I'm well aware of your study. One limitation of the study is that the acoustically treated “RRF” condition is not described in sufficient detail to establish its frequency-dependent behaviour. The paper states that movable drapes were repositioned and “a few patches of sound-absorbing materials” were added to attenuate early reflections and reduce reverberation, but the type, thickness and absorption characteristics of these treatments are not specified. Therefore, while the study clearly compares a normal room, a relatively reflection-free room and an anechoic condition, it does not provide a controlled comparison between broadband absorption, frequency-selective absorption and diffusion.The audibility and audible effects of early reflections are undoubtedly important, which is why many years ago Sean Olive and I conducted extensive experiments to determine those thresholds.
Olive, S. E. and Toole, F. E. (1989). “The Detection of Reflections in Typical Rooms”, J. Audio Eng. Soc., vol. 37, pp. 539-553.
The abstract is:
“Reflected sounds influence the timbre and spatial character of live and reproduced
sounds. Most investigations of reflections have focused on the performance of live
sounds in large halls. Current interest in the acoustical interactions of rooms, loudspeakers,
and listeners requires further and possibly more relevant data than have been
available. In this study, the effects of reflected sounds were examined as they would
occur in stereophonic reproduction in rooms of domestic or control-room size. Thresholds
were determined as a function of level relative to the direct sound, the angle of incidence,
spectrum, the temporal form of the signal, and reverberation. The relationships between
audible effects and measurements, such as energy-time curves and frequency response,
are discussed.”
If you read this, and some of the data from it repeated in my books (Chapter 6 in the 4th edtion), it is evident that common interpretations of the Haas effect, Haas kickers, the effect of additional reflections, etc. are simplistic or wrong. The widely assumed “fusion zone” of 30 ms applies only to speech in an anechoic environment, yet it gets trotted out for numerous applications involving music in rooms. Few people have taken the trouble to understand the established science, and persist in perpetuating old school notions.
I'm wondering if you are misinterpreting what Don Davis wrote about the Haas effect if that was what you were referring to. He defined the Haas effect in the same paper as the following:
"The Haas effect is the inability of the brain to discriminate between echoes and delays of sounds that arrive approximately 10 to 20 ms after the original waves."
"The auditory system temporarily fuses sounds in this 20-ms zone"
There's no indication that he says discrete reflections cannot be heard. Quite the opposite, he strongly believed they were, and LEDE was very much about avoiding such reflections with a certain time window to hear the recorded signal better.
Davis appears to have regarded around 20 ms as a particularly useful delay: long enough for the direct sound to establish clear localization, yet still within what he considered the range of maximum Haas effect. He also noted the coincidence that Beranek had identified a similar ITDG (around 20 ms) as characteristic of the best concert halls at that time.
The Haas kicker/trigger was an experiment early on before diffusers arrived. The idea was the with a strong late arriving reflection, it would cover over many of the late arriving reflections. The mind hears the first strong reflection a lot better and thus naturally isn't able to focus on the other ones. It was an attempt to avoid making rooms dead with only absorption of later arriveal reflections. It works that way also IMO (have tested it), but the strong first arriving reflection is also detrimental. And it was disposed when diffusers entered the scene. So it basically wasn't part of what the LEDE eventually developed into, but an experiment along the way.
I think this may be missing Davis's actual point. He was not arguing that a LEDE control room should reproduce the same reflection pattern as a concert hall, nor that all desirable spatial effects are produced by a single 20 ms reflection. His point was that the LEDE principle could produce an initial delay gap in roughly the same region that Beranek had associated with the best concert halls, and that Davis regarded this as coinciding with a favourable region for what he called the maximum Haas effect.This confuses a number of things. Concert halls need diffusion to convey the orchestral energy to all parts of a widely distributed audience. What is heard in a concert hall is a complex comb filter that is very pleasant, but mostly for those who can afford the good seats. Not all reflectied sounds are perceived the same way. The reflected sounds that create the desired envelopment arrive after delays of 60-80 ms and more, arriving from about 60 deg (in the good seats), not 20 ms arriving from the rear in a LEDE room. As shown in the above paper and in my books the precedence (Haas) effect is very different for different sounds, and it changes enormously with the addition of reflected sounds whether they are in the room or in the recording. There are no “single number” generalized solutions. The Haas "kicker" is a based on a significant misinterpretation of published literature, as was tested in the above paper.
Also, the research on concert halls does not seem to support such a simple contrast between reflections around 20 ms and desirable spatial effects appearing only at 60–80 ms. Barron's experiments showed that lateral reflections could produce spatial impression over a broad delay range of approximately 10–80 ms, with direction and level being at least as important as delay itself.
This is also consistent with the work of Schroeder, Gottlob and Siebrasse, whose comparison of European concert halls found interaural coherence to be a much stronger correlate of subjective preference than ITDG alone. Ando's work is also relevant here. While he did not propose a universal optimum delay, his preference studies showed that the preferred delay of an early reflection could fall around 20 ms under particular conditions. More generally, he found that the optimum initial delay depended on the temporal characteristics of the music, as well as the total strength of the early reflections and subsequent reverberation. Thus, a preference for a delay around 20 ms is not inconsistent with Ando's work; it is simply not a universal rule.
So I agree that there is no single-number solution, but that cuts both ways. The concert-hall literature does not support the idea that the desirable spatial effects are uniquely associated with reflections arriving at 60–80 ms or later. Early lateral reflections, including those in the region around 20 ms, can clearly contribute to spatial impression and subjective preference, depending on their level, direction, temporal structure and interaction with the rest of the sound field.
I think this is where your argument becomes too categorical. Your an Olive's experiments on reflection audibility were highly valuable, but they did not, as far as I can see, compare specular reflections with the same reflected energy redistributed by physical diffusers. A diffuser does not simply “leave the energy in the room”; it changes the spatial and temporal structure of the reflection. That distinction seems particularly important given the extensive research showing that lateral and temporally distributed early energy can affect spaciousness and preference. D'Antonio and Cox, for example, explicitly describe diffusion of early reflections as a means of mitigating coloration, image shift, harshness and acoustic glare.Diffusion of the sound field in listening rooms is of no benefit. In fact, if too energetic and from the wrong direction it degrades the delivery of directional and spatial information encoded in the stereo recording. Reflected sounds arriving from the sides are most effective in creating impressions of space - that is how the binaural hearing system works. Even a single synthesized reflection from the side in an anechoic chamber is perceived as "spaciousness" not comb filtering. Humans are programmed that way.
Diffusers in small rooms should be used to attenuate reflections by sending most of the energy off in other – hopefully non-destructive – directions. However, the energy remains in the room, which may not be an advantage. The performance of diffusers that matters most in small rooms is how they alter the attenuated sound reflected in the direction of the listener, a performance factor that is rarely specified. The gratings cause spectrum changes and phase shifts in the reflected sound. Fortunately humans are not sensitive to phase shifts, and the magnitudes are trivial compared to the delays involved, but the spectrum changes in some diffusers are significant. Peter D'Antonio has discussed this and published some data.
And as desribed earlier; the type of diffuser and how it's used matters greatly. In LEDE rooms, they were used only for late arriving reflections (rear of the room). And while it isn't possible to attain the same result as shoe horse concert halls, there's still diffused energy that arrives, but laterally from the rear as oppose to from the front. Not quite the same, but still highly audible.
Diffusers used this way don't interrupt with the IDTG of the recorded room but adds spaciousness while dealing with late high gain reflections. Very seldom do you see a diffuser used this way change the frequency response much. If deviations happen they are generally very narrow and likey not very audible, if at all. A moden diffuser has been improved compared to the old standard QRD.
That being said an absorber is spectrally more neutral and would be the better choice if one only focuses on the frequency response. But as those of us who have epxerimented with different types of treatment; if only absorption is used for later arriving reflection we end up with something mush less preferable.
I do sell acoustic products, but I do not sell to my customer base here. My customers are almost exclusively in Norway, and primarily in the professional sector. I also develop loudspeakers and would probably have far greater commercial potential by promoting multiple loudspeakers and selling into the audiophile market. My interest in this discussion is therefore not related to any financial benefit.I understand that you are in the business of selling acoustical devices, and I hope your business prospers, but what is stated in this quote are your personal opinions - subjectivity.
I have my opinions too, but from the time I began my research in 1966 I have used the opinions of others, hundreds of them in controlled experiments to focus attention on specific variables. The accumulated data provide a basis for understanding the relationship between what is measured and heard. A single omni mic in a room is not the same as two ears and a brain, and raw ETCs are spectrum blind, not providing the necessary information to assess the audibility of reflections or the multitudes included in diffuse sound fields. The final answers to all questions are not yet discerned, but a genuine scientific approach is needed to clarify those remaining.
In the meantime, so long as recording engineers listen in circumstances that please themselves as individuals the opportunity for the recorded art to transfer to the listener in typical domestic rooms will remain a lottery. An extreme example of this is described in Section 12.5.5 in the 4th edition, where Blackbird Studio C is described and discussed. The design goal for this extravagantly expensive space was to provide George Massenburg, an influential recording engineer, with a wide sweet spot and an acoustically immersive experience. Peter D'Antonio, a genuine expert on diffusers, designed a room with all surfaces covered with 1 m deep broadband 2D diffusers. Sadly, after the effort and expense no new knowledge appears to have emerged. Recording engineers apparently adapted to the unique sensations of the room, and it was (is?) used as an Atmos mix room. Whatever is heard in this room it is not replicated in any audiophile listening space - the ultimate target audience - much less in the rooms occupied by masses of everyday audio enthusiasts, who should be the target.
The widespread use of near-field listening, with or without acoustically dead, reflection eliminated rooms is the extreme opposite of the Blackbird studio. The middle ground is occupied by countless variations.
Recordings need to work in ordinary rooms or they miss the target audience. Now that most listening is done using headphones that argument becomes a somewhat academic question.
At least there are strong indications that monitor loudspeakers have moved in the direction of "high fidelity", neutrality, technical accuracy. This aspect of the circle of confusion is showing signs of rationality.
.
Blackbird Studio C has little in common with a LEDE design, and I do not think it is particularly relevant as an example of diffusion in small listening rooms. Covering virtually all surfaces with very deep 2D diffusers means that the reflected energy is repeatedly redistributed and diffused. This is essentially the opposite of the LEDE approach, where preserving useful reflected energy was important and diffusion was used in a much more controlled manner.
Such an extreme application of diffusion may well create a highly immersive and unusual acoustic environment, but it is not surprising that it could also compromise the precision of the stereo image. It was an interesting experiment, and I am sure Peter D'Antonio learned from it, but I do not think the apparent shortcomings of this extreme room can reasonably be used as evidence that diffusion itself provides no benefit in listening rooms.