Completing the Magician, a documented process

At last, I have found the image I want to use for The Magician tarot card, and it is time to complete the final tweaks on the art work. This will use a combination of upscaling and inpainting, as well as perhaps some masking in order to enhance features of the image and get it up to scratch, as there were quite a few opportunities to restore the coherence of the image which I'm using as the starting point.

holoz0r_the_magician_played_by_beautiful_alluring_woman_holding_e4c67842-7a9b-4844-8337-038c146cb2ed.png

I have the first task of running this through codeformer, to help enhance any of the facial features. I've also asked the script, at the same time, to conduct out painting.

What I got as a result was certainly not what I expected, and I'm not sure if I am upset, or happy. I quite like the murky, mystical doorway it gives to the feature image.

image.png

I don't think I will use this as the final, but instead, I'll just run the image through codeformer normally to restore the faces, then have another good hard look at the image (and the original card of The Magician!) to see if there is anything else that I might add using inpainting to enhance the symbolism in the piece.

A few in-painting attempts to try and fix the facial features didn't go exactly as I expected, instead removing the face entirely!

image.png

So, instead, what I will do is use img2img in conjunction with a negative prompt (and no other prompt) to see what the outcome and result of the image is. I like being surprised. It is always fun. I certainly didn't expect this as the outcome!

image.png

So instead I will do some simple upscaling, combined with the face correct algorithm to make it a little bit better as the next step, instead of just being a forest.

image.png

Here, the left hand is fixed, but the right still needs some more work. I will still also need to rerun this (once it is fixed) through an upscale process to increment the detail and resolve the faces a little more.

I will continue working on the other hands in the image. It only took 64 iterations to get that single hand completed, let me see how many it is going to take for the next hand to be fixed.

After generating a few hundred more iterations, here is where I've gotten up to. It isn't perfect; but its pretty close, and the hands look a lot more realistic.

00000-1999621596.png

I think that's where I am going to call this one done for the time being, but first, I will need to upscale the image a little bit more and run it through codeformer to see if I can add more detail in the faces.

image.png

Again, just like the fool before her, The Magician will need to get a border added; and some other things, like words and stuff.

For now, that's all, until I start culling images toward the idea of "The High Priestess".


Want more content from me?

Witness my futile efforts to play my Steam Game collection in alphabetical order.

Are you aware that I love photography? Check out my work in a collection.


If you haven't started playing Splinterlands, you should do that immediately. It's very good fun.

If you want to see my Splinterlands antics and rants live, Find me on Twitch

If you prefer sleeping in your designated time zone, go watch replays on YouTube.


Thanks as always for your time!



0
0
0.000
2 comments
avatar

By delegating your HP or some token to the @thranax bot, you are helping to support the community and the bot and earning healing rewards daily! and also you will get token with healing

0
0
0.000