listening environment.
80 milliseconds seems way too long for a small room. More like 20-30 ms. Even then, these measures for large venues seldom work the same for small rooms, to wit RT-60. Unless the measure has been shown to be effective in a small room it must be assumed that it is not.
IACC would likely indicate "spaciousness", but it is blind to "imaging". (Large rooms don't generally have imaging problems.) We know that increasing one will likely decrease the other. How do you balance out that dichotomy? Without knowing the time intervals for the reflections one cannot determine this tradeoff.
I would be tempted to exclude the first 10-20 ms and then use the 20-60 ms. range for "spaciousness". I am not sure how one would measure image quality except to look at the first 20 ms of the impulse response.
IACC would likely indicate "spaciousness", but it is blind to "imaging". (Large rooms don't generally have imaging problems.) We know that increasing one will likely decrease the other. How do you balance out that dichotomy? Without knowing the time intervals for the reflections one cannot determine this tradeoff.
I would be tempted to exclude the first 10-20 ms and then use the 20-60 ms. range for "spaciousness". I am not sure how one would measure image quality except to look at the first 20 ms of the impulse response.
Attached is an example of an Acourate IACC measure of my stereo speakers and room. It covers 0- 10ms, 0-20ms, 0-80ms, and 80ms - pulse end. This would suggest that Acourate does measure the time intervals for reflections.
The longer the impulse response is as close to identical for each channel, including room reflections, the better the stereo imaging. I suggest this has been known for sometime in the recording studio/control room world (where I come from). See Chips Davis AES paper on LEDE concepts as it was using TDS and a TEF computer that started to uncover/document this in the 80's.
Having been involved in a ground build of a few control rooms, this was the big test at the end, comparing L/R impulse responses for equality. Control room design revolved around ensuring one half of the room was identical to the other half, no parallel surfaces, proper diffusion, etc. Costs a small fortune.
With modern DSP software and the incredible computing power at our fingertips, it seems we can not only measure this, but can also apply correction algorithms to improve it. I am speaking from my anecdotal experience (and others like Ackcheng) linked above.
The longer the impulse response is as close to identical for each channel, including room reflections, the better the stereo imaging. I suggest this has been known for sometime in the recording studio/control room world (where I come from). See Chips Davis AES paper on LEDE concepts as it was using TDS and a TEF computer that started to uncover/document this in the 80's.
Having been involved in a ground build of a few control rooms, this was the big test at the end, comparing L/R impulse responses for equality. Control room design revolved around ensuring one half of the room was identical to the other half, no parallel surfaces, proper diffusion, etc. Costs a small fortune.
With modern DSP software and the incredible computing power at our fingertips, it seems we can not only measure this, but can also apply correction algorithms to improve it. I am speaking from my anecdotal experience (and others like Ackcheng) linked above.
Attachments
Interesting, but I am not sure what it means. I don't see why one would want a IACC that is high or low in the first few ms. I don't see what it tells you. A low IACC > 20 ms would mean good spaciousness, that's clear.
High IACC just shows how L-R is matched to get stable phantom imaging.
LOC from Griesinger is the new interesting thing in large-room acoustics. Measures image localizidedness(sp?). It basically predicts how well the cocktail party effect works in a noisy and/or reverberant environment. Havent seen so far how well it agrees with current phase-distortion data (at 700-4000Hz) but the implication is huge. Up till now phase-distortion treshold was considered quite high.
What this means is any reflection above treshold could interfere with separating different instruments from the mix, even in the 'spaciousness' region. Also linear phase at the midrange could be more important than previously thought. There is also the subtility of the effect - it could not be audible per-se, but could lead to lower fatigue levels. Just like a good acoustics in classrooms leads to better learning ability. Heretic thoughts.
LOC from Griesinger is the new interesting thing in large-room acoustics. Measures image localizidedness(sp?). It basically predicts how well the cocktail party effect works in a noisy and/or reverberant environment. Havent seen so far how well it agrees with current phase-distortion data (at 700-4000Hz) but the implication is huge. Up till now phase-distortion treshold was considered quite high.
What this means is any reflection above treshold could interfere with separating different instruments from the mix, even in the 'spaciousness' region. Also linear phase at the midrange could be more important than previously thought. There is also the subtility of the effect - it could not be audible per-se, but could lead to lower fatigue levels. Just like a good acoustics in classrooms leads to better learning ability. Heretic thoughts.
^
Heretic? Not in my mind 🙂 There was a guy some 80 years ago talking about a "shuffler". His name was Alan Blumlein:
http://www.audiosignal.co.uk/Resources/Stereo_shuffling_A4.pdf
Heretic? Not in my mind 🙂 There was a guy some 80 years ago talking about a "shuffler". His name was Alan Blumlein:
http://www.audiosignal.co.uk/Resources/Stereo_shuffling_A4.pdf
How about that, history repeats itself? Soon we will build pyramids to honour our leaders (voluntarily?)
Or the extra terestrials will return to help us build more 😉
I beleive Orban Parasound used these priciples in some of their equipment I used back in the '70s and '80s.
😀
I read Griesinger's papers on LOC and I have to agree that, to me, that measure makes a lot of sense. His whole concept of phase integrity in the 700-7k Hz range seems completely plausible. Its not that the ear is sensitive to phase, per see, but it is sensitive to modulation, which is itself phase sensitive. It explains so much. I buy it completely.
The very early IACC just doesn't fit with me for imaging judgments.
The very early IACC just doesn't fit with me for imaging judgments.
Very educational walkthrough - thanks! Considering playing with Acourate in near future. Just cannot find FIR-capable 64bit DSP that could be coupled with multichannel I2S + multiple amps. The moment mini-PCs will settle inside speakers is approaching. Sorry, I meant tablets, not PCs 😀
Thanks for your positive opinion. Do you recall the name of these papers?
Citation about interesting findings about human hearing timing sense accuracy here : "Animals and humans use small differences in the arrival time of sound at each ear to locate its source. In fact, the human brain can detect differences in the arrival time of a sound at the two ears of about 10 millionths of a second." Well, if I got it right, that's distance as short as wavelength of 1GHz. Wavelength of this frequency in free air equals 0.34mm. Is this the target margin for accurate time alignment of drivers inside multi-way speakers? And also for correct placement of the stereo speakers? Rather hard goal, but getting as close as possible may result in more "easy" listening as ear/brain doesn't need to compensate and mask obviously present timing errors relative to different frequency bands to adapt and make up everything sound "realistic enough".
And the second citation: "For lower frequencies the UCL team believes that humans use a different system also used by small mammals like gerbils and guinea pigs." It would be interesting to know which is the frequency boundary for "switching" to "other hearing method" so we could define these areas as critical by different means. My guess is it would match Griesinger's lower boundary of 700Hz.
^
PRTG,
you might be interested to look into this:
Mechanisms of Sound Localization in Mammals
BENEDIKT GROTHE, MICHAEL PECKA, AND DAVID McALPINE
It is a later in-depth publication of some UCL team members.
PRTG,
you might be interested to look into this:
Mechanisms of Sound Localization in Mammals
BENEDIKT GROTHE, MICHAEL PECKA, AND DAVID McALPINE
It is a later in-depth publication of some UCL team members.
however such loudspeakers are losers in listening tests according to Floyd Toole
again this is contrary to the scientific evidence
ditto
Toole, "Sound Reproduction: The Acoustics and Psychoacoustics of Loudspeakers and Rooms ", page 119:
Last edited:
I think we need be careful to draw too conclusive conclusions from Toole's tests in this area. He's not very conclusive himself if you look at this answer below.
The "Sensitive Listener's" article is interesting in that regard and Toole touches that himself in the comment below. It may be that listening over both for a period of time changes the preference. Personally I've always preferred to remove the side reflections in small rooms. Especially when it's combined with a later arriving diffuse field from behind. The diffusion from behind is the closest you get in a small room to the 20-25 ms lateral contribution in concert halls. Toole by the way, never conducted a test with a highly diffusive rear wall to my knowledge, and he also point out that stereo is limited in spaciousness. To increase spaciousness, is one of the reasons to use diffusion in a LEDE/RFZ matter. Multichannels with rear speakers is his approach rather then diffusion and they are certaintly not the same. One uses a passive contribution, the other an active. Passive with diffusion sounds much more natural and pleasant IMO. In other words, there can be variables here that can change the result of such tests.
His comment on reverberation time shows IMO that he misunderstands what true reverbaration is. It simply doesn't exist in a small room within the strictest definition. Well known among those who have studied small room acoustics.
Last edited:
Come on man, read everything.
In this studies it varied with type of music and he only concluded with something like "it seems like there's a preference for lateral reflections with some music material".
And isn't worth taking notice that most hard core audiophiles end up with removing sidereflections when they try both?
Last edited:
well, I read everything:
perhaps I am wrong but the above quoted for me clearly reads: "whenever there are spatial cues in the recording itself the recreation of this space is not hindered by a low IACC stereo setup and sometimes it is even enhanced by it". Right? Perhaps with one exception of pure Blumlein recordings which are rather a rarity today.
Well, most "hard core audiophiles" are seriously misguided and strongly biased persons strictly following fashions and whose biases are formed under decisive influence of their gurus which are typically audio journalists and sales people.
I agree with SL:
Maybe. I woudn't draw a conclusion based on this alone though.
Don't fall into the trap of only making arguments based on a desire to "win the debate".
I think you well that I'm not talking about audiophooles who are worried about whether to use rhodium or gold plated contacts. And how could people be influenced by audio magazines or sales persons on this topic when it's never addressed by them! When did you last read an article in one of those magazines about attenuating specular reflections with broadband treatment? Or a salesman that talked about it?
Start sticking to the topic please.
Those who experiment with acoustic treatment almost all prefer attenuting the closest sidewall reflections and in most cases also the opposite wall. And often it's done by too thin panels that only absorp part of the reflected energy. In other words, far from optimal. As a side-note here, the way Harman Int. room is treated is almost a joke to a small room acoustic professionalist.
Like mentioned before there too many variables here in order to draw a conclusion. The studies on the preference on this is extremely limited and without taking other considerations into account. What for example if you not only have a diffusive rear wall but also removes sidereflections with angled panels instead of absorbtion. Or even better, build a room with splayed sidewalls which is often done in studios. Will this change people's preferences? There are many different approaches here and it would be wise to not be too conclusive.
The reason why Toole is so often quoted is basically because he published his work. The fact that sidewall reflections can sometimes be preferred was well known decades ago by the SydAud Community. However, they publishes very little of their researches and kept their knowledge hidden and for those attenting seminars. Whether it's preferred or not is however depended on several factors. One has to experiment for oneself to find the answer to the specific room, with the specific speakers and the music material one listens too.
Perhaps, ok, almost, ok.
But could it be that this is because they learned to like a specific sound? Exactly just like those professionally biased people that Toole points at?
A lot of choices is made out of such a accumulated bias or just out of a habit.
which is? For example?
ok!
The point I make is exclusively against such a conclusion, the one proposed by Mitchba:
is it all right?
- Status
- Not open for further replies.