hide is a steganographic tool for hiding data in, and retrieving data from, images.
Warning: hide is currently in the beta stages of development!
hide is a steganographic tool for hiding data in, and retrieving data from, images; this is all it does. It (currently) supports BMP, JPEG, PNG, TIFF and Webp images, though might support other image formats in the future. With the exception of JPEG's, the image with the hidden data will be a 32-bit pre pixel image; 3 × 8-bits per RGB and a further 8-bits for transparency.
A list of things the hide isn't:
- An encryption tool
- A compression tool
If you need either of these things doing (which you really should - don't use just one layer of security but many complimentary layers) then checkout encrypt as it does both of those things.
There are three possible usages of hide:
hide <image> <document> <output>
hide <image> <output>
hide <image>
The first will hide the document in the first image and produce an output image which contains the hidden data. The second will retrieve, and place into the output file, a previously hidden file. And finally, the third will show the images capacity for hiding a file in the given image.
You can also use the script:
./truly-hide [file]
This will first encrypt the file (using encrypt, of course) and then hide it. Using this script there is no metadata leakage about what may, or may not, be hidden in a particular image, as it uses the advanced functionality of encrypt of not creating a recognizable container. Omit [file] to recover a previously encrypted and hidden file (it will be extracted to the current directory).
Different image formats will provide different capabilities when hiding data. More often than not the resultant image will have a greater file size than the original, sometimes by an order of magnitude.
Uncompressed BMP images are (currently) the only image format where the file size does not change. PNG and TIFF images often have larger file sizes with hidden data but this isn't always the case. JPEG's have the smallest capacity of all the currently supported formats because of how the hidden data is stored, and can often increase the file size dramatically. Webp images can suffer from massive file size increases especially if the original was lossy, otherwise can be as capable as PNG and TIFF.
The sample images provide an example to demonstrate what could be considered a worse case scenario. Each image with hidden data is filled to capacity with random data, plain text, or a single byte of hidden data.
Image | File Size | Ratio Format | Original Image | Random | Text | 1 Byte | Random | Text | 1 Byte -------+----------------+------------+------------+------------+--------+-------+-------- BMP | 3,978,218 | 3,978,218 | 3,978,218 | 3,978,218 | 100% | 100% | 100% JPEG | 386,716 | 1,136,084 | 1,136,099 | 1,135,270 | 294% | 294% | 294% PNG | 1,823,156 | 2,248,197 | 2,137,849 | 1,814,704 | 123% | 117% | 99% TIFF | 1,955,268 | 4,154,698 | 3,393,644 | 3,779,120 | 212% | 174% | 193% Webp | 1,478,702 | 2,125,684 | 1,800,902 | 1,473,176 | 144% | 122% | 99%
- Debian/Slackware
- WebP library doesn't compile
- Missing definition for lossless image encoding (Debian)
- Missing entire header for encoding (Slackware)
- WebP library doesn't compile