PCM WAV Spectrogram Image Generator
Plot a bounded local WAV as a time-frequency PNG with explicit FFT windows, channel and amplitude scales.
Loading tool…
How to use PCM WAV Spectrogram Image Generator
Explore how audio frequencies change over time using a periodic Hann window and one-sided FFT-bin amplitude. The image records chosen window centers and actual displayed bin frequencies, making its time and frequency resolution explicit.
- Choose a supported local WAV or load the original generated chirp.
- Set the image and analysis options, then run locally.
- Read the explicit scales and report; download the original image.
Good to know
- One ordinary uncompressed PCM16 WAV, mono/stereo, 8–192 kHz, at most 10 MiB, 120 seconds and 2,880,000 frames. RF64, compressed/float/extensible WAV and malformed chunks are refused. No browser audio decoder or microphone access.
- A local module Worker gets 10 seconds for preparation and 2 seconds for computation. Image output is limited to 1400 by 800 pixels; decoded sample processing and FFT work are bounded. No upload, playback, account or saved history.
- The source must contain at least one complete FFT window. There are 32–384 selected time windows, at most 2048 frames/window and a 12,000,000-operation estimate. Sparse windows can skip intervals; duplicated rows or time columns do not add resolution.
- Frequency bounds are snapped inward to actual FFT-bin centers. Log frequency needs a positive lower bound. Color is window-corrected bin-centered sine peak amplitude relative to full-scale PCM; DC/Nyquist are not doubled. Off-bin leakage, stereo mean cancellation and mean removal affect results. This is not PSD, loudness, speech recognition or an audio-quality certificate.
Common questions
Does a bright stripe prove the recording has that exact pitch or loudness?
No. The stripe represents selected finite FFT bins and window centers. Window leakage, frequency resolution, channel choice and sparse time sampling limit interpretation; the reference is sample amplitude, not physical or perceived loudness.
Does it need AudioContext, a microphone or an online service?
No. It parses the bounded PCM samples locally, computes in a Worker and draws a PNG with native canvas. The image generation does not depend on browser audio codec support.
Keep going with related tools
All music & audio →PCM WAV Waveform Image Generator
Create a local waveform SVG or PNG image using min/max sample envelopes and an explicit channel layout.
MathSampled Signal Spectrum Workbench
Inspect a bounded real-sample DFT with explicit window gain, single-sided amplitude, phase threshold and a Parseval energy check.
Music & audioPCM WAV Inspector and Trimmer
Inspect, frame-trim, peak-normalize or swap channels in a bounded local16-bit mono or stereo PCM WAV file.