The second picture contains no new pixel data

Only the 1152×1152 source JPEG and a destination-to-source address map are included. The browser decodes the source once; every output RGB value is read from that fixed decoded array. The generated fern was a target layout during preparation. Its pixel bytes are not loaded by this widget. A bijective map uses each of the 1,327,104 source references exactly once.

Intermediate full-detail states are also permutations. The transition exchanges read bindings, with reverse playback applying those exchanges in reverse order. Source fingerprint: —. Mapping validation: checking.

How the lattice and the image share one state

The toroidal and unwrapped projections use the same row/column addresses as the image read. A selected destination is resolved through the active read map; its original reference supplies the pixel color, source marker, torus path and destination marker. Both lattice axes wrap modulo the image side.

The geometry keeps the handedness and 3:2.4 major/minor radii of the existing Holoseed datafield view. The displayed native-fiber sample uses 24×24 seats at spacing 48 on the 1152 carrier. Its phase comes from the selected source address. This is a projection and sampling of address geometry, not a display of private controller values.

Algebraic behavior and declared routing policy

Fixed-store binding transport reads the destination through an inverse address map. Registered scale coordinates retain coarse position and local phase. Demonstration paths use wrapped row and column offsets, expanded into adjacent logical steps; their endpoints match the actual read references.

The image-specific map is an explicit public address program. It is not a native FPGA permutation plane or a private routing scheduler, and it does not establish that one native operation transforms arbitrary photographs. Route sampling, camera motion, and reveal timing are presentation choices. Logical step counts are not hardware hop-depth measurements.

Scale coverage and memory scope

Full detail includes every source pixel exactly once and preserves the complete RGB histogram. Larger coverage reads one retained phase per block and paints that value over the block. Displayed frequencies can change, while the source remains complete. A complete chart change alone does not duplicate stored values or recover omitted fine detail.

The explicit image map, source decoding, and display buffers occupy browser memory. This is not a compression result or a zero-memory browser implementation. Private recurrence sequences, generated controller tables, kernel tables, and FPGA route programs are not embedded.

Verified on-board payload views

• 24 permutation planes: a complete 20,736-position scan per plane, retaining each underlying reference once.
• 973,968 view-result records across three payload fixtures, with zero mismatches.
• 124,416 before/after readback words: zero mismatches and zero writes while sealed.
• Payload, kernel and recovery passed together in one bitstream; 36 expected payload negative tests behaved correctly.

Verified scale and fabric achievements

The 144×144 campaign covered 186,624 structural outputs across nine family fields with 36 interleaved 24×24 controller fibers. Multiscale/Universal Scale Kernel evidence included 10,782 cases / 467,280 exact records, 47,880 chart round trips, and 12,672 aligned returns. Composed-control coverage included 64 ordered chart pairs across 24 planes.

These workloads overlap and must not be added into one total. The HD artwork and this logical route visualization were not on-board test fixtures.