The double vertical strokes are the letter N (60% sure). CONGNACI, perhaps. Sims-Williams' Celtic Inscriptions of Britain is particularly helpful in understanding early medieval name forms.

Currently stuck on reading a very eroded part of an inscription. Even on the monument just a couple of chalk strokes indicate that there's something there. The 3D scan shows much more when you filter it correctly. But it's still quite far gone. Going to use all of the enhancements I can think of.

Today is all about file organisation and data processing. I'm also comparing datasets collected with two cameras, one of which is an iPhone 16 Pro Max using its 48 megapixel files.

The second monument is perhaps Cornwall’s only inscribed stone situated next to a lovely warm radiator. It’s the ‘Selus’ stone in St Just Church. We’re using 3D enhancement techniques to check for missing details.

First monument of the day: the early medieval ‘Qonfal’ (personal name meaning “great wolf” according to Prof Charles Thomas in 1993) memorial in Madron Church.

Time for new strings on my mandolin.

Today I am mainly working on Sourton cross, analysing 3D capture data of the supposed chi-rho to see if it’s really there, what form it takes, and to make some clear images of the early medieval inscription (6th-9th century).

Very atmospheric conditions for a bit of recording today.

I've been playing with ChatGPT 4 to see how useful it can be for research and decided to blog about it. I focus on not using the giant erratic pool of knowledge in the LLM but instead on using it with my own data plus some other uses. TL;DR It's useful but not perfect: tom.goskar.com/2024/02/13/u...

Just having a read through the BlueSky privacy policy. They don’t respect the “Do Not Track” preferences and collect a lot of data. I had hoped that it would be a little more enlightened as the company is allegedly a “public benefit corporation”. I’ll watch this space…