Acoustic design, 3D scanning and calibration — from calculation to the first movie.
This is not an opinion or a salesman’s stance. It is the conclusion of peer-reviewed acoustics. Harman’s research (Floyd Toole, Sean Olive) showed that below the transition frequency around 300–500 Hz what a person hears is determined primarily by the interaction of the speakers with the room — modes and boundary effects — rather than by the «box» itself.
According to Harman research, speaker and listener placement alone produces a spread of more than 18 dB in the room response below 300–500 Hz, and between different rooms the spread reaches 25 dB below 100 Hz. For comparison: a competent amplifier or DAC is transparent in level-matched blind tests — less than 1 dB difference. The room error is an order of magnitude larger than any electronics error.
We are not claiming that multichannel sound is a hoax. In a properly built and treated space, 5.1 and Atmos deliver their result. The problem is different: Atmos is designed for a specially built room, and most buyers do not have one. And then physics rules, not the channel layout.
In an untreated space, every additional channel is one more source of reflections and standing waves. Below the transition frequency the room modes dominate: SBIR nulls up to 6–25 dB deep, boundary effects, comb filtering. Adding a sixth or eleventh speaker to such a room multiplies the chaos, not the spaciousness. This half of the sound is cured by no amplifier, no processor, and no channel count — only by designing for the room.
A designed installation demonstrably removes 10–25 dB of audible room error — an order of magnitude larger and more audible a gain than any electronics upgrade at the same budget.
The main argument against the chase for channels was formulated by the engineers at Grimm Audio — a maker of studio monitors on which music is mixed. Their thesis: stereo does not physically reproduce the original sound field. It is a psychoacoustic illusion, and that is exactly why a pair of accurate sources can build a convincing soundstage.
The phantom image between the speakers is built by the brain — so what matters is not the number of drivers but the precision of how the system radiates energy into the room. Above the transition frequency, timbre is set not by a flat on-axis response but by directivity and the ratio of direct to reflected sound. According to Harman research, the listener’s ear weights the sound roughly as 12% direct / 44% early reflections / 44% late energy — meaning about 88% of perceived timbre comes from the room, not directly from the speaker.
The same Grimm engineers dropped the beryllium tweeter in their flagship and switched to a carbon-fiber one with near-identical measurements. The reason was the toxicity and manufacturing difficulty of beryllium, not the sound. The conclusion is direct: an exotic membrane is not what determines the result. What determines it is systems engineering: designing for the room, directivity, and integration.
There are third-party-measured precedents that controlled directivity matters more than a flat on-axis response and channel count. We design by these same principles.
A cardioid pattern reduces the sound going backward and thereby sets the ratio of direct to reverberant field. Above the transition frequency this improves intelligibility and reduces the audibility of reflections.
Placing the woofers close to the front wall merges the speaker and the wall into a single source: the front wall provides bass boundary gain rather than damage. Measurements show up to +6 dB of bass headroom below ~80 Hz.
The bass boundary dip (the Allison effect) is predictable: with the woofer ~0.9 m from the wall the dip falls around 113 Hz, at ~1.2 m around 85 Hz. This was measured by Roy Allison (AES, 1974). Close placement lifts the dip out of the bass range. We calculate this before installation rather than chasing it «by ear.»
We do not sell a number of channels or the most expensive box. We sell a «room + integration + directivity» system — the variables that Harman science proved to be dominant.
It works — but in a specially designed and treated room. Most buyers don’t have one, and then the sound is determined not by the channel layout but by room physics: below 300–500 Hz the modes and boundary effects dominate, giving an 18–25 dB spread. We don’t deny multichannel sound — we say that without room treatment it doesn’t open up, while two accurate speakers designed for the room give a predictable result for less.
Stereo is a psychoacoustic illusion: the phantom image is constructed by the brain, not physically reproduced from the original field (as the Grimm Audio engineers put it, on whose monitors music is mixed). So what matters is not the number of drivers but the precision of how the system radiates energy and interacts with the room. A pair of speakers designed for the space builds a convincing soundstage — and without spawning uncontrolled reflections.
Because it is proven by measurement. Sean Olive’s preference model (Harman) predicts blind-test results with a correlation of 0.86 from anechoic speaker data alone — meaning the speaker and the room set the quality, not the electronics. Meanwhile the room error (18–25 dB) is an order of magnitude larger than the difference between amplifiers (less than 1 dB). Money spent on treatment and designing for the room delivers an audible gain that a more expensive box will not.
Because we don’t buy channels that create chaos in an untreated room. On a real project a competitor offered 7.1.4 for 12M ₽ in an untreated space. We built a two-speaker system designed for the room with acoustic treatment for 4M ₽ — and the sound came out better, because the budget went into removing 10–25 dB of room error rather than into extra drivers.
Our offices are in Moscow and St. Petersburg, and we travel across Russia. The initial 3D scan and acoustic audit are done on site. Calculation and design are done remotely. For sites outside Moscow and St. Petersburg we discuss travel terms individually.
Describe the room — we’ll reply with a preliminary estimate. We’ll show how many decibels of error can be removed at your site.