What causes acoustic feedback in a sound system?

Feedback occurs when sound from a loudspeaker reaches a microphone, gets amplified, sent back through the speaker, and picked up again. This regenerative loop causes a sustained tone at frequencies where the loop gain exceeds unity (0 dB). Room resonances, microphone polar pattern, and speaker-to-mic distance determine which frequencies feed back first.

How does SonaVyx detect feedback frequencies?

SonaVyx continuously analyzes the FFT spectrum for narrow peaks with Q-factors above 10. Natural program content rarely produces peaks this narrow, so high-Q spectral peaks reliably indicate feedback modes even before they become audibly sustained. The detection threshold and Q sensitivity are configurable.

What is a Q-factor in the context of feedback?

Q-factor measures the sharpness of a spectral peak. Higher Q means a narrower peak. Feedback tones are essentially pure sinusoids with very high Q values (often Q > 50). SonaVyx uses Q > 10 as the detection threshold because real program material almost never produces peaks this narrow outside of specific instruments.

How many dB of notch cut should I apply for feedback?

Start with 3-6 dB of cut at the detected frequency with a Q matching the feedback mode width. SonaVyx suggests specific cut depths based on peak severity. Avoid cutting more than 9 dB at any single frequency, as deep narrow notches can cause audible phase distortion in the surrounding frequency range.

Why does feedback happen at specific frequencies?

Feedback occurs at frequencies where three conditions align: the room has a resonance (reinforcing that frequency), the microphone is sensitive at that frequency and angle, and the speaker is efficient at that frequency. The combined loop gain must exceed 0 dB. Room modes at low frequencies and hard surface reflections at high frequencies create the most common feedback conditions.

Can I prevent feedback without notch filters?

Yes, several approaches reduce feedback risk without EQ. Increase the distance between mic and speaker. Use cardioid or hypercardioid microphones aimed away from speakers. Apply acoustic treatment to reduce room reflections. Position speakers in front of microphones (not above). Each technique increases gain-before-feedback by 3-6 dB.

What is gain-before-feedback (GBF)?

GBF is the maximum system gain achievable before feedback occurs, measured in dB. A higher GBF means you can make the system louder before feedback becomes a problem. Typical GBF targets are 12-18 dB above the needed operating level. SonaVyx measures this by tracking peak levels as you increase gain.

Should I use a graphic EQ or parametric EQ for feedback?

Parametric EQ is superior for feedback control because you can match the filter width (Q) to the feedback mode width. A 1/3-octave graphic EQ applies a fixed Q of approximately 4.3, which is too wide for feedback and removes too much adjacent musical content. Parametric allows surgical Q values of 10-30 for minimal collateral damage.

What causes acoustic feedback in a sound system?

Feedback occurs when sound from a loudspeaker reaches a microphone, gets amplified, sent back through the speaker, and picked up again. This regenerative loop causes a sustained tone at frequencies where the loop gain exceeds unity (0 dB). Room resonances, microphone polar pattern, and speaker-to-mic distance determine which frequencies feed back first.

How does SonaVyx detect feedback frequencies?

SonaVyx continuously analyzes the FFT spectrum for narrow peaks with Q-factors above 10. Natural program content rarely produces peaks this narrow, so high-Q spectral peaks reliably indicate feedback modes even before they become audibly sustained. The detection threshold and Q sensitivity are configurable.

What is a Q-factor in the context of feedback?

Q-factor measures the sharpness of a spectral peak. Higher Q means a narrower peak. Feedback tones are essentially pure sinusoids with very high Q values (often Q > 50). SonaVyx uses Q > 10 as the detection threshold because real program material almost never produces peaks this narrow outside of specific instruments.

How many dB of notch cut should I apply for feedback?

Start with 3-6 dB of cut at the detected frequency with a Q matching the feedback mode width. SonaVyx suggests specific cut depths based on peak severity. Avoid cutting more than 9 dB at any single frequency, as deep narrow notches can cause audible phase distortion in the surrounding frequency range.

Why does feedback happen at specific frequencies?

Feedback occurs at frequencies where three conditions align: the room has a resonance (reinforcing that frequency), the microphone is sensitive at that frequency and angle, and the speaker is efficient at that frequency. The combined loop gain must exceed 0 dB. Room modes at low frequencies and hard surface reflections at high frequencies create the most common feedback conditions.

Can I prevent feedback without notch filters?

Yes, several approaches reduce feedback risk without EQ. Increase the distance between mic and speaker. Use cardioid or hypercardioid microphones aimed away from speakers. Apply acoustic treatment to reduce room reflections. Position speakers in front of microphones (not above). Each technique increases gain-before-feedback by 3-6 dB.

What is gain-before-feedback (GBF)?

GBF is the maximum system gain achievable before feedback occurs, measured in dB. A higher GBF means you can make the system louder before feedback becomes a problem. Typical GBF targets are 12-18 dB above the needed operating level. SonaVyx measures this by tracking peak levels as you increase gain.

Should I use a graphic EQ or parametric EQ for feedback?

Parametric EQ is superior for feedback control because you can match the filter width (Q) to the feedback mode width. A 1/3-octave graphic EQ applies a fixed Q of approximately 4.3, which is too wide for feedback and removes too much adjacent musical content. Parametric allows surgical Q values of 10-30 for minimal collateral damage.

Feedback Frequency Finder

Acoustic feedback occurs when a microphone captures amplified sound from a loudspeaker, creating a regenerative loop at specific resonant frequencies. SonaVyx detects feedback frequencies in real time using spectral peak analysis with Q-factor measurement above 10, provides automatic notch filter frequency and bandwidth suggestions, and tracks multiple simultaneous feedback modes across the 100 Hz to 8 kHz range where feedback most commonly occurs.

Try It Now

Open the feedback finder in your browser — identify problem frequencies in real time.

Open Tool

Technical Specifications

Parameter	Value	Standard
Detection Method	Spectral peak tracking (Q > 10)	Prominence-based severity
Frequency Range	100 Hz - 8 kHz	Primary feedback region
Frequency Resolution	< 3 Hz (16384 FFT at 48 kHz)	Sufficient for parametric EQ
Simultaneous Peaks	Up to 12 tracked	Priority-ordered by severity
Q-Factor Measurement	> 10 triggers detection	-3 dB bandwidth method
Notch Suggestions	Frequency, Q, depth (dB)	Parametric EQ parameters
Processing	Rust WASM real-time	< 10 ms latency
Alert System	Visual + severity grading	Critical / Warning / Monitor

How to Find Feedback Frequencies

Open Microphone

Connect your measurement microphone and grant browser access. Position the mic at the mix position or in the coverage area where feedback is most likely. SonaVyx begins continuous spectral monitoring immediately.

Ring Out the System

Slowly increase the microphone gain or system volume. As you approach gain-before-feedback, SonaVyx highlights frequencies where spectral peaks begin to emerge with Q-factors above 10. These are your incipient feedback modes.

Identify Feedback Frequencies

The display shows each detected feedback frequency with its exact center frequency, Q-factor, and severity level. Critical peaks (sustained, high amplitude) are marked red. Building peaks (growing amplitude) are amber. Potential peaks (narrow but stable) are green.

Apply Notch Filters

SonaVyx generates parametric EQ settings for each detected frequency: center frequency, bandwidth (Q), and suggested cut depth. Apply these to your graphic or parametric equalizer. Start with 3-6 dB cuts at the suggested Q.

Verify Improvement

After applying notch filters, increase gain again. SonaVyx shows whether the previously problematic frequencies are now controlled and identifies any new feedback modes that emerge at the higher gain setting.

Understanding Acoustic Feedback

Acoustic feedback is the most common problem in live sound reinforcement. The Nyquist stability criterion tells us that feedback will occur at any frequency where the open-loop gain exceeds unity (0 dB) and the total phase shift is a multiple of 360 degrees. In practice, room resonances and speaker-microphone coupling create multiple potential feedback frequencies, each requiring individual treatment.

Ring-Out Technique

Ring-out is the standard method for identifying feedback frequencies. Starting at a low gain, the engineer slowly increases the microphone level until the first feedback frequency begins to ring. That frequency is notched with a parametric EQ filter, typically 3-6 dB deep with a Q of 10-30. The gain is then increased further until the next feedback frequency appears. This process repeats 5-8 times, each notch increasing the system gain-before-feedback by approximately 3 dB.

Automatic vs Manual Detection

SonaVyx automates the frequency identification step of the ring-out process. Traditional ring-out requires the engineer to identify each frequency by ear or by reading a spectrum analyzer display under time pressure. SonaVyx detects emerging feedback modes before they become audible by monitoring Q-factor trends, alerting the engineer as soon as a spectral peak narrows beyond the Q > 10 threshold.

Common Feedback Frequency Ranges

Feedback most commonly occurs between 200 Hz and 4 kHz. Below 200 Hz, room modes dominate and cardioid microphones have reduced rear rejection. Between 500 Hz and 2 kHz, speech and vocal fundamental frequencies create the strongest coupling. Above 2 kHz, microphone directivity improves but hard surface reflections can still cause problems. Above 8 kHz, feedback is rare due to air absorption and reduced microphone sensitivity.

Feedback Finder Comparison

Feature	SonaVyx	Smaart v9	REW	OSM
Real-time feedback detection	Yes (WASM)	Visual only	No (offline)	No
Auto notch suggestions	Yes (freq, Q, dB)	No	Manual	No
Q-factor measurement	Yes (> 10 threshold)	Visual	Yes	No
Severity grading	Critical/Warning/Monitor	No	No	No
Browser-based	Yes	No	No	No
Multi-peak tracking	12 simultaneous	Visual	Manual	No
Price	Free	$898	Free	Free

Frequently Asked Questions

Related Tools & Resources

Spectrum Analyzer

Full FFT spectrum for detailed frequency analysis

Feedback Elimination Workflow

Step-by-step guided feedback elimination process

Problem Detector

Comprehensive audio problem detection suite

Feedback (Glossary)

Technical definition and physics of acoustic feedback

Standards References

IEC 60268-16:2020 — Sound system equipment: Objective rating of speech intelligibility (feedback impact on STI)
AES-2id:2023 — AES information document for room impulse response measurements (loop gain analysis)
IEC 61260-1:2014 — Octave-band filters (1/3 octave analysis for feedback identification)