ByteLite - Dictionary-cascade compression

Not Incremental. Architectural.

ByteLite is a deterministic lossless compression system built on Bit Motion Encoding over the Sigma-9 alphabet. Every byte normalizes bitwise. Every round narrows toward a terminal payload. The model travels with the data. Controlled test on structured text corpora: payload convergence to under 64 bytes. Full S-formula (payload plus decompressor) under independent verification. Results vary by data type.

Patent pending  |  Lossless, SHA-256 verified

Try It

Try ByteLite Compression

Upload any file to see the compression in action

📁

Drop a file here or click to upload

Any file type, up to 2GB

Three Stages in the Compression Pipeline

01

Normalization

  • Source stream enters the engine
  • Bitwise normalization into canonical form
  • Source bit 0 becomes a single normalized 0. Source bit 1 expands to the eight-bit sequence 00000001
  • Stream is now deterministic across any input
02

Motion Encoding

  • Bit Motion Encoding operates over the Sigma-9 alphabet
  • Each motion value is a REF-relative successor choice
  • 510-sequence motion dictionary governs choices
  • Mode Ladder A through F governs which representation is active
03

Convergence

  • Sequential rounds reduce structural redundancy
  • Payload reaches terminal form
  • Output: compressed payload plus round-count in 255-carry format
  • Decompressor replays motions in reverse to reconstruct bit-for-bit

Round / Stage Reduction

Stage 1
Normalization
input
bitwise normalization
Stage 2
Motion Encoding
reduced
REF-relative motion over Sigma-9
Stage 3
Mode Convergence
further reduced
Mode Ladder A-F active
Stage 4
Output Key
compact key
controlled test: under 64 bytes on structured text

Educational pipeline visualization. Relative proportions are illustrative. Controlled test on structured text corpora: payload convergence to under 64 bytes. Full S-formula (payload plus decompressor) under independent verification. Results vary by data type. Patent pending.

How Results Are Measured

Compression ratios are measured against the Hutter Prize S-formula, which sums the compressed payload size and the decompressor binary size. Dictionary tables, when required, count in S. The S-formula is the standard for the Hutter Prize competition on the enwik9 corpus.

Random binary data outside the Sigma-9 alphabet domain receives no advantage from BME. Results reported on this site reflect structured text corpora. Performance on other data types varies.

What Is Real Today

Live Now
  • Universal dictionaries implemented
  • Lossless compression engine
  • SHA-256 output verification
  • Patent pending
Interactive Preview
  • Browser-based demo at /demo
  • Educational stage visualizer
  • Sample determinism proof
In Development
  • Demonstration application
  • Public verification harness
  • Performance benchmarking suite
Planned
  • Public API access
  • Enterprise licensing portal
  • Open-source verifier

Integration Interface

API access is in planning. Enterprise partnerships available now.

POST /api/v1/compress
Content-Type: application/json
 
{
  "data": "<base64-encoded-input>",
  "dictionary_profile": "universal-v1",
  "mode": "structured"
}
 
// Response
{
  "compressed_key": "a4f8c2b7...",
  "original_bytes": 1000000,
  "compressed_bytes": 15,
  "verification_hash": "sha256:a4f8c2b7..."
}
HeartStrings Play is launching. Follow the ByteLite LLC ecosystem.