An ambient and imaginative moment for female chant, cymbals, gongs, rabab, bamboo flute, fiddle, synthesizers, percussions, crystals, bowls, rainstick and beats.

https://youtu.be/C-Dx6VU5GZM?si=vzvROkPdnREFyG6Z

Don't know how useful this will be on here, but I figure it's worth posting for posterity's sake. I've had to remove most of the links to get it past the filter so please forgive.

In the late 1970s, a research team started by Yuriy Berendyukov under the supervision of Yuriy Kovalgin at the Leningrad Electro-Technical Institute of Communications (now Bonch-Bruevich Saint Petersburg State University of Telecommunications) investigated the psychoacoustical shortcomings of Western quadraphonic formats such as SQ, QS, and CD-4. Much like Wendy Carlos (Wendy Carlos' Quadraphonic Folly) and Michael Gerzon (Surround-sound psychoacoustics, Wireless World, Dec. 1974) the Soviets found the typical four-corner square layout yielded poor phantom imaging and localization.

After many statistical psychoacoustic tests, Berendyukov found that the best four-speaker layout for accurately locating phantom images was a trapezoid, with the listener situated near the bottom baseline. Eventually, a simple polarity-based encoding/decoding matrix was developed, and the system was dubbed ABC in reference to the three phantom zones the system tries to retain. The system's stated goal was to reproduce a 360-degree panorama, such as the full ambience of a concert hall.

Eventually, a demonstration record was pressed in 1978 along with a full album in 1980 with Berendyukov's country-folk band. In 1982, an op amp decoder along with a general overview of the system was published in the Soviet DIY electronics magazine Radio (#9, 1982).

As it stands, I've been in contact with Berendyukov for the last few months and he's graciously allowed his dissertation abstract to be made public. It can be found here. Unfortunately, the full dissertation is lost to time and even Berendyukov himself no longer owns a copy. On the plus side, a book worth checking out is Kovalgin's Акустические основы стереофонии (Acoustic Foundations of Stereophony (1978)) which not only touches on Blauert-level stuff but also has a small chapter on Berendyukov's initial SQ/QS testing.

Both men are still alive and agreed to meet up and reminisce toward the end of the summer. If anything archivable is shared, I'll be glad to post some links.

Hi everyone,

I'm Philip, a solo developer and founder of Dimenwave Technologies from Trinidad & Tobago. Over the past few years I've been developing a cross-platform media player called DimenPlay, and alongside it I've built a custom spatial DSP system called TASS (Time-Aligned Spatial System).

The reason I'm posting here is not to make a “my technology is better than everything else” claim. Quite the opposite - I'm at the point where I want people who understand spatial audio to challenge the work.

A little context about DimenPlay

DimenPlay is designed as a cross-platform media player that brings together local high-resolution playback, network libraries, and streaming sources within one application. The player is built around flutter framework, with the TASS engine integrated directly into the playback pipeline. The feature can be enabled and disabled directly in the media bar.

Some of the features surrounding the spatial engine include real-time spatial controls, different listening modes, high-resolution audio playback, and visualization/analysis tools. The intention is for TASS to be part of an actual listening workflow rather than a standalone DSP experiment.

I'm mentioning this because the application itself is part of the equation: the goal is not simply to demonstrate an effect, but to investigate how this type of spatial processing behaves when integrated into a complete playback system.

What is TASS?

TASS takes a different approach from conventional HRTF-based spatialization. Rather than attempting to model the listener's ears and head acoustically, its spatial processing is centered around controlled temporal relationships between stereo channels.

The original observation that led me down this path was surprisingly simple: when two sources are positioned very close together around the center, there is a point where they begin to perceptually blend even though they remain technically separate signals. 

That led me to investigate whether manipulating temporal relationships could produce useful spatial separation without relying on an HRTF model.

Since then, I've built the system into a real-time audio engine and integrated it directly into DimenPlay.

What I'm testing

I'm currently working toward a more formal comparison between:

• conventional stereo
• an established binaural/spatial rendering approach
• My DimenPlay app and how it handles how one enjoys the features that compliment the TASS engine.
• TASS effects and how it plays with tracks not mastered for surround sound like effects.

I'm measuring things such as timing relationships, phase behavior, spectral changes, channel correlation, dynamics and other signal characteristics, while also conducting controlled listening tests.

I'm particularly interested in determining which observed differences are genuinely spatial and which might simply be artifacts of spectral coloration, phase behavior, level differences, precedence effects, or other psychoacoustic phenomena.

And this is where I would really appreciate criticism.

If there are measurements, controls, test methodologies, or known spatial-audio considerations that you think I'm overlooking, I'd genuinely like to hear them.

There is a technical overview and demonstration of the current DimenPlay/TASS implementation here:

DimenPlay

I'm especially interested in feedback from people experienced with spatial audio, psychoacoustics, binaural processing, stereo imaging, or real-time DSP.

I'm not presenting TASS as a finished scientific conclusion. I'm presenting the current implementation and asking the community to help pressure-test it.

TASS system

If the data supports the hypothesis, great.

If it doesn't, that's useful too.

Thanks for taking the time to read this, and thanks to the moderators for allowing me to share the project. Just trying to do some honest work here.