Audio provenance

Watermark audio with a hidden ID

Embed a 40-bit identifier into an audio file, or inspect a file for an existing SilentCipher watermark.

Model choice is part of the ID record A file must be decoded with the same model used to encode it.

Choose audio

The encoded output is returned as a mono WAV.

Drop an audio file here or WAV, FLAC or AIFF

Exactly 10 hexadecimal characters (5 bytes).

Message ID

0x 0 / 10

Use this as the stable identifier in your own records.

3Select a modelStore this choice with the ID.

44.1 kHzDefault model 16 kHzCompact model

The original, encoded output and metadata are also saved by the API for audit.

Use the original model selection if it is known.

Drop an audio file here or WAV, FLAC or AIFF

2Select the encoding modelA mismatch returns “not detected”, not an error.

44.1 kHzDefault model 16 kHzCompact model

Crop-resistant decoding

Try phase shifts for cropped audio. This is more robust, but slower.