top | item 23455033

Discret 11, the French TV encryption of the 80s

373 points| atomlib | 5 years ago |fabiensanglard.net

83 comments

order
[+] rwmj|5 years ago|reply
The system sounds similar to VideoCrypt. The French system delays lines. VideoCrypt worked by rotating each display line by a number derived from a PRNG. To decode the line the hardware had to buffer the line into a small piece of fast (for 1980s) RAM, and then using a number derived from the key would send the line to the TV starting at that offset and at the end wrapping around to the beginning.

I know all this because in my first job we did a bit of experimentation to see if it would be possible to reconstruct the encrypted images without the key, and in fact it was quite easy because images are very highly correlated from one line to the next, so you just had to rotate each line through each of the 256 combinations and compare it to the previous line, and pick the one with the least difference. Although our system was far too slow to decode a live feed (this would be early 90s), and had some other problems.

I didn't explain this very well - the Wikipedia page has a better explanation:

https://en.wikipedia.org/wiki/VideoCrypt

[+] mcbridematt|5 years ago|reply
The BBC ran a service called BBC Select in the early 90s with a VideoCrypt derivative, rather than line cut they shuffled blocks of lines around the screen - which worked better over more lossy and interference prone terrestrial transmissions.

The technical paper from BBC R&D is quite interesting: http://downloads.bbc.co.uk/rd/pubs/reports/1995-11.pdf

There are a couple of captures of the scrambled transmission on Youtube - it's definitely far more advanced than the earlier Canal method https://youtu.be/HcCdw97_ESA

[+] PoachedSausage|5 years ago|reply
I see that has a reference to European Scrambling Systems 5 – The Black Book by John McCormac. I actually have a copy of that, it gives a rather entertaining overview of the arms race between the broadcasters and the hackers.
[+] mytailorisrich|5 years ago|reply
In general you don't even need to use the whole line to do the correlation.

I remember writing a small program to decrypt image stills in the late 90s and it was surprisingly easy and straightforward, though obviously doing it in real time may have been another story...

[+] antman|5 years ago|reply
If you created a graph where the lines where the nodes and the difference between the lines where the edges then a traveling salesman algo might recreate it efficiently (or its vertical mirror I guess)
[+] philjohn|5 years ago|reply
Growing up in Belgium in the 80s and 90s I also remember Canal Plus and their encryption scheme - and trying (rather fruitlessly) to watch through a movie regardless - but of course, without the sound, that was a fool's errand.

I always wondered how the decoding worked, and for its time, using the serial number of the equipment as part of this calculation was a good one.

Also brings to mind the stories in the US of DirecTV pirating and how they put a stop to scores of pirates on Black Sunday - https://blog.codinghorror.com/revisiting-the-black-sunday-ha...

[+] anthk|5 years ago|reply
> to watch through a movie regardless

Ahem, yes, "movies".

[+] bulgr0z|5 years ago|reply
This brings back so many memories, in the late nineties my dad brought a PCI TV Tuner card for the family computer and ditched our real TV in order to avoid paying the French TV Tax ("redevance").

It also allowed us to pirate Canal+, decrypting the video stream with a program called Moo TV (if I remember correctly). The audio stream required another program.

It was all a bit janky, frames were frequently dropped, and sometimes you had to tweak a few parameters to get a proper audio stream but the fact that it was (mostly) working was simply amazing to my brother and I.

Good times :)

[+] blincoln|5 years ago|reply
If it was decoding a broadcast signal, how could the unique serial number of an individual decoder play any useful part in the decryption key? I didn't see any discussion of this in the article. Sorry if I missed it.
[+] baud147258|5 years ago|reply
I've heard that a former employee at the company where I'm working was assembling pirate decoders on the side, but I don't really how they did for the keys, maybe shared between friends and family?
[+] 082349872349872|5 years ago|reply
France also having been SECAM might explain why russian retro-80's concerts[1][2] often feature "Emmanuelle" (1974, NSFW) K7s in their "do you remember" interludes?

[1] https://ru.wikipedia.org/wiki/Дискотека_80-х seems to be the main one, but it has competitors.

[2] from the 90s, but if the instigators of the US PMRC had ever seen eurodance like https://en.wikipedia.org/wiki/E-Rotic they'd have been clutching their pearls.

(the retro-encrypt screen's "pour en savoir+" mdr)

[+] fred256|5 years ago|reply
In the Netherlands, I remember electronics stores carried DIY kits with names such as "TV signal quality enhancer". I only later realized what these were actually for.
[+] jimmaswell|5 years ago|reply
I remember some weird little black box with a red button my parents had between the cable box and the wall coax. I wonder if that was it.
[+] pnako|5 years ago|reply
IIRC, aside maybe from some technical limitations, the reason it was not "proper" encryption where you can't make any sense of the signal is that it gave some idea of what was playing. The sound was completely scrambled but from the picture you could tell the program and tell if it was a football game, movie, talk show, etc.

I had that channel for a while and watching WWE with French commentary was a favorite when I was a kid :)

[+] gregoriol|5 years ago|reply
The "make sense of what could be" might have been a good way to make people tempted to subscribe actually: it was so frustrating to identify a program you could be interested in and not be able to fully see it.
[+] throw_m239339|5 years ago|reply
> I had that channel for a while and watching WWE with French commentary was a favorite when I was a kid :)

Tu veux dire le p•rn• du premier samedi du mois ;)

[+] danw1979|5 years ago|reply
Fabien Sanglard's blog is a true gem of the Internet. An excellent tech historian, in my humble opinion.

If you're enjoying this, you might also enjoy Creatures Of Thought at https://technicshistory.com/. Their history of the transistor series was superb reading.

[+] ajuc|5 years ago|reply
I think it was used in 90s in Poland by Canal+? At least when you tuned to the encrypted channel it looked exactly like it says in the article - every line was shifted randomly.

I know people bought cheap pirated decoders that decrypted the channel without the need to pay the subscruption fee.

It couldn't be the exact same system because we have PAL, but it was something very similar.

[+] pewu|5 years ago|reply
Analog encryption schemes are infinitely interesting to revisit today. The ingenuity of these folks trying to create an unbreakable encryption scheme with limited resources of 80's and 90's never ceases to amaze me. BTW A piece of software exists that can generate images encoded in various ways, that are supposed to be compatible with original hardware: http://cryptimage.vot.pl/cryptimage.php I have yet to lay my hands on some decoder and try this out myself.
[+] praveen9920|5 years ago|reply
I wonder how this would have worked in practice with lot of noise being added to signals. Is the noise too negligible to "delay" that is added?
[+] FraKtus|5 years ago|reply
The idea of the Radio Plans decoder is that the video black level was a little bit higher than the start of the signal after the horizontal pulse signal. Because of that, it was possible to detect if the delay was 0 1 or 2...

Canal Plus reacted quickly by adding a little bit of noise to the signal after the horizontal pulse.

[+] mpalfrey|5 years ago|reply
Pretty fascinating. CAS and broadcast tech is a pretty big interest of mine and I've been lucky enough to work for 2 of the biggest players in this space in the past (although not directly on the CAS side of things and sadly I'm not in the video space these days).

Some very cool engineering was done without much in the way of technology being available!

[+] aivisol|5 years ago|reply
> The audio stream is interleaved at the end of the blocks.

If I remember correctly, audio was transmitted on a subcarrier some 6 or so MHz higher?

Edit: Later in the article it is actually stated: "A normal SECAM signal uses FM on a 6Mhz carrier"

[+] FraKtus|5 years ago|reply
The sound just had the spectrum inverted.

That's why when inverted it was sounding like crickets speaking.

You could unscramble the sound just by multiplying it with the right sine wave at the right frequency.

I remember prototyping that with a modular synthesizer :-)

[+] themodelplumber|5 years ago|reply
It seems like three or four channels must have prompted at least some people to experiment with TV DXing. I wonder if that was very much a thing at the time.