Microphone Spectrum & Noise Analyzer Online: Live FFT & Acoustic Diagnostics

Q: Is my microphone audio recorded, stored, or sent to any server?

Strictly no. This tool operates 100% locally on the client side using the standard W3C Web Audio API. Audio buffers are processed purely in your browser memory (RAM) and are never recorded, cached, or transmitted over the internet.

Q: What is considered a good ambient noise floor for voice recording?

In treated studio environments, noise floors measure below −55 to −60 dBFS. In quiet home settings, −50 to −45 dBFS is very good. If your room reads above −38 dBFS, computer fans, open windows, or room flutter echoes are being picked up.

Q: Why do I see 50 Hz or 60 Hz spikes on the spectrum in a quiet room?

This is mains ground hum caused by unshielded microphone cables, ungrounded power supplies, or cheap USB adapters. Using a balanced XLR connection or engaging an 80 Hz high-pass filter eliminates this issue.

Q: How should I set my microphone preamp gain (Gain Staging)?

Calibrate your audio interface gain so that normal speech sits around −20 to −14 dBFS RMS, while your loudest shouts peak no higher than −6 dBFS. This preserves dynamic headroom and prevents digital clipping.

Q: What is the difference between RTA bars, oscilloscope, and waterfall spectrogram?

The oscilloscope displays the raw waveform in the time domain, the 31-band RTA partitions sound into standard 1/3-octave frequency bands, and the waterfall spectrogram plots a rolling 3D thermal history of frequency decay over time.

Q: How does the voice pitch tuner work?

The tuner employs a normalized autocorrelation algorithm with parabolic interpolation to detect the fundamental vocal fold vibration frequency (f0) and map it to musical notes relative to A4 = 440 Hz with cent detuning accuracy.

🎵 Voice Pitch Tuner (Autocorrelation)

-- Hz (0 cents)

Sing or speak steadily into your microphone to detect your fundamental voice pitch and nearest musical note.

🔊 Audio Frequency & Noise Tone Generator

Synthesize precision reference frequencies and acoustic noise to test speaker response and microphone pickup.

Waveform / Noise Type:

Frequency (20 Hz – 20 kHz):

440 Hz

Volume Level:

20%

📖 Microphone Tuning Guide & Acoustic Physics

🎚️ Gain Staging taisyklė

Adjust your preamp gain so peaks hit between −18 dBFS and −12 dBFS during your loudest speech, maintaining safe dynamic headroom without digital clipping.

⚡ 50 Hz Hum šalinimas

Eliminate 50/60 Hz ground loop spikes by utilizing shielded XLR cables, separating signal lines from AC power adapters, and activating an 80 Hz high-pass filter.

🎙️ Artumo efektas (Proximity Effect)

Directional cardioid capsules boost low-frequency warmth when positioned 5–10 cm from the mouth. Always employ a pop filter to disperse plosive bursts.

🛋️ Patalpos aidas ir rezonansai

Hard parallel surfaces trigger flutter echoes and room modes around 150–400 Hz. Soft furnishings, acoustic curtains, and foam diffusers dramatically clean your sound.

Microphone Spectrum & Noise Analyzer is a browser-based diagnostic suite engineered for creators, streamers, vocalists, and remote professionals who demand pristine acoustic precision. In an era where audio clarity defines communication quality, uncovering hidden room resonance, fan rumble, or 50/60 Hz mains hum before important sessions is essential. While human hearing quickly fatigues, digital signal processing (DSP) algorithms and Fourier analysis provide absolute scientific truth.

Comprehensive Audio Engineering Glossary: What Every Metric Means

This diagnostic station tracks dozens of acoustic, physical, and digital telemetry variables in real time. Here is what every parameter on your screen represents:

31-Band ISO 266 1/3-Octave RTA (Real-Time Analyzer): A standard studio spectrum visualizer dividing human audible frequencies (20 Hz to 20,000 Hz) into 31 standardized 1/3-octave center bands. Human perception of pitch is logarithmic (each ascending octave doubles the frequency). This filter bank instantly pinpoints where your vocal energy or unwanted room resonance is concentrated.
Time-Domain Oscilloscope: Visualizes the actual mechanical vibration waveform across time. Our engine incorporates a Zero-Crossing trigger that locks onto the waveform as it transitions from negative to positive voltage, stabilizing the visual trace on screen without jitter during sustained vocal tones.
Waterfall Spectrogram (Heatmap): A 3D acoustic waterfall flowing downward at 60 FPS. It simultaneously maps time (vertical axis), frequency (horizontal axis), and energy amplitude (color temperature – from deep midnight blue for silent passages to burning red and white for high-energy harmonic peaks).
FFT Continuous Spline Curve: A continuous Bezier interpolation of raw Fast Fourier Transform data plotted along a logarithmic scale. It provides an immediate visual summary of your microphone capsule's frequency response curve.
RMS Audio Level (Root Mean Square dBFS): The effective average power of the audio signal. Psychoacoustically, RMS closely correlates with perceived loudness as decoded by the human brain.
True Peak Level (dBFS): The instantaneous crest of the waveform. Keeping your loud vocal peaks below −6 dBFS guarantees ample headroom without clipping.
Dynamic Range (dB): The mathematical difference between maximum peak amplitude and average RMS energy. A higher dynamic range indicates dynamic, uncompressed vocal delivery.
Room Noise Floor (dBFS): The ambient acoustic level when the room is in complete silence. In commercial soundproof booths, this drops below −55 dBFS. In typical home workspaces, −50 to −45 dBFS is great; anything above −38 dBFS means fans, AC units, or traffic rumble are polluting your signal.
Signal-to-Noise Ratio (SNR): The acoustic clearance between your spoken voice and the ambient room noise floor. An SNR above 30 dB delivers crystal-clear broadcast intelligibility.
50 Hz / 60 Hz Mains Ground Hum: Electromagnetic interference induced into unshielded cabling from alternating current power lines (50 Hz across Europe, 60 Hz in North America). It manifests as an irritating low-frequency buzz.
Fan & Desk Rumble (< 120 Hz): Low-frequency mechanical vibrations transmitted through desk legs, keyboard strikes, and computer fans directly into the microphone stand.
Digital Clipping (0 dBFS Threshold): When incoming analog voltage exceeds the maximum integer value of your ADC (e.g., +32,767 in 16-bit or +8,388,607 in 24-bit audio), the top of the wave is truncated into a harsh square wave, generating abrasive odd harmonics.
Voice Pitch Tuner (f0) and Musical Cents: Detects the fundamental frequency of your vocal folds using autocorrelation. 1 semitone equals 100 musical cents; staying within ±5 cents reflects pitch-perfect vocal stability.
DSP Audio Filters (Echo, Noise, AGC): Built-in browser filters: Acoustic Echo Cancellation, Noise Suppression, and Automatic Gain Control. For raw microphone fidelity testing, keep them disabled.

Understanding Fast Fourier Transforms (FFT) and Hann Windowing

Acoustically, sound travels as pressure waves through air, which the microphone's diaphragm converts into microvolt electrical fluctuations. Your audio interface digitizes this continuous analog stream at 44.1 kHz or 48 kHz. While an oscilloscope displays the composite time wave, Fourier analysis decomposes it into constituent harmonic frequencies:

2048-Point Buffer: At 48 kHz sampling, a 2048-point buffer divides the spectrum into 1024 distinct discrete frequency bins with roughly 23.43 Hz resolution per bin, ensuring fine-grained harmonic separation.
Hann Windowing: Slicing continuous audio into discrete algorithmic blocks creates artificial border discontinuities known as spectral leakage. Applying a cosine-bell Hann window tapers the block edges smoothly to zero, rendering pure harmonic peaks without false sideband artifacts.

The Physics of Proximity Effect, Pop Filters, and Balanced XLR Cables

Optimizing recording quality at home relies on fundamental physical principles:

Cardioid Proximity Effect: Directional gradient capsules naturally amplify low frequencies (100–250 Hz) by up to +6 to +10 dB when positioned 5–10 cm from the mouth, lending rich broadcast warmth.
Pop Filter Aerodynamics: Plosive consonants (P, B, T) generate turbulent air jets moving at several meters per second. Dual-layer mesh pop filters disperse this directional kinetic energy away from the delicate diaphragm.
Balanced XLR Common-Mode Rejection: Ground hum occurs when unshielded cords act as inductive antennas. Balanced XLR cables carry hot and cold inverted signal pairs; the receiving preamp inverts one line and sums them, completely canceling out the shared hum.

❓ Frequently Asked Questions (FAQ)

Is my microphone audio recorded, stored, or sent to any server?

Strictly no. This tool operates 100% locally on the client side using the standard W3C Web Audio API. Audio buffers are processed purely in your browser memory (RAM) and are never recorded, cached, or transmitted over the internet.

What is considered a good ambient noise floor for voice recording?

In treated studio environments, noise floors measure below −55 to −60 dBFS. In quiet home settings, −50 to −45 dBFS is very good. If your room reads above −38 dBFS, computer fans, open windows, or room flutter echoes are being picked up.

Why do I see 50 Hz or 60 Hz spikes on the spectrum in a quiet room?

This is mains ground hum caused by unshielded microphone cables, ungrounded power supplies, or cheap USB adapters. Using a balanced XLR connection or engaging an 80 Hz high-pass filter eliminates this issue.

How should I set my microphone preamp gain (Gain Staging)?

Calibrate your audio interface gain so that normal speech sits around −20 to −14 dBFS RMS, while your loudest shouts peak no higher than −6 dBFS. This preserves dynamic headroom and prevents digital clipping.

What is the difference between RTA bars, oscilloscope, and waterfall spectrogram?

The oscilloscope displays the raw waveform in the time domain, the 31-band RTA partitions sound into standard 1/3-octave frequency bands, and the waterfall spectrogram plots a rolling 3D thermal history of frequency decay over time.

How does the voice pitch tuner work?