Room modes
Before correction, energy between 35 and 50 Hz sat about 6 dB above the target band, as shown in the correction software’s design view.
Case Study
The CQS Reference Theater.
A dedicated Dolby Atmos home theater of about 14 by 16 feet. It is our own room, used to develop and test the method before we apply it to yours.
RoomAbout 14 × 16 ft, dedicated
FormatDolby Atmos
ToolsREW, Dirac Live, calibrated microphone
DocumentedMay 2026 record, revised as the room changes
Most write-ups of a theater show the equipment. This one shows the process: what we measured, what we changed, what the data showed afterward, and what it does not show.
The challenge
A mixed-brand system in a real room.
Subwoofers fighting
Two subwoofers cancelled each other around 60 Hz, because their tuning and phase behavior did not match.
Mixed speakers
Several brands in seven channel groups needed one consistent target, not channel-by-channel guesses.
The center channel
Dialogue depends on the center. Its response had to be measured on its own, with everything else out of the way.
The method at work
Measure. Optimize. Document. Repeat.
- 01
Measure.
Every speaker was placed against a laser plane set at ear height, so the physical placement was right before any filter was designed. Then we measured with a calibrated microphone and REW.
- 02
Optimize.
One subwoofer was phase-inverted, and the result was checked at five seating positions. One target curve was applied across all channel groups in Dirac Live.
- 03
Document.
Measurement files, correction projects, the target curve and the physical geometry were saved, so the whole state can be reloaded and checked from the records.
- 04
Repeat.
Our first center-channel result did not survive our own review. We corrected the measurement signal path, measured again under controlled conditions, and rewrote the claim to match what the data showed.
What the measurements showed
Each claim, with its evidence and its limit.
Subwoofer cancellation near 60 Hz
Result. Substantially reduced after a 180-degree phase correction on one subwoofer.
Evidence. Paired measurements with the phase at 0 and 180 degrees, at five positions across the seating area.
Limit. Reduced, not eliminated. Cancellation in a real room cannot be removed by phase alone.
Center-channel dialogue range
Result. Across 200 Hz to 4 kHz at the main seat, the response varies by about 12 dB from peak to trough. That is dominated by one narrow notch near 250 Hz. Excluding the notch, most of the vocal range stays within about 4 dB either way of its midpoint.
Evidence. Center-only measurements with correction on and off, plus an immediate repeat that matched. The notch appears in both, so the correction did not cause it.
Limit. One seat. We have not confirmed what causes the notch, and we do not claim a perfect response.
Consistent target across mixed speakers
Result. Tonal cohesion across the system was improved by applying one target curve to every channel group.
Evidence. Correction-software filter designs for all seven channel groups.
Limit. In-room measurements after correction exist for the front stage. Surrounds and overhead channels are still to be measured.
Stated limits
What this case study does not claim.
- It is one room, measured mostly at one seat. It says nothing about the other seats yet.
- It is our own theater, not a client project. We show it so you can see how we work.
- It is an engineering record, not a listening review. How a room sounds is subjective, and we keep that apart from the measurements.
- The room changes over time. Each change is measured again and the record is updated.
What you receive.
The same approach, scaled to your room and your service. Depending on the service, that includes a written summary or full report with before-and-after measurements, the settings and filter files, and a record of what was changed so it can be maintained.
Gallery
The room, in pictures.
Photos from our own home theater.