📊

Room EQ Advisor

This room EQ advisor plays pink noise, measures your room’s response at the listening seat, and returns parametric EQ cut suggestions (frequency, gain, Q) for the peaks — with a before/after simulation and an honest call on which problems EQ can fix and which need treatment.

ℹ This is an advisor, not a real-time corrector — you apply the filters in your own DAW/DSP (or a unit like miniDSP). It’s an uncalibrated measurement (speaker + room + mic, relative dB) at one seat, and EQ only helps there. Crucially it only cuts peaks and never boosts dips: a dip is usually a cancellation (null) that EQ can’t fill — boosting just wastes headroom. EQ tames modal peaks (roughly below 300 Hz) well; reflections and reverberation need physical treatment. The test tone plays through your speakers; the mic is analyzed live and never recorded.

Microphone is off. Set a comfortable level, click “Start measuring” from your seat, let it settle, then “Suggest EQ”.

Share or embed this tool

Free to use on your own website — WordPress, Wix, or any platform. Paste one line and it works instantly, resizing to fit.


How It Works

The tool plays pink noise through your speakers, measures the combined response at your listening seat, and removes pink’s natural tilt to give a relative response curve — similar in principle to an RTA (real-time analyzer), but averaged and processed for room correction advice rather than instant display. When you press Suggest EQ it finds the peaks — frequencies where the room is too loud — and turns each into a parametric cut: the centre frequency is the peak, the gain is enough to flatten it, and the Q is estimated from how wide the peak is (a narrow modal peak gets a high-Q surgical cut; a broad bump gets a gentle one). It then simulates the result by applying those peaking filters back to the measurement and drawing the corrected curve over the original, so you can see the peaks come down. Apply the listed settings in your DAW, plugin, or hardware DSP and re-measure to confirm. For quantitative work, dedicated measurement software such as REW (Room EQ Wizard) paired with a calibrated USB mic will go much further. For a deeper look at where those peaks actually originate, the room frequency analyzer overlays predicted axial standing-wave frequencies so you can see which modal peaks are dimensional.

The most important thing this tool does is tell you what not to EQ. A dip in the response is almost always an acoustic cancellation — two sound paths arriving out of phase at your seat — and you cannot fill it with EQ: boosting just pushes more energy into a null that swallows it, wasting amplifier and speaker headroom for no benefit. The fix for a dip is physical: move the seat, the speakers, or the subwoofer, or treat the boundary causing it. EQ also only works at one position and only really helps with low-frequency modal peaks (typically in the 30–300 Hz range); reflections, comb filtering, flutter echo, and reverberation in the mids and highs are jobs for absorption and diffusion, not filters. If you want to know how much low-frequency bass trap coverage you need to tame those modes physically, use the bass trap calculator before reaching for EQ. Cut, don’t boost; treat first, then EQ the residue; and trust a measurement over a graph.

When entering the suggested filters into your software or hardware, each band needs three numbers: the centre frequency (Fc in Hz), the gain (a negative dB value — it’s a cut), and the Q factor (or its reciprocal, bandwidth in octaves, depending on what your EQ calls it: bandwidth = 1/Q). Common destinations include parametric EQ plugins in a DAW, the four-band parametric EQ inside a miniDSP 2x4HD, Dirac Live’s manual-override filter slots, or the PEQ section of an AV receiver that supports manual parametric filters. If your hardware only offers shelving or graphic EQ, a parametric plugin inserted before your monitoring chain is the most practical alternative.

Frequently Asked Questions

Why won’t you suggest boosts for dips?
Because dips are usually cancellations (nulls) at your seat, and EQ can’t fill a null — boosting just burns headroom while the cancellation still swallows the energy. Move the seat/speakers/sub or treat the room instead.
What can EQ actually fix?
Mostly low-frequency modal peaks (roughly below 300 Hz) at one listening position. It can’t fix reflections, comb filtering, reverberation, or the fact that the response changes as you move.
How is the Q chosen?
From the measured width of each peak: Q ≈ centre frequency ÷ its −3 dB bandwidth. Narrow resonances get a high-Q (surgical) cut; broad bumps get a low-Q (gentle) one. It’s clamped to a sensible range.
Is the before/after curve what I’ll really get?
It’s an idealised simulation — the suggested filters applied to your measurement. Real results depend on your actual EQ and the room, and the measurement is uncalibrated and single-position, so always re-measure after applying.
Should I treat the room or just EQ it?
Treat first. Bass traps and absorption fix the underlying problems (modes, reflections, decay) that EQ can only mask at one spot. Use EQ to polish the remaining low-frequency peaks after treatment.
What test level should I use for an accurate room EQ measurement?
Use roughly your normal listening level — loud enough that the room's low-frequency modes are properly excited and the microphone signal is well above its noise floor, but not so loud that nearby surfaces rattle or the measurement is disrupted by ambient noise. The 40% default is a sensible starting point; raise it slightly in noisy environments. Keep the level consistent between before- and after-treatment measurements so the curves are directly comparable.
How do I enter the suggested Q and Fc into a miniDSP or DAW EQ plugin?
Each suggested filter has three numbers: Fc (Hz), gain (a negative dB cut), and Q. Set the EQ band type to "Peak" or "Bell", enter the Fc and gain directly, then enter the Q. If your EQ shows bandwidth in octaves instead of Q, convert with bandwidth = 1/Q (e.g. Q 4 = 0.25 octaves). On a miniDSP 2x4HD, all three parametric bands use exactly this interface. Always re-measure after applying to confirm the intended peak has come down.
Why does the corrected simulation look flatter than what I actually hear after applying the EQ?
The simulation is idealised — it shows the mathematical result of applying those filters to this measurement. Real results differ because the measurement is uncalibrated and single-position, the room response changes slightly with temperature and humidity, your EQ implementation may define Q differently, and bass modes are not perfectly minimum-phase (so the cut doesn't fully undo the peak at all time offsets). Always verify with a fresh measurement at the same seat position rather than trusting the simulation alone.