← Controls/Stage 08 — Pay TV Scrambling

Pay TV Controls

Pay TV scrambling operated entirely on the composite video signal before transmission. These were not encryption systems in the cryptographic sense — they modified sync pulses, picture polarity, or scan-line order in ways that were reversible only with the correct descrambler hardware. All four major techniques used from the 1970s to the 1990s are simulated here from first principles, operating on the same composite waveform used by the rest of the signal chain. Signal Domain Mode is required and is enabled automatically when any scramble mode is selected.

Controls

SCRAMBLE MODE None / Sync Suppress / Video Inversion / Cut-and-Rotate / Line Shuffle

Selects the conditional-access scrambling technique applied to the composite waveform. Each mode is modelled at signal level — sync pulses, colour burst, and active video are modified exactly as the original transmitter hardware did. Selecting any mode other than None automatically enables Signal Domain Mode.

None — No scrambling. Signal Domain Mode is not affected.

Sync Suppression (Oak Orion / Jerrold / early cable) — The horizontal sync tips are raised from −40 IRE to blanking level (0 IRE), destroying the reference the receiver uses to lock horizontal and vertical scan. Simultaneously, the colour burst on the back porch is replaced each line with a random-phase carrier at the correct frequency and amplitude. This breaks the colour PLL as well as the sync PLL. A receiver without a descrambler loses both sync lock and colour: the picture rolls vertically, tears horizontally, and shifts through a rainbow of hues as the colour PLL hunts without a stable burst reference. The simulation drives the vertical oscillator into free-run, adds per-line H-shear via the active-video circular shift kernel, and replaces burst samples with random-phase sinusoids derived from a per-line hash.

Video Inversion (VideoCipher II — HBO, ESPN, C-band, 1986–1994) — Combines sync suppression (above) with per-line pseudo-random polarity inversion of the active video. An LFSR seeded by the decryption key determines which lines are inverted each field; approximately 50% of lines are inverted at any time. Sync-suppression effects (V-roll, H-shear, colour cycling) are fully active. The additional inversion makes content completely unrecognisable: inverted lines have their luma complemented (white becomes black, mid-grey stays mid-grey) while preserving scan-line sharpness. The result is a rolling field of interleaved normal and inverted lines.

Cut-and-Rotate (Videocrypt / Sky BSkyB / Eurocrypt, 1989–1996) — Each active scan line is circularly rotated by a different pseudo-random number of samples. The sync pulse and colour burst are left completely intact, so the receiver locks normally and the picture is geometrically stable. The cut point changes each line according to a PRNG seeded by a Videocrypt-period key (re-keyed approximately every 2.5 seconds / 5 PAL fields). The visual result is every horizontal line shifted sideways by a different amount; flat-colour regions appear as a staircase of correctly-coloured strips at random horizontal positions, with a noise burst at each seam. The cut-and-rotate kernel operates entirely within the active video region — sync and burst samples are never disturbed.

Line Shuffle (Nagravision / Canal+, 1989–present) — The active scan lines of each field are divided into groups of 32 and the lines within each group are permuted according to a Fisher-Yates shuffle seeded by the current key. The pixel content of each line is completely untouched — sync, burst, colour, and sharpness within every line are preserved. Only the vertical order of lines within each 32-line group is scrambled. The key is refreshed approximately every 5 fields. Without a descrambler, the image looks like a correctly-exposed photograph sliced into 32-line horizontal strips and randomly reassembled: fragments of the image appear at correct horizontal positions but at wrong heights.

Interactions: All modes require Signal Domain Mode (enabled automatically). Sync Suppress and Video Inversion interact with Vertical Hold — increasing V-Hold offset accelerates the free-running roll. Cut-and-Rotate and Line Shuffle work on any standard; Sync Suppress and Video Inversion are most visible on NTSC where sync-loss behaviour is most dramatic. VCR recording of a scrambled signal (enabling both Pay TV and VCR) produces the authentic second-generation degradation seen when subscribers tried to tape premium content through their converter box.

SCRAMBLE STRENGTH 0 – 1.0

Controls the magnitude of the scrambling effect. At 0.0, scrambling is bypassed regardless of mode. At 1.0, the full effect is applied. Intermediate values simulate a receiver that is partially decoding — a weak descrambler, a mismatched standard, or a degraded key signal.

Sync Suppression: At full strength, sync tips are raised to blanking level (0 IRE) and the colour burst is fully randomised per line. As strength decreases, sync tips are only partially raised, the burst randomisation is attenuated, and the V-oscillator approaches lock more easily. At intermediate values the picture rolls slowly rather than continuously.

Video Inversion: Controls the H-shear amplitude and V-roll rate independently of the sync suppression component. At full strength, both components are maximally active. The inversion itself is binary (a line is either inverted or not) so Strength does not affect inversion probability — it affects the sync-loss artefacts that accompany it.

Cut-and-Rotate: Sets the maximum cut-point range as a fraction of the active line width. At 1.0, cut points are distributed uniformly across the full active line (~910 samples for PAL). At 0.5, cut points are limited to the first half of the line, producing a less severe staircase. The seam noise burst at each rotation point is also scaled by Strength.

Line Shuffle: The Fisher-Yates permutation is always a full permutation (all 32 lines within each group are permuted). Strength attenuates the visual impact by blending scrambled and unscrambled line positions. At 1.0, every line is at a fully permuted position; at 0.5, approximately half the lines are at their original positions.

How to test: Set any mode, set Strength to 1.0, then slowly drag Strength toward 0. Observe the partial decoding artefacts that appear at intermediate values — these match the visual behaviour seen on real receivers with a descrambler that had the wrong card, wrong key generation, or a marginal RF input.

Signal-domain requirement

All Pay TV modes require Signal Domain Mode because the scrambling operations are defined at composite waveform level — they modify individual samples of the sync pulse, colour burst, and active video waveform. There is no pixel-domain equivalent that could faithfully model the downstream consequences (sync loss, colour PLL break, seam noise, interaction with VCR colour-under). Selecting any scramble mode enables Signal Domain Mode automatically; it can be manually disabled after selection if you need to compare, but the scrambling effect will be absent.

Historical notes

None of these systems were designed to be unbreakable — they were designed to be cheap to implement in consumer hardware while being difficult enough to defeat with 1980s electronics. Sync suppression was broken almost immediately by hobbyists with a sync-restoration circuit. VideoCipher II was defeated by pirate descramblers that extracted the LFSR seed from the VBI; VC-II RS (Renewable Security) extended the system’s life by moving the seed to a smartcard that could be re-keyed, but each successive card generation was compromised within months of release. Videocrypt was similarly defeated in stages by emulator cards. Nagravision’s cryptographic layer was stronger, but the physical scrambling is otherwise the same. Digital conditional access (DVB-CA, Mediaguard, Irdeto) rendered all of these systems obsolete in the late 1990s.

Experience the real thing.

Every control in this reference reflects a physical component or circuit. Physics-accurate, from first principles.