This playground is part of an interactive demonstration for an example super language variant, StegaPhone.
Encode and decode secrets in the mispronunciation of words using the text input modes below.
For additional information, see the super language paper.
[Step 1] choose mode
The encode (or sender) mode is for the encoding of payloads into a cover to create a package.
The decode (or receiver) mode is for the decoding of payloads from a package and cover.
[Step 2] configure variant
Set key parameters to the StegaPhone variant, including the payload alphabet and the chunk size.
Defaults are provided and can optionally be updated. Note a chunk size greater than one will be non-deterministic since any word in the chunk may be mispronounced.
[Step 3] run encode/decode
Input a cover and then either (i) a payload for encoding into the cover or (ii) a package for
decoding the hidden payload. Run encode or decode, respectively.
COVER
PAYLOAD
Characters must be in the payload alphabet:
(chunk size =
)
PACKAGE
PACKAGE
transformed package will be rendered here
In the markup output, each line in the table is one symbol in the string.
Each word that is in purple is a mispronunciation and encodes a 1, the
non-highlighted words otherwise each encode a 0.
PAYLOAD
transformed payload will be rendered here
[Step 4] generate TTS
Generate text-to-speech (TTS) audio from package.
Note rate limiting for the TTS endpoint is enforced. Do not excessively run this function to avoid losing access.