If the last retouching attempt was more concrete, now that I'm more or less getting the hang of it, I feel drawn to experimentation. I wanted something more sophisticated and unusual (that desire always gets me down!) Also, I finally figured out how to work in the neural network editor to get better images.
When I say "better", I mean images that take me several times less time to edit. The "neural network will do it for you, just get the request right" level is still beyond my reach. XD
Anyway, it started out as a mirror image of the previous art, in which the AI had a bit more freedom and so Harry looks like he fell out of a parallel universe on it (and I really love the parallel universe theme!) Here it is, by the way.
So I really wanted to connect the two arts somehow. And because they were compositionally identical, the idea of the mirror-portal came up by itself.
The second point was to get this portal-mirror itself. Since it was clearly a non-standard request, the easiest way to explain to the AI what I wanted from it was to use my fast drawing (I don't think I'm going to show it). Of all the options, I liked this one:
And now for the thing that caused the most problems (well, apart from the back angle, which was easier for me to deal with on my own), the background. It was at this point that I started to think that sometimes it's easier to draw something yourself than to get the right option from the AI...
But through a lot of variation and brainstorming, I got the right one. Like Harry, I put it together from bits and pieces, so it's a little different from the final version.
It is difficult for me to judge how successful the implementation of the original idea was, but it was definitely a very educational and developing experience))
snowghoul
theartofmadeline
kianamaiart
stormborns
dededos
tangentwithtanner
dailybridgerton
