Analogue television colour systems and regional standards

Read the layers before the names

An analogue colour picture is easier to follow if six terms are fixed first:

• In the 2:1 interlaced systems discussed here, a field is one time-separated scan of alternating line positions; the next field scans the positions between them. Together the two fields form one raster frame. The raster is that complete scanning schedule: 59.94 fields/s is about 29.97 frames/s, while 50 fields/s is 25 frames/s.
• Composite video places brightness, synchronizing information, and encoded colour in one electrical waveform.
• Luma Y′ is the brightness-bearing signal formed from nonlinear, gamma-corrected R′, G′, and B′. The prime marks these nonlinear signal values; unprimed RGB denotes linear-light values.
• A colour difference says how far a channel lies from luma, for example R′−Y′ or B′−Y′. Two independent differences plus Y′ are enough to reconstruct the three colour channels.
• A subcarrier is the high-frequency wave used to carry those differences. NTSC and PAL send a short reference burst on the back porch, the blank interval after the line-sync pulse. SECAM instead uses frequency modulation and identifies which difference is present on each line.
• QAM means quadrature amplitude modulation: two values control two perpendicular components of one subcarrier. A QAM vector draws those two values as one arrow. Its angle represents hue and its length represents chroma magnitude; it is a diagram of the signal coordinates, not a separate transmitted object.

The colour-system names are historical abbreviations:

• NTSC — National Television System Committee, the committee that standardized the US system.
• PAL — Phase Alternating Line.
• SECAM — from the French Séquentiel couleur à mémoire, “sequential colour with memory”.

All three compatible colour systems retain Y′ for monochrome reproduction; they differ mainly in the encode–carry–decode stages. Keep this chain in view:

The two fields arrive at different times rather than forming one instantaneous snapshot; motion can therefore expose their interlace. The diagram is conceptual: its eight lines show the interleaving pattern, not the half-line timing that establishes 2:1 interlace.

A standard is a stack, not a nickname

“PAL video”, “an NTSC country”, and “a SECAM television” are useful everyday phrases, but each compresses several independent technical choices into one label. A complete terrestrial analogue service needs at least three layers:

1. Raster and timing: line count, field rate, interlace, blanking, and sync.
2. Colour encoding: how two colour differences share one composite signal with the monochrome picture.
3. RF transmission: channel width, vision-modulation polarity, vestigial sideband, sound-carrier offset, and sound modulation.

The diagram shows one documented combination: a 625/50 raster, PAL colour, and System N RF transmission form the Argentine PAL-N profile. The layers can also vary independently. The following are comparisons, not conversion paths:

• Same raster and RF, different colour encoding: NTSC-M and PAL-M both use the 525/59.94 System-M environment, but encode colour differently.
• Same colour encoding, different RF system: PAL-B/G and PAL-I both use PAL colour, but their RF plans differ in vision bandwidth and sound spacing; B/G channel width also depends on the VHF or UHF band.
• Same colour encoding, different RF polarity and sound: SECAM-D/K uses negative vision modulation and FM sound, while SECAM-L uses positive vision modulation and AM sound.

NTSC, PAL, and SECAM principally name the second layer. Letters such as M, N, B/G, H, I, D/K, K1, L, and L′ name standardized combinations of raster and radiated-signal parameters. Historical practice joined these layers in familiar regional profiles, but the joints remained real: Brazil used PAL colour with System M, Argentina used PAL with System N, and SECAM appeared with B/G, D/K, K1, L, and the local L′ carrier arrangement.

The monochrome contract underneath colour

Compatible colour systems preserved the existing brightness picture. Encoded red, green, and blue first form nonlinear luma Y′; two remaining coordinates describe colour differences. A monochrome receiver does not decode that added structure as colour. It reproduces mainly Y′; receiver bandwidth and visual integration usually reduce the high-frequency chroma pattern, although residual dot or cross-luminance structure can remain visible.

Read the expression from left to right:

• Y′ is the nonlinear luma produced by the sum. It is the brightness-bearing output used by a compatible monochrome receiver.
• R′, G′, and B′ are the nonlinear red, green, and blue input signals. The prime says that their transfer characteristic has already been applied; these are not linear-light quantities.
• 0.299, 0.587, and 0.114 are the three contribution weights. They add to one, so equal white inputs still produce white luma. The larger green coefficient means that green contributes more to Y′ than red or blue for the same signal value.

These coefficients and the assumed display conditions belong to the named standard, not to light in general. The separate Colour models and spaces Principle explains why Y′, CIE Y, linear RGB, and modern digital Y′CbCr must not be substituted for one another.

Sync, blanking, and the colour reference occupy known parts of each line. The companion analogue television signal-path Principle follows that complete line through composite video, RF modulation, reception, and picture reconstruction.

Follow one red patch through the chain

For an illustrative full-red input, R′=1, G′=0, and B′=0. The shared luma equation gives Y′=0.299, so the raw differences are R′−Y′=0.701 and B′−Y′=−0.299. NTSC, PAL, and SECAM scale and carry those two values in different ways, but a correct decoder ultimately recovers the same pair. Adding the red difference back to luma gives R′=0.299+0.701=1; adding the blue difference back gives B′=0.299−0.299=0. The green channel then follows from the luma matrix using those recovered red and blue values: G′=(Y′−0.299R′−0.114B′)/0.587=0. The monochrome picture remains in Y′, while the information needed to restore colour rides beside it.

The next diagram is a map of the three colour mechanisms. Its labels introduce the axes and references that the following sections then derive in sequence.

NTSC: colour as a referenced QAM vector

NTSC places two colour coordinates on quadrature components of a suppressed subcarrier. At any instant the chroma contribution can be written as two orthogonal terms:

In this expression:

• C_NTSC(t) is the combined chroma contribution at time t, not the complete composite-video line.
• I′(t) and Q′(t) are the two time-varying, scaled colour-difference coordinates defined by the NTSC convention. They set the amplitudes of the two terms.
• sin(…) and cos(…) are two components of the same suppressed subcarrier. Their quarter-cycle, or 90°, separation makes the axes independent.
• fsc is the chroma-subcarrier frequency in hertz, so 2π fsc t is its phase advance in radians at time t, measured in seconds.
• 33° rotates the I′ and Q′ axes into the NTSC convention. It is a fixed convention term, not a variable standing for the literal burst phase.

For the red patch above: I′ = 0.74(R′−Y′) − 0.27(B′−Y′); Q′ = 0.48(R′−Y′) + 0.41(B′−Y′). With R′−Y′=0.701 and B′−Y′=−0.299, this gives I′ ≈ 0.599; Q′ ≈ 0.214.

This equation is a baseband modulation model; the standard also specifies scaling and bandwidth.

Vector magnitude represents chroma magnitude; vector angle represents hue relative to the decoder’s axes. A short burst on the back porch of each active line lets the receiver rebuild the missing subcarrier. Its phase provides the angular reference, while its amplitude provides a stable reference for chroma gain circuits.

This makes several effects physically distinct:

• incorrect burst amplitude changes the decoder’s chroma gain reference;
• incorrect burst phase rotates the recovered colour axes;
• chroma amplitude changed without the same change in burst produces a different gain error;
• subcarrier-frequency error makes the reference relationship drift rather than remain at one angular offset.

M/NTSC uses a nominal chroma subcarrier of 3.579545 MHz. The 525-line colour timing is slightly below the earlier nominal 60 fields per second, giving about 59.94 fields/s. Neither fact means every System-M service must use identical black setup: Japanese and US System-M operation is a useful counterexample.

PAL: alternate one colour axis

PAL retains quadrature amplitude modulation but reverses the V-related axis on successive lines. In a simplified notation:

In this expression:

• C_PAL,n(t) is the chroma contribution on line n at time t.
• U′(t) and V′(t) are the two scaled colour-difference coordinates. U′ keeps the same sign from one line to the next.
• n is the integer line number. The numbering origin in this simplified notation is chosen so that the illustrated line n is even and carries +V′; the next line carries −V′. Because (-1)^n alternates between +1 and −1, it reverses the sign of V′ on every other line.
• sin(2π fsc t) and cos(2π fsc t) are the two perpendicular QAM components of the same subcarrier, not two separately transmitted carriers.

For the red patch above: U′ = 0.493(B′−Y′); V′ = 0.877(R′−Y′). With R′−Y′=0.701 and B′−Y′=−0.299, this gives U′ ≈ −0.147; V′ ≈ 0.615.

The burst phase swings with the line sequence so that the receiver knows which V sign applies. A delay-line decoder performs the following two-line correction:

1. line n arrives with +V′ and is stored for one line period;
2. line n+1 arrives with −V′; the decoder restores its sign and combines it with the delayed neighbour, so a fixed phase error appears with opposite signs and largely averages out.

The price is a vertical assumption: adjacent lines are treated as having closely related colour.

PAL therefore did not abolish phase errors. Excess error, incorrect line-alternation, a bad delay path, or rapidly changing vertical colour can leave desaturation, alternating-line colour, or Hanover-bar structure. Burst amplitude and burst phase remain separate references just as they are in NTSC, but the decoder’s response is PAL-specific.

Most 625-line PAL profiles use a 4.43361875 MHz subcarrier. PAL-M and PAL-N use different relationships because their raster and channel constraints are different. This is why “PAL equals 625/50” is a common but false shortcut.

SECAM: frequency and one line at a time

SECAM does not put two simultaneous colour axes on a QAM vector. It alternates scaled red- and blue-difference signals, D′R and D′B, from line to line and frequency-modulates the selected component:

In this expression:

• i selects the red-difference or blue-difference line, so i ∈ {R, B}.
• fi(t) is the instantaneous subcarrier frequency in hertz on the selected line. It is the result on the left side of the equation.
• f0,i is that line type’s resting frequency: the frequency produced when its colour-difference input is zero.
• D′i*(t) is the selected, scaled colour difference after baseband pre-correction. Its sign moves the subcarrier below or above the resting frequency.
• Δfi is the proportionality coefficient that converts one unit of D′i* into a frequency shift in hertz. The product Δfi D′i*(t) is the actual signed shift at time t.

For the red patch above: D′R = −1.902(R′−Y′); D′B = 1.505(B′−Y′). With R′−Y′=0.701 and B′−Y′=−0.299, this gives D′R ≈ −1.333; D′B ≈ −0.450. SECAM then applies baseband pre-correction to obtain the starred D′* value used by the frequency equation.

Accumulating this instantaneous frequency over time gives carrier cycles; multiply that integral by 2π to obtain phase in radians. The equation is a low-frequency modulation model, not the transmitted waveform itself.

The two line types have nominal resting frequencies of 4.40625 MHz for D′R and 4.250 MHz for D′B. Baseband pre-correction is first applied to the colour-difference signal before FM. A separate frequency-dependent “bell” amplitude characteristic then shapes the chroma subcarrier, with reciprocal processing in the receiver. A receiver limits and discriminates the FM chroma, then combines the current line with the other difference stored in memory:

1. on a D′R line, the decoder demodulates red difference and reads the delayed D′B value from the one-line memory;
2. on the following D′B line, it demodulates blue difference and reads the delayed D′R value. Line identification keeps this sequence aligned.

Line identification can use reference-frequency intervals on the back porch; older equipment could use identification signals in selected field-blanking lines. These are not NTSC/PAL burst phase and amplitude references. Calling a SECAM identification-level fault a “burst amplitude error” would hide the different detector, state, and visible failure.

SECAM is consequently insensitive to QAM phase in the narrow NTSC/PAL sense, but it has its own sensitive quantities: FM deviation, resting frequency, pre-emphasis and bell-filter response, correct D′R/D′B sequencing, line identification, limiter/discriminator behaviour, and delay-line state.

Same colour family, different regional service

The headline RF row defines a system family, while its footnotes can narrow the contract for one colour encoding, administration, or optional transmitter arrangement. These exceptions do not create another colour family: they change which radiated spectrum and carrier tolerances a faithful model must reproduce.

Sideband widths are measured outwards from the vision carrier. A negative mask offset lies below that carrier in frequency; a positive offset lies above it. A point such as ≥20 dB at −1.75 MHz requires at least 20 dB of attenuation at that offset, relative to the stated reference level. The explorer presents these as compliance constraints, not as a continuous filter response. An effect that needs the shape between them must use a documented synthesis model or a measured transmitter/receiver response rather than interpolating the listed points.

The standard letter matters whenever an effect reaches outside composite baseband. System M uses a 6 MHz RF channel and a +4.5 MHz sound offset. PAL-I uses an 8 MHz channel and a +5.9996 MHz sound offset. D/K uses 8 MHz and a +6.5 MHz sound offset. French SECAM-L is especially instructive: its vision carrier uses positive modulation and its sound is AM, while SECAM-D/K uses negative vision modulation and FM sound.

The wider list matters too. PAL-H retains PAL colour but a System-H RF mask. Its 1.25 MHz vestigial sideband is wider than System G’s 0.75 MHz vestige even though both use an 8 MHz UHF channel and 5.5 MHz sound spacing. SECAM-B/G uses 5.5 MHz FM sound, while SECAM-K1 uses 6.5 MHz FM sound and negative vision modulation. French L′ is a local VHF carrier-placement variant: it keeps positive vision modulation and AM sound but inverts the RF spectrum, placing sound below the vision carrier. None of these RF distinctions changes the SECAM colour encoder.

Some footnotes are narrower still. B/SECAM and G/SECAM replace the PAL-related far-side mask point with ≥30 dB at −4.33 MHz, allowing ±0.10 MHz in its offset. France also permitted an optional System-L arrangement with a 0.75 MHz vestigial sideband and its own typical mask points. Some former-OIRT administrations added a regional D/K supplement with passband and outer-spectrum constraints. The explorer labels the latter two as optional or regional profiles so their requirements are not mistaken for universal System-L or D/K behaviour. Sound-carrier offsets are nominal; where the recommendation publishes a tolerance, the explorer shows it separately in kilohertz.

Those differences do not change how a correct red patch becomes D′R inside the SECAM encoder. They do change overmodulation boundaries, envelope-detector behaviour, vision/sound interference, channel filtering, and what a mistuned receiver sees. An effect article must therefore state both the colour family and the relevant RF system whenever its mechanism crosses that boundary.

Local conventions can also refine a shared combination without creating a new colour family. Japanese M/NTSC did not use the US 7.5 IRE black setup. Brazil’s PAL-M joined PAL phase alternation to a 525/59.94 System-M environment. Argentina’s N/PAL combination retained a 625/50 raster inside a 6 MHz channel.

Equipment modes are not countries

Multi-standard televisions, VCRs, laserdisc players, game consoles, and standards converters produced useful hybrids. Their menu labels often look like broadcast standards even when they describe only a baseband output or a tape recording method.

• PAL-60 commonly carries PAL-style colour with 525/59.94-like timing so an NTSC-rate recording can be viewed on suitable PAL equipment.
• NTSC 4.43 uses NTSC-style colour interpretation around a 4.43 MHz subcarrier convention, typically for conversion or playback.
• MESECAM is a consumer tape-recording convention for SECAM sources using a heterodyne, down-converted colour-under path in multi-system tape machines; it is not another terrestrial SECAM transmission standard.

These names are valuable evidence when reproducing a device-specific artifact, but they do not define RF channel width, vision polarity, or a national allocation. The device manual and actual signal measurements outrank folklore about what the label “usually means”.

Earlier systems took different paths

NTSC, PAL, and SECAM were not the only possible family tree. The British 405-line System A and French 819-line System E were monochrome raster/RF systems. They preceded the later 625-line colour combinations and coexisted with them during long migrations; neither was an early flavour of PAL or SECAM. System E’s 11.15 MHz vision/sound spacing could appear with either sign because the carrier orientation depended on the assigned French channel.

The CBS field-sequential colour system went further. It sent red, blue, and green picture fields in sequence at 405 lines and 144 fields per second, with a synchronized colour wheel at capture and display. Two interlaced fields form a 72 Hz raster frame, while six fields form a complete colour sequence at 24 Hz; those are different rates and the explorer labels both. The FCC adopted it in 1950, and limited commercial broadcasts began in 1951. Its raster and field sequence were incompatible with the installed 525-line monochrome receiver base. The later compatible NTSC approach instead preserved the monochrome picture and placed colour information around a subcarrier.

Historical systems belong here when they expose a discarded design choice: compatibility versus replacement, simultaneous versus sequential colour, or more lines versus practical channel width. A catalogue of every national assignment would obscure those mechanisms.

Profile reference

The reference uses signal profiles rather than a country-colouring map. Borders and broadcast policy changed, one administration could operate more than one standard during a transition, and studio interchange did not always match the radiated service. The signal parameters are the durable evidence.

Each profile combines raster timing, colour encoding, and—where applicable—an RF system. The diagrams summarize those layers; the fields below retain the published parameters and state which implementation details remain unmodelled. Choose a profile class, then a named combination.

References

• ITU-R BT.470-6 — Conventional television systems (superseded detailed edition) — Tables 1–4 and their notes
• Federal Register, 17 October 1950 — FCC field-sequential colour television standard
• ITU-R BT.1700 — Characteristics of composite video signals for conventional analogue television systems
• ITU-R BT.1701 — Characteristics of radiated signals of conventional analogue television systems — BT.1701-1 (08/2005), Annex 1, Table 1 and notes
• ITU-R BT.2043 — Analogue television systems used throughout the world
• CCIR Düsseldorf 1990 — Soviet tests for Systems D and K/SECAM — Report 959-2, Table III
• ITU-R BT.624 — Characteristics of television systems
• Rohde & Schwarz — analogue television L/L′ technology and spectral inversion
• Philips L01.1E service manual — SECAM B/G and L/L′ receiver conventions
• Panasonic operating instructions — equipment compatibility table for NTSC 4.43 and PAL-60
• Multi-system VCR technical specifications — separate PAL, SECAM, and MESECAM recording paths
• Panasonic VCR operating instructions — MESECAM equipment mode
• CCIR New Delhi 1970, Volume V Part 2 — historical monochrome television systems — Report 308-2, Table I, items 1–17 and note 10
• BBC Research Report 1963/55 — conversion between 405-, 625-, and 525-line television
• BBC Research Report 1962/14 — comparative tests with 405-, 625-, 819-, and 525-line systems
• FCC Annual Report, fiscal year 1951 — adoption and early operation of field-sequential colour