smithy, Subject: Amplitude in the 408 and 340 Sat Sep 01, 2012 3:19 pm
I tend to kill 340 threads it seems, so I’ll start one!
Now this is pretty damn esoteric. It’s rubbish, but I found it entertaining!
IF the 340 is prepared in the same general way as the 408, just "more so", this might work, perhaps.
IF the pluses in the 340 are random overlays of other substitutions and not something to do with "nulls" or "pivot points" or something, and last,
IF there’s no transposition, it might be useful. (Naaa, it will still be rubbish!)
Get coffee, here we go.
If you put the 480 into ZKD, you get a lot of numbers against the symbols. These are a representation of the appearance of the symbols. Simple!
Here’s what that looks like out of the ZKD and back into EXCEL.
So, the orangey numbers are what the word "KILLING" looks like in the raw numbers of the 408. Exciting eh?
Now because the numbers in the 408 and 340 (the amount of symbols and therefore the number of characters, and probably the multiply-represented characters too) are very different, I can’t go looking in the "raw" 340 for those strings. Wouldn’t it be nice, huh??!
They are indeed useless.
But those strings in the 340 have "amplitude", if you like. A simple line graph of the word "killing" looks like this.
It’s a graph, it has points at different heights. Just like my other award-winning diagrams on the other thread have. Amazing.
Is it any use to know that? I don’t know, really! 
Here’s what the 408 and 340 plotted together look like, using ZKD for the numbers.
The parts I highlighted in the 408 are the words "KILL[ING]" and "BECAUSE", since I thought the 340 likely to have one or both in it.
The highlights in the 340 could be anything of course. I just liked them. Some of them are strings you can see and recognise in the symbols of the 340, looking at the cipher text. What a coincidence! Some of them aren’t. How interesting.
What are the chances we can look at the entire 408 and the 340, like a wave-form on an oscilloscope? Tweaking the maths and looking for "trends"?
Run one backward and forward against the other – fiddling with the "amplitude" as we do?
Will that mean we’ll actually be recognising words in the 340 which look like those in the 408?
What about creating other ciphers with "low amplititudes" (more substitutions in) which contain the words (like "killing") we think might be in the 340?
Pointless?
It was fun to share, whatever the answers! Thanks!
