Originally Posted by craig john
Let's change the subject, as it's clear that no one's minds will be changed on the topic of imaging.
@Floyd Toole This a purely theoretical question. I was reading your article entitled "Subjective Measurements of Loudspeaker Sound Quality and Listener Performance" published in 1984. In Section 1.2 on the history of subjective testing you make the following comment:
You were referring to studies that had been done in the '40's and '50's, 30+ years earlier.
I then read this comment you made the other night in this thread:
We are now 30+ years past the publication of your papers. Audio technology overall has advanced far beyond what it was back in the '80's. Back then, the sources were analog vinyl or tape, both noisy and deficient in other ways. These days, we now have "object-based" 15+-channel systems with a separate channel or two just for LFE, Digital Signal Processing, Room Correction, Bass Management and EQ, and much more powerful amplification. The sources are digital, noiseless and lossless. They have Dynamic Range capability of 105 db, (compared to tape at 55 dB and vinyl at 70 dB.) They come to us on shiny little discs or over a wire connected to the internet.
In light of all the advancements that have been made in audio, do you think it is time to re-evaluate these findings? For example, one of the primary determinants of sound quality back then was "spaciousness." With the advent of high speaker-count, multi-channel systems, (like yours and mine, 9.3.4), is spaciousness less valuable and high directionality more valuable? Imaging no longer depends on fragile phantom sounds placed in the sound field, but only "heard" from the sweet spot. Imaging is now done with "sound objects" that move from speaker to speaker to speaker, and can be heard all over the room. How do you think monophonic testing applies to speakers used, not in 2-channel, but in high speaker count systems? In general, are there properties of speakers that are more beneficial in high speaker-count systems, that were not realized as being important 30+ years ago in 2-channel systems?
If you were still actively working in the field, what new studies would you be looking to do?
Thanks for your thoughts.
I don't know how one would implement "high directionality" loudspeakers in each of the many locations in a high-channel count immersive system - to be directional at long wavelengths they need to be large. Dolby Atmos can go to tens of "immersive" channels in cinemas. The only practical method is to use conventional speakers and cover all internal surfaces with highly absorbing material - an anechoic space - thereby accentuating the direct sound, but it is not a nice place to live.
A short story. I was given a private demonstration of Atmos by Dolby when they introduced the home version. To begin with I was exposed to the "real thing" in a carefully calibrated (they said) dubbing stage, where soundtracks are created. They played an excerpt from "Gravity" which has sounds emerging from different directions. They did, but the sound quality was truly mediocre. I complained, because, truthfully, the night before I had played the film at home through my then 7.1 system. It was colored, and it sounded as though more than X-curve calibration was involved.
Then they played a male voice panned around the room. In 7.1 it predictably jumped from speaker to speaker on the sides and rear, but it always sounded like the same voice. When panned in Atmos, the movement was more uniform - no surprise with more channels - but the voice quality changed dramatically with location, becoming quite timbrally distorted in some locations. I naturally complained, and was greeted with the comment from a senior executive that "most people don't notice that". So "gee whiz' localization took precedence over timbral consistency and accuracy. I told my colleagues at Harman, and when some of them went the next day for supposedly the same experience, that demo was mysteriously "not available".
What was going on? With object-based directional panning the sound is shared among multiple loudspeakers in an ever changing manner as sounds move around. The resulting acoustical interference among the active speakers cannot be constant; it changes with position as different speakers in different locations are engaged. This is not mysterious - it is predictable. And it will be different for every installation, especially with different channel counts.
When I discussed this with Wilfried van Baelen, the inventor of Auro3D he understood, explaining that this was why his scheme was an addressed-channel system.
So, to answer your question about what studies I would like to see undertaken, I guess one of them would be to establish a scientific foundation for immersive audio, and force all the profit-motivated players to abide by the directives so that we don't end up with the mess we have. Right now SMPTE itself is debating where to put loudspeakers so cinemas can play all three of the formats. Commercially Atmos has won, but that does not make it be best or an optimum system. Perhaps that has yet to be created.
Fortunately for the corporate interests, consumers are easily amused by sound coming from different directions, and don't pay attention to fidelity as some of us in this forum do. For me, the most impressive demonstration of a concert hall and a cathedral performance was another private demo by Wilfried using his recordings in Auro3D. Absolutely breathtaking, was my response - I was "there" and could walk around the home theater and it didn't change. But this was a manufacturer's demonstration, and the circle of confusion did not exist - although it took place at a CEDIA event. Still, if it can be done once, it can be repeated. We need more . . .
For those of you with the playback capabilities there is an interesting recording: Delphian is the label, and it is a performance of Praetorius "Missa Tulerunt Dominum meum" recorded in Dolby Atmos, 9.0 Auro 3D, 5.0 DTS HD MA and stereo 24/192 PCM. One can switch at will among the options as it plays. The mixes were done by reps from the companies. For my taste, Auro 3D won hands down. But that is me, in my room