Trying a continued focus on the shape of faces and position of eyes. I think it's getting better, though the tilt of the face is off a bit; at least it is decently proportioned and the eyes aren't all over the place.
This is the last of the Porsche character sheets in this series, and I'm going to dive in to finishing the hands and feet Goldman book before switching gears to faces. Still, this exercise turned out not too bad ... and I am now seven days ahead on drawing every day AND scheduling them for posting.
Drawing every day, on average, posting as far ahead as I can.
Next up: editing work on SPECTRAL IRON, Dakota Frost #7!
Another study of a Midjourney-generated character sheet for Porsche the Centaur. (All of these similar-looking ones were generated using variations of a similar prompt; they never got much better, but they're good variations on a theme). Overall, getting better on the eye placement, now that I'm focusing on it; also the head to hair ratio is better but not great; also the shoulderpads are high or the chin is low ...
Drawing more or less every day; posting as regularly as I can.
Reviewing my past art, I noticed that mis-aligned and mis-spaced eyes, and generally poorly placed and scaled faces, were a problem. So I tried to focus on that problem this time, and ... it didn't turn out bad.
As usual, it's Porsche the Centaur, via a Midjourney character sheet, again. Though her hair is WAY too huge ... I guess I really do romanticize big hair.
Another drawing of Porsche from a generative AI character sheet (I think this one was from ChatGPT+DALL-E, which seems to be a bit better than Midjourney at taking art direction and creating centaurs). I modified the face to more closely match Porsche, whose ears are located more closely to a normal human's ears.
I've started to build up a buffer, like I am for the Blogging Every Day series, by trying to do two drawings at every sitting. I can't manage to draw for an hour and a half every single day, but if I do it most days, then I slowly creep ahead, and can put more effort and thought into each drawing.
According to my spreadsheet, I'm now about six drawings ahead, drawing-wise, and two posts ahead, posting-wise. Maybe I can take some time to, you know, write about these characters now.
Quick sketch of a young woman with GREAT hair - I'd say almost Cinnamon-style hair, if there hadn't been a woman with almost EXACTLY Cinnamon's bi-colored hairstyle at the same Barnes and Noble about a week ago - whom I decided to sketch after finishing my last drawing.
She moved too fast through the store to get a really good picture, and from where I was sitting it was hard to see what she was looking at, but I think you've got to go beyond just drawing from practice books and start drawing from life, or you're just regurgitating other people's drawings, like an AI.
And I like AI, but regurgitating other people's drawings is NOT why I am drawing every day.
Me drawing from yet another Midjourney character sheet for Porsche. Misproportioned, and I am still having trouble on the eye placement and face proportions, something I've seen on a fair number of old drawings; once I finish the hands and feet book I think it's probably time to go back to practicing faces.
This certainly is rendered better, but if you look closely at the edges of the hair, especially where it meets the armor, it doesn't make too much sense. Oh well, it's still inspirational.
My rendering of the last pose from the DALL-E character sheet for Porsche:
Not entirely terrible, though I can see my proportions are a bit cartoonish. These systems can't take art direction yet - I had to clean the character sheet up in Photoshop to really make it suitable, and even then the middle pose should have had the legs more spread apart, which it tried to do erroneously on the right-hand pose with a fifth leg - but they sure can render the heck out of an image.
Another sketch of Porsche, this one based on a "character sheet" I generated with Midjourney based on my own character descriptions. Yes, yes, I understand there are many issues with AI art generators, some of which are real and some of which are not; my own writing has been used to train large language models, I know artists who have been discouraged from art by generative AI, and I expect the use of these systems for final product will be very disruptive. HOWEVER, there are some things which AI generators can do which can't be feasibly done by any other method - such as producing character sheets for your own practice.
As I mentioned in an earlier blog post, I managed to convince ChatGPT / DALL-E to cough up the above "centauress in a spacesuit" drawing after many iterations, then got it to transform that image into the following character sheet of Porsche from several different angles:
It ain't perfect - note she has five legs on the right, and the perspective differences aren't really that well done - but it's close enough for me to use these as reference as I do my own drawing practice, shown at the very top. Seeing these "drawings" gives me new perspectives on how to render my own ideas at a higher level than I can currently do, enabling me to bootstrap my artistic skills.
There's a long way to go for AI generated art to be able to respond to art direction - but it is helping me chart a new direction in my own art, which is great, because if I can draw it, I can then art direct myself.
We have a black cat, so we got a black cat condo (just barely visible to the left). But of course, our cat-shaped void is a cat, and so prefers the blue couch, where its voluminous shedded fur is easily visible. My wife caught him in the act, so, enjoy this picture of our cat-shaped void, doing cat-styled things.
-the Centaur
Pictured: Loki on our couch. Interestingly, this picture was taken at an angle, so I rotated it, then used Adobe Photoshop's generative fill to recover the outer edge of the picture. The very outer edge is ... mostly right. Some weirdness is visible in the carpet patterns on the lower left, the brick pattern on the upper left, and whatever it is on the table on the right isn't there in reality. Otherwise, not a terrible job.
Okay, I'm going to start out with the best of the images that I produced trying to create Porsche the Centaur using ChatGPT's DALL-E interface. The above is ... almost Porsche, though her ears are too high (centaurs in the Alliance universe have ears a little more like an elf, but mobile like a dog's). And, after some coaxing, the ChatGPT / DALL-E hybrid managed to produce a halfway decent character sheet:
But both of these images came after several tries. And when I tried to get ChatGPT / DALL-E to generate a front and back view of the same character sheet, it just disintegrated into random horse and human parts:
Similarly, the initial centaur image came only after many prompt tweaks and false starts, like this one:
There are legitimate questions about whether the current round of AI art generators were trained on data taken without permission (they almost certainly were), whether they could displace human artists (they almost certainly will), and whether they will have destructive effects on human creativity (the jury is out on this one, as some forms of art will wane while new forms of art will wax).
But never let anyone tell you they've worked out all the bugs yet. These systems are great renderers at the image patch level, but their notion of coherence leaves a lot to be desired, and their lack of structural knowledge means their ability to creatively combine is radically limited to surface stylistics.
One day we'll get there. But it will take a lot of work.
Just Loki, on the back patio, looking at a leaf ... with a little added magic (full size).
Producing this relatively simple image actually involved a fair number of Photoshop tools, several of which are new "generative AI" tools, but many others of which are just plain old machine vision magic:
Layers (stacked images) used extensively to save original or alternate versions of things.
Perspective Warp (a pretty impressive tool in its own right) to distort the image into a rectilinear shape.
Content Aware Fill (a new Photoshop generative AI tool) to extend the warped stone tile to fill the frame.
The Clone Stamp tool to complete the grout lines which were only partially filled in by Content Aware Fill.
Quick Selection tool to isolate Loki and the leaf into their own layers for later.
Selection > Modify > Expand and Selection > Modify > Feather to get the fine hairs on Loki's boundary.
Generative Fill (another generative AI tool) to eliminate many of the leaves.
More Clone Stamp to eliminate more leaves and minor imperfections.
Layer duplication to create an original and to-be-colored tile backdrop.
Swatches, the Rectangular Marquee tool, the Polygonal Lasso, and the Fill tool to create the colored tile.
Color Burn layer blend mode (with 57% opacity) to create the primary Mondrian effect.
Another layer duplication to create a new version of the colored tile to enhance the grout.
Filter Gallery > Colored Pencil which fortuitously greyed out the colored tile and colorized the grout.
Magic Wand tool set to Contiguous and 0% Tolerance to cut out the greyed tiles from the grout layer.
Darker Color layer blend mode to enhance the grout.
Drop Shadow on the leaf to make it stand out.
Duplicating Loki into a layer with Darken to make him stand out against the colored grout.
Adding Inner Glow modified to Darken as well (with a Choke of 14 and Size of 87) to eliminate some of the white halo around Loki.
Adding a second Loki layer, Normal blend with 50% opacity, to get his sheen.
I like how it came out, especially given how it started:
I looked at that and thought, "You know, that's almost a Mondrian backdrop" and I was right!
I still have misgivings about using AI-generated art to create final designs without human intervention, and I think AI art needs to address the copyright issue in a meaningful way, but speaking as an artist into cosmic horror, it sure can create some creepy images that are great food for thought. Here's a couple of cool ones from a recent project that I've been working on - great design concepts, whether or not they get used.
Bonus points if you can guess which work this art is designed to illustrate.
SO, one of my favorite characters is Porsche, a centauress warrior from the thirty-first century who populated many of my first tranche of as-yet-unpublished science fiction stories. (I think she only appears in one published story, "Stranded" in the anthology of the same name, and even that, just as a cameo). And while I have worked a lot to improve my art, I wondered what Midjourney could do. And I got the above result from the following prompt:
a centauress with long, rich curly purple hair, very beautiful, with half-asian, half-english appearance, and pointed ears, wearing an armored space costume like a combination of ghost in the shell and star trek, and bearing a double-bladed scythe with black glowing blades
- the Centaur
Wow. This is really spot on. Her hair is right, her face is right, her skin tone is right, her armor is right ... heck, even her slightly haunted look is right, and to go beyond even that, one of the variants looks like a slightly older, more grizzled variant, which completely checks out with her storyline:
Kudos to you, Midjourney, except ... she's not a centauress. She's just a person, a fetching one, I admit, but not a half-woman, half-horse creature with the pointed ears and black twin-bladed scythe of the prompt.
Well, shoot. What if we look at some of the other variants?
This one is creepily good in a sense ... it's got her forehead dot (she's a First Contact Engineer, and wears a pheromone bead she used to communicate with a scent-based alien hive species) and even hints of her mechanical arm and possibly ear. But this is just coincidence. Look at this other variant:
What appears is just chance. Here, her ears are rounded, the dot's gone, and the weapon looks even less on point. A lot of what looked right to me is just random features onto which I was projecting, like cloudbusting.
Well, double shoot. What if we refine the prompt? What do we get?
a centauress (a creature with the upper body of a woman and a lower body of a horse) with long, rich curly purple hair, very beautiful, with half-asian, half-english appearance, and pointed ears, wearing an armored space costume like a combination of ghost in the shell and star trek, and wielding a scythe with black glowing blades
- the Centaur
Yerk. That's ... just jumbled nonsense. Tweaks to the prompt to make it simpler just produced women on horses. Midjourney does not apparently understand the concept 'centaur' in any meaningful sense. I tried just the prompt "centaur", and ... um ... yeah ... no, I'm not going to show you those. They're just a guy with a horse, or sort of on a horse, or ... sort of ... in a horse? A centaur as envisioned by The Thing.
Okay, one last try. What if I give it one of my pieces of art, and then ask it to render it anime style? Let's hold that piece of imagery till the last, but the prompt is:
an anime style centauress with purple hair and a double-bladed scythe jumping in front of a waterfall
--the Centaur
Oh, lordy. And I'm not going to show you the one it tried to generate from the prompt "anime style".
Oh wait, I am!
Wow. Evocative - the top left reminds me of Cinnamon Frost - but it has little to do with the image I put in, and the attempt in the top right especially is nonsensical.
I am inspired by Midjourney. It's definitely a better renderer than me, and has good ideas about composition which I have already used in my artwork.
But I stick by my comment that it is an amateur which has taught itself to render very well, and cannot take meaningful art direction. As limited as I am, I'll stick with my own drawing, thank you!
Like this one, the image I gave to Midjourney above. It's not perfect, it's not well rendered ... but it is mine:
And she has four legs, a scythe, and pointy ears, dag nabit.
SO automatic image generation is a controversial thing I think about a lot. Perhaps I should comment on it sometime. Regardless, I thought I'd show off the challenges that come from using this technology using a simple example. If you recall, I did a recent post with a warped bookstore picture, and attempted to regenerate it using generative AI with Midjourney. Unfortunately, the prompt
a magical three-dimensional impossible bookstore in the style of M.C. Escher
me
failed to pick up the image for some reason. After a few iterations with the Midjourney Discord interface, I got the very nice, but nonsensical and generic, AI generated image you see up top. After playing around with the API, I realized that I likely had formulated my prompt wrong, and tried again to include this image:
On the second pass, I got another, more on-point, yet still nonsensical image as you see below:
These systems do LOOK impressive. But they work like ... amateurs who've learned to render well. They can produce things that are cool, but it's very hard to make them produce something on point.
And this is above and beyond the massive copyright issues that arise from a system that regurgitates other people's copyrighted art, much less the impact on jobs, much less the impact on the human soul.
