On this picture below you can see that the Atari 2600 graphics are rather colourful, and some games like Pitfall! I think use the many shades of each colour and have rather good graphics for the 2600. I have seen a video of that game on youtube and it must have been NTSC or possibly PAL.
My question is kind of twofold. Why is the SECAM palette so garish and ridiculous, and why are there only 8 colours?
The TIA manages a very large palette on NTSC and PAL systems because it takes advantage of the way that composite colour is encoded: three bits produce luminance, and the colour subcarrier is always exactly the same sinusoid, but four of the other colour bits set its phase.
So for both NTSC and PAL:
To verify this, look at an NTSC colour wheel, and look at which angle each of the Atari's NTSC colours appears. They're just steps around the outside of the colour wheel.
Even PAL suffers a little from the NTSC-first logic; it has 12 hues instead of 15 because the NTSC subcarrier is approximately 12/15ths the frequency of PAL's, and the delay steps are NTSC oriented.
SECAM doesn't work in the same way. Performing a phase shift doesn't actually make any difference — the colour subcarrier is a single channel in frequency modulation, not two channels in quadrature amplitude modulation. Nothing about a SECAM output is determined by phase. The colour-shifts that result from phase errors is exactly what SECAM sought to fix.
So more complicated electronics are required, and the elegant hack of just doing a phase shift isn't available. More logic is required, and Atari did what it was cost effective to do, which is to implement a palette much like the other RGB-oriented machines of the era.
In short: the Atari's disproportionately-good palette on NTSC and PAL systems is because Atari exploited the way QAM composite encodes colour. SECAM does not use QAM. Therefore a much more basic palette, closer to other machines of the era, was implemented.
NTSC, PAL and SÉCAM are all colour-carrying signals based on existing monochrome transmission signals. This means that colour information has to be added to the existing monochrome signal, without disturbing it (or as little as possible). The missing information is represented as red and blue differentials with regard to the luminance.
NTSC and PAL carry all the colour information in every line, for every pixel. This results in a certain amount of artifacting, especially in NTSC (PAL avoids this to a large extent by halving the vertical colour resolution).
SÉCAM however only carries one of the two pieces of information in every line: alternating lines provide red and blue signals, and the missing information is taken from the previous line, using a delay line (hence the name “SÉCAM”: “Séquentiel Couleur À Mémoire”, or “Sequential colour with memory”). This is troublesome for the 2600’s TIA since the latter can change its colour output on every line, and doesn’t have a framebuffer (or even enough memory to store the previous line and re-use the colour information).
The simplest fix for SÉCAM, which was the last market to be addressed by the 2600, was to make the luminance constant, which allows chrominance to be calculated — Atari had to choose between colour and saturation, and chose colour. Afficionados of black-and-white TVs would disagree, but imagine trying to sell a black-and-white gaming system in the early eighties...
The use of seven bits out of eight corresponds to the definition of the colour palette in the TIA (see COLUP0 here): four bits determine the colour (on NTSC, none, gold, orange, red-orange, pink, purple, purple-blue, blue — twice —, light blue, turquoise, green-blue, green, yellow-green, orange-green, light orange), and three bits the luminance.
There is an explanation listed in the Atari Mania FAQ:
Jerry Jessop explains why French Ataris produce fewer colors:
I will tell you why it only has monochrome out, because it's SECAM and a SECAM GTIA was never produced. The PAL GTIA is used in France and the Lum outputs are run into an onboard encoder to produce a "psudo" color depending on the Luminance output, composite only. This is why a SECAM VCS or 800 has nowhere near the same number of colors (16) availible as a PAL or NTSC unit (256).
The FGTIA was never completed as the market size did not warrant the expense. The largest SECAM market is not France but the Soviet Union (former) and in 80-84 sales of these items there were not possible.
The GTIA is the (by today's standards, very simple) video chip used by the Atari 2600, of which there are PAL and NTSC versions. This explanation sounds like they decided it wasn't worth the expense to make a third, SECAM, version, solely for the French and Soviet markets, so they decided instead to use the PAL chip (which, unlike the NTSC chip, would output video that was compatible with SECAM TVs for black and white pictures) and added some external, crude electronics to convert the different possible luminance output values of this chip into different colours (it was much easier to make electronics to look at the luminance values than the chrominance ones, since the latter would effectively involve decoding PAL).
Since the chip (being crude) only looked at the luminance values, and there were only 8 luminance values output by the PAL chip, this meant there could only be 8 colours. Therefore, as Stephen Kitt rightly said in his answer, they decided to sacrifice the ability to display different luminances at all for the SECAM output, and locked the luminance on maximum.
I suspect ignoring the complexities of SECAM and outputting just the colours was reasonably trivial as long as you were willing to put up with artefacts on colour transition, which they presumably were. It'd be nice if someone who actually used a SECAM Atari 2600 could verify that this was the case, and how visible these artefacts were.
Regarding the "only 8 colours" point, these are the standard colours available with RGB controls when the only choices are to turn the red, green and blue channels either on or off. Running through the truth table for the options available:-
