1. Quality Randomness
JFES Fractal Stream quality is measured directly by the JFES Gauge, with optimal entropy results. The raw stream is available for independent testing using Download Raw Stream.

2. Key Size = Payload Size
The JFES Fractal Stream is generated to exactly match the payload size x passes. Every pass applies unused fractal stream entropy.

3. One-Time Use
Fractal OTP can be switched ON or OFF, demonstrating the one-time component of the JFES technology. When ON, the stream is different and unique for every encryption.

JFES achieves Shannon's Perfect Secrecy: cryptographic impenetrability. The ciphertext contains no information that enables an attacker to distinguish the original plaintext from any other plaintext of the same length. Every possible ciphertext is a valid encryption of every possible plaintext, leaving no cryptographic foothold from which the plaintext can be recovered.

This view renders the raw JFES fractal stream (available with the Show Raw Stream button).

This visualisation is designed to expose weakness, not conceal it. Any structural bias, correlation, or pattern within the stream will become visible through alignment, repetition, or clustering. A clean, fully dispersed image indicates absence of exploitable structure under this mapping.

Each circle represents three bytes mapped directly to RGB colour, using matrix-stepped sampling to preserve spatial alignment across the stream.

The output is fully deterministic—identical inputs always produce the same visual pattern.

This is not random noise. The image is a direct visual expression of your password and transformation settings, revealing both distribution and structural characteristics of the stream.

RGB channels are aligned to the same matrix positions across three sequential layers of the stream. This preserves spatial relationships rather than mixing adjacent bytes, allowing structural or correlated patterns to emerge visibly if present. In the absence of such patterns, the result appears fully dispersed.

Open two browser windows side by side and vary one option at a time.

Uncheck Fractal OTP in both browsers.

If you change the Password text, use the Test Fractal Stream button to update the matrix and stats. Changing Transformation Options will update immediately.

Even the smallest change will produce a completely different pattern.

Compare results to see how each setting reshapes the fractal stream.

Uniform colour distribution indicates balanced byte output.

Total pattern change from small input variation demonstrates strong diffusion.

A lack of visible structure or repetition suggests low bias in the stream.

In the matrix-stepped RGB view, absence of aligned patterns (horizontal, vertical, or diagonal) indicates no spatial or layered correlation within the stream.

Compression resistance is the primary measure of JFES stream quality. Compression algorithms are highly effective at detecting and exploiting structure, repetition, and bias. If a stream can be compressed, it contains exploitable organisation.

Stream Size is the total number of bytes in the current visualisation.

Compressed Size is the size of the stream after ZIP compression.

Compression Ratio compares compressed size to original size.

Resistance measures how well the stream resists compression. Values near 100% indicate no exploitable structure.

JFES does not attempt to generate conventional randomness and does not utilise random functions. The following measures provide a statistical view typically applied to random or pseudo-random byte streams.

Average is the mean byte value (0–255, ideal 127.5).

Entropy reflects distribution uniformity (maximum 8).

Chi-Square compares byte distribution to uniform expectation (≈255).

Serial Correlation measures dependence between adjacent bytes (ideal 0).

These measures describe statistical properties only and do not capture structural or sequential organisation within the stream.

The JFES stream quality is determined by its resistance to exploitable structure and predictability, not by adherence to conventional randomness measures.

Conventional cryptographic stream evaluation borrows heavily from randomness testing frameworks, most notably NIST SP 800-22, which was designed to evaluate pseudo-random and true random number generators. JFES is neither — it is a deterministic, reproducible fractal transformation engine. While JFES streams perform extremely well against randomness measures, this represents an overlap in properties, not a JFES design objective.

Compression resistance represents a fundamentally deeper test. Modern compression algorithms are universal pattern detectors — they do not probe for a single statistical property but search for any exploitable structure whatsoever. A stream that resists compression has survived a general sweep rather than a specific probe. NIST SP 800-22 asks targeted questions. Compression resistance asks "is there anything here at all?"

The strong performance of JFES streams across conventional randomness measures — entropy, chi-square, serial correlation, average byte distribution — indicates meaningful overlap between the two evaluation frameworks. However, compression resistance is the primary and more general quality measure for JFES purposes.

A further distinction must be kept in view. JFES streams match payload length both before and after transformation. JFES Gauge transforms a payload of identical repeated characters — the least possible structure, cryptographically neutral. Because the payload contributes nothing, the output carries only the character of the fractal stream itself — the ciphertext is the stream. This means JFES Gauge simultaneously measures the raw fractal stream quality before transformation and the ciphertext quality after transformation in a single unified test. This collapses the before/after distinction entirely. AES ciphertext randomness, by contrast, is measured after transformation only, on output that has already been processed, and the generative mechanism cannot be observed independently. JFES Gauge offers direct measurement of both — a level of transparency and testability that block cipher architectures structurally cannot provide.

The JFES Silo selector sets one of four example Compartmentalisation Silos.

An organisation can generate an unlimited number of unique Silos, and the ones listed are a sample set to demonstrate encryption level compartmentalization.

Each Silo is in effect an entirely unique encryption algorithm.

The Passes selector sets the number of fractal transform passes applied to the payload.

As the unique and infinitely complex Fractal Stream size is unlimited, it is extended to match the payload size x number of passes, continuing to transform the payload with as many passes as desired (limited to 7 passes here) with infinite and unique fractal complexity.

The first pass transforms structure. The second pass removes residual structure, resulting in a fully distributed (incompressible) stream.

Further passes reinforce this state and increase resistance to residual or emergent structure.

You can select one or more Overwrite Functions that overwrite the payload with the Fractal Stream:

1. sBox: Fractal driven byte substitution.
2. xor: The default bit-wise cryptographic operator. This is always active.