Lesson 01 of 05
How a hash is generated
Learn how a SHA-256 hash is generated, and what it represents
Try it yourself
What just happened
The file is read as bytes and the SHA-256 algorithm is run over them. The 64 hexadecimal characters are the result.
- Deterministic
- The same bytes give the same resultRun the same bytes through SHA-256 on another conforming implementation and the result is identical.
- Sensitive
- A change to any byte changes the resultEven a tiny edit normally produces a very different-looking fingerprint. Lesson 02 lets you measure that effect.
- One-way
- Not a readable copy of the fileSHA-256 is designed to make recovery of the input from its digest impractical.
In the browser file-submission flow, Quantum Notary can receive the digest without receiving the file body. Optional filename, size, or type reaches the service only when you choose to include it, and it can then appear in the proof report or help with record management in the dashboard. None of that metadata is written into Base transaction calldata — only the hash goes on chain.
Technical version
SHA-256 processes a byte stream in 512-bit blocks and produces 256 output bits, usually printed as 64 hexadecimal characters. It is designed for preimage and collision resistance. Because an unlimited set of inputs maps to a fixed-size output, collisions must exist in principle; a matching digest is strong comparison evidence, not a claim of mathematical identity.
Why it's always 64 characters
One word or a large document, SHA-256 still returns 256 bits. The box below computes the real fingerprint on every keystroke.
Fingerprint
Technical version
SHA-256's internal state and output are fixed at eight 32-bit words. Input length changes how many blocks are processed, not the output length. Fixed length makes the digest practical to store and compare; it does not make the digest unique for every possible input.
What it's good for
A fingerprint by itself is a comparison value, not a date or a verdict. If retained data processed under the same rules reproduces the digest, that strongly supports that the bytes match the input that produced it. It says nothing by itself about who made or owns the file, whether its contents are true, or when it existed. A chain record adds the chronology evidence; lesson 04 shows that part.