Decodificador de audio Morse - Decodificar Morse de gravacoes

An audio Morse code decoder converts recordings of Morse beeps into readable text by analyzing timing patterns between tone and silence. Unlike typing dots and dashes by hand, you can start from a WAV or MP3 captured from a practice session, an on-air recording, or exported video audio.

This tool is built for ham radio operators checking a short CW clip, students verifying homework, and anyone who already has Morse as sound rather than as written symbols. It follows International Morse Code (ITU) timing rules and runs entirely in your browser for the audio analysis step.

After decoding, you can cross-check results in the main translator or use the picture translator when your source is a photo or screenshot instead of audio. translator, picture translator.

When you upload a file, the decoder uses the browser Web Audio API to decode the file into a PCM waveform — no audio is uploaded to our servers for this step. The waveform is scanned in short frames (about 10 ms each). For every frame, the tool calculates RMS energy (a measure of loudness) to build an envelope of how strong the tone is over time.

That envelope is lightly smoothed, then compared to an adaptive threshold (roughly 30% of the peak level). Frames above the threshold count as “tone on”; frames below count as silence. The decoder merges consecutive frames into segments and measures how long each tone and each gap lasts in milliseconds.

Next, it estimates the dot length by looking at the shortest common tone durations in the recording. Durations near one dot unit become dots (.); durations near three dot units become dashes (-). Gaps are classified using standard ITU ratios: a short gap keeps building the same letter, a medium gap starts a new letter, and a long gap inserts a word break.

Before showing results, the tool checks whether the timing pattern actually looks like Morse (rather than speech or random noise). If timing is too irregular, it asks for a cleaner recording. The Morse string is then mapped to letters using the same International Morse table as the main translator. An optional AI organize step sends the decoded text to Google Gemini on the server to fix spacing and casing — only the text and Morse string are sent, not the original audio file.