Question 1

How do I use a transfer function measurement to set EQ?

Accepted Answer

The general workflow: measure the transfer function at the primary listening position. Apply 1/6 octave smoothing to see the broad trends without getting distracted by narrow-band room artifacts. Identify broad deviations from your target curve (typically ±3 dB from flat or a chosen house curve). Apply corrective EQ using parametric filters: boost or cut the identified problem frequencies using moderate Q values (0.5-2.0). Re-measure after each adjustment to verify the change. Focus on the 200 Hz to 4 kHz range first, as this is most critical for tonal balance and intelligibility. Avoid chasing narrow dips (these are often room modes that shift with listener position) and never boost more than 6 dB at any frequency. SonaVyx AI diagnostic generates EQ recommendations automatically.

Question 2

Should I EQ to make the transfer function perfectly flat?

Accepted Answer

No. A perfectly flat transfer function at one position does not sound natural and creates problems at other positions. Room modes cause narrow peaks and dips that are position-dependent. Cutting a peak at one seat may create a dip at another. Boosting to fill a dip requires a lot of energy and may cause distortion. Instead, target a smooth, gently contoured response. Most sound engineers prefer a house curve that is flat from 200 Hz to 4 kHz and rolls off gently above 8 kHz (-3 to -6 dB at 16 kHz). Low frequencies below 200 Hz may be slightly elevated (+3 dB). Apply smoothing (1/3 or 1/6 octave) to see the trend, and EQ only the broad features, not the fine structure. SonaVyx displays target curves for comparison.

Question 3

What Q value should I use for corrective EQ filters?

Accepted Answer

Q (Quality factor) determines the bandwidth of a parametric EQ filter. Higher Q means narrower bandwidth. For corrective EQ: broad tonal corrections (overall brightness, warmth) use Q of 0.5-1.0, covering 1-2 octaves. Mid-range corrections (presence, nasal coloration) use Q of 1.0-2.0, covering about half an octave to one octave. Feedback notch filters use very high Q (10-30) to surgically remove a single frequency. Avoid using high-Q filters for tonal correction because they create audible ringing artifacts and only affect a tiny frequency range. A general rule: use the widest Q that addresses the problem without affecting adjacent frequencies. SonaVyx EQ recommendations suggest Q values based on the width of the detected deviation from the target curve.

Question 4

Is it better to cut than boost with EQ?

Accepted Answer

Yes, cutting (subtractive EQ) is generally preferred over boosting for several reasons. Cutting reduces the signal level, moving further from the system distortion and clipping threshold, while boosting increases it. Cutting a resonance is permanent and effective at all positions (the resonance is a physical phenomenon that exists everywhere). Boosting to fill a null only works at the measurement position and wastes amplifier power. The maximum recommended boost for corrective system EQ is 6 dB, while cuts of 12 dB or more are acceptable for removing resonances. Start by identifying the most prominent peaks and cutting them. Then evaluate if the remaining response needs any gentle broad boosts. SonaVyx AI diagnostic prioritizes cuts over boosts in its recommendations.

Question 5

How many measurement positions should I use for EQ decisions?

Accepted Answer

For sound system EQ, measure at least 3-5 positions in the primary listening area. More positions give a better spatial average but reach diminishing returns after about 8 positions. Use these guidelines: one position at the mix position (your primary reference), two positions at front and rear of the main audience area on the center axis, and two positions at the left and right edges of coverage. Compare all traces to identify which features are consistent (these are system characteristics that should be EQed) versus which vary between positions (these are room artifacts that should not be EQed). A feature that appears in all positions at the same frequency is a good EQ candidate. SonaVyx trace overlay with up to 10 stored traces makes spatial comparison straightforward.

Question 6

What smoothing level should I use when EQing from a transfer function?

Accepted Answer

Smoothing reduces the visual complexity of the transfer function by averaging adjacent frequency bins. For EQ decisions: 1/3 octave smoothing is best for initial assessment and identifying the broad tonal balance. It removes position-dependent fine structure and shows what parametric EQ can realistically correct. 1/6 octave shows more detail, useful for identifying specific resonances or comb filtering. 1/12 and 1/24 octave show the raw room response including narrow modes, which generally should not be EQed. Start with 1/3 octave to set your broad EQ, then switch to 1/6 octave to fine-tune. Avoid making EQ decisions based on unsmoothed data because the narrow peaks and dips are position-dependent and not correctable with EQ. SonaVyx provides smoothing from 1/1 to 1/24 octave plus an unsmoothed option.

Question 7

How does room EQ differ from monitor EQ in a studio?

Accepted Answer

Room EQ and monitor EQ have different goals. Room EQ corrects the combined response of the monitor speakers and the room at the listening position. It compensates for room modes, boundary reflections, and speaker-room interactions. Monitor EQ adjusts the speaker response itself, typically done in an anechoic environment or with windowed time domain measurements that exclude room reflections. In a studio control room, the approach is usually: first, optimize speaker placement to minimize room mode excitation. Second, add acoustic treatment to control early reflections and reduce RT60. Third, apply minimal EQ to correct the remaining broad deviations. Modern room correction systems like Sonarworks or Dirac use windowed measurements to separate the speaker response from the room response. SonaVyx transfer function with IR windowing helps separate these components.

Question 8

Can I use the transfer function to identify which speaker is causing a problem?

Accepted Answer

Yes. The standard diagnostic approach is to measure each speaker individually (muting all others) and compare their transfer functions. Differences in magnitude response between identical speakers indicate: driver failure (one speaker lacks high frequencies or low frequencies), processing error (different EQ settings, incorrect crossover), wiring problems (reversed polarity, missing conductor), or acoustic problems (one speaker is closer to a wall, obstructed, or angled incorrectly). Overlay the individual speaker traces and look for discrepancies. A speaker with a failed tweeter will show a steep high-frequency rolloff. A polarity-reversed speaker will show phase opposition at all frequencies. SonaVyx trace overlay with color-coded traces makes comparison between multiple speakers quick and visual.

EQ from Transfer Function: How to Make Measurement-Driven EQ Decisions

Frequently Asked Questions