The Journal / Articles & reviews
Room Correction vs Acoustic Treatment: What DSP Fixes, and What It Costs You
Correction can measure better and still sound less alive; treat the room first and the box has less to do, sometimes too little to justify the conversion.
I took the acoustic panels off my walls and let the McIntosh MEN220 measure the room. It came back wanting to correct 25%. The panels went back up, the same processor measured again, and the number fell to 7%. Same box, same room, a fraction of the work. For years I had changed gear to chase what I heard. That pair of numbers was the proof of where the problems really lived.
So, room correction or acoustic treatment? Both, and the order is the whole answer. DSP room correction is superb at a narrow job: bass peaks and balance at the listening seat. It cannot do the room's job, and in a chain with tubes and vinyl it charges a price the graph never shows: air, decay, life. The numbers below come from my own processor, and they point one way: walls first, filters last.
Twenty-five percent, then seven: the test that settled the order
The McIntosh MEN220 is a dedicated room correction processor built around Lyngdorf's RoomPerfect, under license from Lyngdorf Audio A/S. I knew there were other systems out there. I chose McIntosh for the least scientific reason in this hobby, and admitted it on camera: I'm a McIntosh boy.
Before the test I had done the cheap work: the speakers positioned, then absorption on the back wall and parts of the front wall, with diffusers carefully placed among the panels. First reflections were my biggest problem, because my seat sat very close to the back wall.
The result is in my October 2025 film on room acoustics, and my reaction on camera was not scientific either: "This is staggering. This is a colossal difference." All I had really done was play with the speakers and install a few thousand or less in acoustic treatment. The percentage is the MEN220's own yardstick, and its meaning is the before and after: on that scale, 7 is less than a third of 25.
The panels did most of the processor's work, without touching a single note of the signal. I put the whole test into a single Daily: From 25% correction to 7%.
What room correction actually does, and what it cannot reach
Every DSP room correction system works in three moves: measure, calculate, filter. A microphone at the seat listens, software builds filters, and the filters go into the signal path. That is the promise. Physics writes the small print.
In the bass, the small print is generous. Floyd Toole wrote in The Acoustical Design of Home Theaters (1999) that room resonances at low frequencies behave as "minimum phase" phenomena: fix both the amplitude and the phase, and the transient response is fixed too. A filter that tames a booming bass note also tames its ringing. This is DSP's home ground.
Higher up, the room behaves less like one big resonance and more like a hall of mirrors: late copies of the music from the side walls, the ceiling, the glass. In the room that held my first real system, reflections caused comb filtering and smeared the soundstage; no amount of gear could fix that. A filter can reshape what the microphone hears at one point. It cannot stop the wall from sending the copy.
Even Dirac's Active Room Treatment stays in the bass. In an April 2023 StereoNET interview, Dirac's Head of Product, Nilo Casimiro, said it uses the speakers "to cancel out room resonances and reflections in the bass region," with a cancellation signal "limited in frequency and level."
Read every label as live text
- Bass region: DSP territory
- Above the bass: the room's territory
- Room resonances behave as minimum phase (Toole, 1999)
- Fix amplitude and phase: the ringing goes too
- 60 Hz: a 10 dB peak or null is what you hear at the seat
- 25 Hz and 50 Hz filters cut peaks and ringing (Mellor, 2011)
- Dirac ART: works in the bass region; cancellation limited in frequency and level
- Late copies from side walls, ceiling, glass
- Comb filtering; smeared soundstage
- A filter cannot stop the wall sending the copy
- Treatment: first reflection points
- low frequencies
- high frequencies
In 2011 Nyal Mellor of Acoustic Frontiers showed two filters, at 25 Hz and 50 Hz, cutting both the peaks and the ringing in a nine-seat home theatre. His conclusion: "Equalization should be thought of as the last component of the optimization process." The last component. Not the first.
Where room correction software earns its place
Correction is not a gimmick. In the October 2025 film I gave the example every room owner should fear: a 10 dB peak or null around 60 Hz. Own the most accurate speakers in the world, and that is still the response you hear at your seat. A peak like that is where correction is strongest, for the minimum-phase reason above. A null is harder; Dirac's answer is Bass Control, which lets several subwoofers cooperate and use phase at a specific frequency to cancel it.
It also earns its place in rooms you cannot touch. Most of us listen in a living room, with all its restrictions; my old room had a full glass wall beside one speaker, and you cannot hang an absorber on a view. And it suits listeners who value balance above all: on my system the MEN220 improved imaging, soundstage, levels and timing. Those gains are real.
Dirac Live: strong in the bass, no substitute for panels
Dirac Live is worth it for bass problems in a room you cannot treat, especially with several subwoofers; as a substitute for first-reflection panels, it is not. Dirac's own product page sells it as correction of both magnitude and timing for the whole system in the room, yet its two extensions, Bass Control and Active Room Treatment, both work in the low end.
One owner's account shows both sides of the bargain. A reader who wrote to SoundStage! Access in April 2016 said Dirac Live let him "hear and feel the really deep bass below 30Hz clearly for the first time," with "no boominess whatsoever." Yet at first, he admitted, "I felt that I missed something": "the exaggerated bass and the time smearing of bass transients" he had taken for power. It took him about two weeks, and a target curve lifted to +5 dB below 100 Hz, to settle. Correction can take away a flaw you had learned to love. It cannot take away a reflection.
What it cost: more correct, less alive
The test proved the room mattered. Living with the processor taught me the second, harder lesson, because my chain carries tubes and vinyl. My McIntosh C12000 preamplifier pairs tubes with solid state; the turntable is a McIntosh MT5 with a Sumiko Starling moving-coil cartridge. In my February 2026 film on the new listening room I said it plainly: the MEN220 is "a digital processing system. It adds things to the sound." Put it in a pure analogue chain and it converts everything to digital. You lose the airiness of tubes, the long tails, the shimmer on metal instruments, the dynamics.
Better balanced, perhaps more correct, but less alive.
In October 2025 I had called the same device the real proof of everything I said about rooms. Both statements are mine. The October verdict came from a measurement; the February one came from records and tubes played through the box. At The Sound Lab, better means more convincing, closer to live sound. A processor can flatten the frequency response and make the music less believable at the same time. Correct is a measurement. Alive is what makes you reach for the next record.
That is why the walls matter twice. Every problem they solve is one no filter has to chase, and in an analogue chain, the less there is left to correct, the easier it is to ask whether the processor belongs in the signal path at all. The air lives outside the conversion. The Daily More correct, less alive has the short version.
The worst room possible: 37% in the old L-shaped space
The February 2026 film describes the old room without mercy: about 800 square feet, L-shaped, the whole system cramped into the small leg. One shared electrical circuit. One possible speaker position. The seat 3.5 metres from the speakers, with zero gap to the back wall. The left speaker exposed to a massive open area, the right pushed against a full glass wall. I told people in the industry many times: "I am cursed with the worst room possible."
In that film I put the MEN220's figure for that room at 37%, the highest correction ratio we have ever recorded in any room. I added panels, diffusion and absorption, and the correction dropped to the mid-teens. But the problems remained.
Two films, two pairs of numbers, both from before the new listening room: 25 and 7 in the October 2025 film, 37 and the mid-teens in the February 2026 film. The films do not reconcile them, so I keep each with its own film and never average them. They agree on the part that matters: in both, treatment took away more than half of the correction the processor wanted.
Read every label as live text
- How much correction the MEN220 wanted, before and after treatment
- Correction reported by the McIntosh MEN220 (%)
- 0
- 10
- 20
- 30
- 40
- 25%
- Untreated
- 7%
- Treated
- 37%
- No panels
- mid-teens
- With panels
- Separate films, not averaged
- October 2025 film
- February 2026 film
What finally fixed the sound was not a processor but the new listening room: a rectangle of about 1,000 square feet, with mild absorption under natural-wood panels behind the system, corner bass traps, custom ceiling absorption, layered curtains and two dedicated 40-amp circuits, one per MC611 monoblock.
The helicopter on Pink Floyd's The Wall came from high on the ceiling for the first time in my system. On "Mining for Gold", from the Cowboy Junkies' The Trinity Session, recorded in the Church of the Holy Trinity in Toronto, you are in the church. Nothing in my signal chain changed: same Yvettes, same C12000, same monoblocks, same Transparent cables, same sources. Room and power changed together and share the credit. But the old room had been holding the system hostage.
The right order: position, treat, measure, then correct what remains
Better speakers make the order more urgent. When I moved from Klipsch Reference Premiere to Bowers & Wilkins 800 Series D4, my first thought was that my dealer had cheated me. He had not; the more revealing B&Ws made the room's problems bigger. The order I would follow:
- Position the speakers. It costs nothing: toe-in, toe-out, distance from the back and side walls, in very small increments. A properly positioned speaker is the difference between a $5,000 pair and a $15,000 pair.
- Treat the first reflection points, with a maximum of $500 at first.
- Trap the corners, on a similar budget. I strongly suspect your low end will come out tighter and cleaner than any subwoofer integration someone recommends.
- Measure before you spend on gear, cables or more panels. A microphone and free software such as REW will do; you do not need a box like mine. I made that case in Measure first, and do it for free.
- Correct only what remains. If a bass peak survives positioning and traps, that is the moment for room correction, and it will have far less to do. In an analogue chain, weigh that peak against the conversion before you insert the box.
Beware the trap I fell into more than most: over-treatment. Go panel by panel: install, measure, reapply. Overdo it and your room becomes lifeless, the acoustic twin of more correct, less alive. Unlike any box, treatment does not depreciate; it stays through every change of electronics and speakers.
| Question | DSP room correction | Acoustic treatment |
|---|---|---|
| What does it act on? | The signal, through filters built from microphone measurements | The room: walls, ceiling and corners |
| Where is it strongest? | Bass peaks and balance at the measured seats | First reflection points and corner bass |
| Where does it fall short? | Reflections above the bass; even Dirac's ART works in the bass region | Overdo it and the room becomes lifeless |
| What did it cost my analogue chain? | Conversion to digital: less air, shorter tails, less shimmer, less dynamics | Nothing added to the signal |
What each approach acts on, where it is strongest, where it falls short, and what it cost an analogue chain.
Sources: my October 2025 and February 2026 films; Toole (1999) via Acoustic Frontiers; Dirac's Nilo Casimiro to StereoNET (April 2023).
Who should buy DSP first: the honest exceptions
Three kinds of listener should buy the box before the panels.
- You cannot touch the room. Position the speakers, then pull the one lever you have.
- Your problem is bass and you run more than one subwoofer. That is exactly the job Dirac's Bass Control does.
- Your music is digital from end to end. My objection was what the conversion did to an analogue chain; with no turntable or tubes in the chain, that particular price is not yours to pay.
Even then, measure first, and know that the reflections stay where they are. The industry will gladly sell you the box because, as I said in October 2025, they cannot sell you a room. In my test the walls left the box 7% to do. Start with the walls.
Frequently asked questions
Is Dirac Live worth it?
Dirac Live is worth it for bass problems in a room you cannot treat, especially with several subwoofers. As a substitute for panels at the first reflection points, it is not.
Does room correction software replace acoustic treatment?
No, room correction software does not replace acoustic treatment. Correction works best on bass peaks at the measured seats; reflections above the bass are the room's job. In my October 2025 film, treatment cut the correction my MEN220 needed from 25% to 7%.
What room correction does the McIntosh MEN220 use?
The McIntosh MEN220 uses Lyngdorf's RoomPerfect, built under license from Lyngdorf Audio A/S. The measurement microphone and stand come in the box.
More from the Journal
Keep exploring.
Loading articles…