There was a lot of posts, so I'm going to just summarise them here:
On "Stylized Morse"
As mentioned previously, the morse is stylised. Any audio morse translation software is going to be listening for "dit dit dit dit dit", not "wip wup wip wup wip". An equivalent of this is if I used OCR software on Wingdings and expected to get Romanised characters out; it's not going to happen. But the underlying Wingding character codes are the same, Wingding A == 65 == Arial A.
Alternatively, ask an Australian and someone from the UK to both say "G'day" and you'll get two very different results which are structurally identical. Doesn't mean one of them didn't say "G'day", they just said it differently.
There *are* slight variations between recordings, but it's only ever in amplitude or pitch; structurally (and that's the important bit) they're the same. For example, if I wrote:
@@@ ### @@@
%%% $$$ %%%
@@@ $$$ @@@
*** ### ***
As representations of different recordings of the same segment.... have I recorded the same thing or not? Consistency of structure is what I looked for primarily. Ultimately, regardless of stylisations, for the letter "A", I listened for "high low", not the exact same noise.
On "Write a galnet post" and see if FD says anything.
Already done, and unsurprisingly, nothing. I've written articles three times now, and nothing has ever been picked up. So, there's numerous explanations;
- I'm a terrible writer (very true, I'm a programmer in my day job, and a radio operator in my secondary job )
- FD don't want to "have their hand forced" just yet.
- Tying it in even slightly to a Power isn't the direction FD want to take.
- FD simply don't have the resources to put towards that sorta thing at the moment.
I don't think FD staying silent is a smoking barrel.
On uninhabited systems
Multiple recordings by wishblend were done in uninhabited systems. At the time, EKURU A 1 was done as a prediction because that's the only system name with letters I knew. Now I have the full alphabet, I could "predict" and uninhabited system relatively easily (not right now though, because I'm going to work).
<snip> Poorly thought through paragraph, which I apologise for.
If it really is morse?
Honestly, I'd be happy to say it's not strictly morse, but a symbol set based off morse, with a set of (as yet, undetermined) rules which cause slight variations within each symbol.
This would be similar to how an english "c" in center is pronounced "ss", but a "c" in "cheat" is more like "tyh". Unfortunately differences in English are quite radical compared to the differences in this. The sort of differences I'm talking about is that three dits (dit dit dit) will be "wip wip wip" with a slightly descending pitch in each one, compared to "wip wup wip wup wip" for 5 dits.
For what it's worth, the "P" symbol is quite "different" from it's morse base, under these circumstances. Listening to multiple recordings, it's supposed to be "dit dah dah dit", but it's coming out more like "dit dah
dit dit". This was consistent for any recording with "P". This could be stylisation, or it could be something.
On "hearing what I want to hear"
For whatever it's worth, while dissecting the audio cuts, there were many times when I either played the wrong audio cut, or had the wrong text in a morse translator. Every time this happened I listened and went "Wait, this is completely wrong",, then realised my mistake. If I was just "hearing what I want to hear", surely that wouldn't happen?
On software not finding an answer
Basically, the Mk I. Human Sensor Array isn't a fast heuristic process compared to most software, but it's unquestionably more accurate than software, in my opinion. That's just my opinion though. I can make a computer think a word document is a JPEG image. I can't make myself do that.
From here?
I'm leaving it from here. If people still think it's wrong, I'm not going to try and prove them otherwise. I've put everything I've got into spitting this stuff out, and have had some great help. Some people never will be convinced I think. But if you want to prove me wrong, it's going to take presenting a recording with the listed location that doesn't match up at all. As mentioned, this diagnosis would've failed over 20 recordings ago if I was just "hearing what I wanted to hear".
EDIT: Oh BTW Wishblend, thanks for the Ekuru A 1 recording, as expected (imo) there's stylistic differences, but they're pretty much the same. Happy for anyone else to verify.