This view renders the raw FES 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.

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

Even the smallest change will produce a completely different pattern.

Compare results to see how each setting reshapes the underlying 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 FES 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.

FES 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 FES 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. FES is neither — it is a deterministic, reproducible transformation engine. Where FES streams perform well against randomness measures, this represents an overlap in properties, not a 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 FES 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 FES purposes.

A further distinction must be kept in view. FES streams match payload length both before and after transformation. FES 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 FES 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. FES Gauge offers direct measurement of both — a level of transparency and testability that block cipher architectures structurally cannot provide.

Fractal OTP is optional.

The Fractal OTP allows you to enter meta-data to generate unique Fractal Streams that meet the Shannon One Time Pad (OTP) requirement that a key can only be used once.

When supplied, FES uses Fractal OTP to navigate from the Key Portal to a second OTP Portal, using the string entered.

The Fractal OTP is also Silo sensitive, so the FES Fractal Stream (equates to Shannon's OTP key) is unique to the Key/OTP/Silo combination (and the other options).

For file and document encryption you can enter the full File path of the document, for communications you can use Session IDs or time-stamps.

The FES 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 Key Space selector sets the total number of bits used for the key.

Required Key Space precision is achieved by expanding the number of Mandelbrot dimensions, each dimension adding 112 bits of precision. There is no conceptual limit to the size of the Key Space; it has been tested to 4,480,000 bits (limited to 57,344 bits here).

Transformation Precision is the same as the Key Space selected.

The Transformation Precision is achieved by scaling the number of Mandelbrot Dimensions. The possible number of Mandelbrot Dimensions is logically unlimited, limits are imposed by hardware memory and processors. Up to 40,000 dimensions have been tested, but in practical terms 512 dimensions is a likely maximum with a 57,344 bit key space.

Key Space identifies a multi-dimensional fractal portal. The Key is then discarded and does not act on the payload.

The payload is then transformed by an infinitely complex multi-dimensional fractal stream that emerges from the portal, using the same Fractal Precision.

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. fBlit: Bitwise fractal stream driven scramble.
2. sub: Fractal driven byte substitution.
3. xor: The default bit-wise cryptographic operator. Cannot combine with add as this cancels the stream.
4. add: The payload and fractal stream byte values are added. Cannot combine with xor.
5. bit split: Bytes are bit-split according to fractal stream values.
6. bit warp: Bit segments are sequentially selected and split according to fractal stream.

The green functions are the core overwrite transforms and the yellow functions are complimentary transforms. You should always use at least one core transform.

Each selected Overwrite Function shifts, swaps, and overwrites in byte and bit aligned modes, submerging the payload in the fractal stream, the payload ceases to exist.

Bit-split delivers additional diffusion and non-linearity. It should be used in combination with another function.

Scramble Fractal re-ordering using fractal value order, a powerful transform that re-orders the entire payload byte positions.

The order is assembled from core fractal values during Fractal Stream generation.

Scramble is unique for each pass.