Most of the modern compression algorithms tend to be «visually» lossless — meaning that they are lossy — but the reduction in quality is only noticed on close inspection. However, at Dotphoton we believe this is not enough for professional photographers. Dotphoton algorithms create no artefacts and retain image information, so they are as suitable for post-processing and archival as regular raw files.
Technically, Dotphoton introduces a small scientifically controlled loss of 0.3 bits of information per pixel at most. Bits are a logarithmic scale, and 0.3 is a very small amount. In fact, this value is so small that the difference between the original pixel value and the compressed value is always strictly smaller than the noise. But the usual lossy vs lossless dualism does not tell the whole story for our solution.
For example, if you take two identical pictures, one after the other, they will not be bit-for-bit an exact replica of each other and the noise structure will be different, but they are still the same picture. Similarly, Rawsie guarantees that the difference between the original and compressed photo is smaller than the natural difference between two "identical" photos.
By stating "preserving full raw quality and resolution" we mean that we preserve details, dynamic range and fidelity of a raw file, i.e. all raw parameters you normally want to tweak in your raw images in a photo editing software are still available and fully supported in Rawsie-compressed images and we don't change bit count of your image.