To be fair, you're quite good at blender.
post
Viewport preview to prove I have the 3D models made, that the original "artist" was unable to provide.

Bonus render with the visor up that the op couldn't do either.

Nooo, don't open your visor! You're standing on the moon!
they tricked you into comitting murder.
A few years ago that was a given. I believe that you did that yourself. But AI frontier models, or some of them, can generate blend files now. They are probably not as good as a person who knows Blender well, and they still can't do the more complex texturing and such. But they might be able to do the base model.
Now, it wouldn't take the 30 mins they took to use AI to do the first one, lots of edits and redos to get things right. To do a lot of models, it's best to just learn Blender. Hell, you might learn Blender by having to help the model correct mistakes. But they can do some basics now.
You can get chat gpt to use blender now too. Watched a video yesterday where you can just straight up tell chat gpt what you want and it feeds python scripts into blender that do the modelling. So you can't even use wireframes or alternate angles to prove its in blender now.
But it took the YouTuber 90 minutes and like $20 of tokens to do one scene. And typical to ai it looked great at first glance but on closer inspection things were weird and required fixing. Like cutlery on the table was rotated wrong and floating, chairs were clipped into the wall, window panes were beveled an insane amount but it was in the frame so it's just wasted geometry.
And for a person who's never used blender the UI and navigation is confusing, so it's still not as accessible as saying their ai slop is a drawing lol. Having the ai render out a fake wireframe or blender viewport is still hard to get right because it requires an even grid or a lot of text, so an experienced blender user could still tell on a closer inspection.
The ai one had a tumor on the bottom of the moon and the reflection on the helmet is.... Odd.
I'm currently learning Blender, and I'm in awe of you!
Yeah there were a lot of inconsistencies that made it obviously ai, and to do them in blender would have been difficult to achieve intentionally so it was pretty dumb for op to try and argue that it wasn't ai.
-
The glass visor has reflections in some places, but not all of them so it looks like there isn't one on the bottom half of her face.
-
There's no distortion of her face from curved glass. If you compare my renders with the visor down vs up, you can see it magnifies get face a bit and you can't see the neck even though it's there in both renders.
-
Her outfit is asymmetrical.
-
The visor shows a light to the top right, which is correct for her helmet and the right side of the cheese, but then her suit is lit from below to the right with the shadows cast upwards. Then the highlight in the cheese is from the top left as if the light is directly over her, but that wouldn't light the side of the moon so evenly.
-
And it's low resolution, with a weird crop. Idk why anyone would set their camera to render 1003x1030, while my render is a 1080x1080 square.
How long does one need to practice to produce your result in 30 minutes?
At the time I had a couple of years of faffing about blender and doing tutorials and little bits of art. Nothing serious.
To be fair, 30 minutes is a speed run and the model isn't really usable for anything except a still render. There are a lot of improvements I could make I reckon. If I wanted to put her in any other pose for example, it would take a few hours to tidy the scene and fix the mesh. If I wanted to animate her or make the model game ready it would probably take me a day or two of full time work.
It's very rewarding though, I highly recommend learning blender. And it's free, which you can't say about much these days.
Thanks for the estimation on further tasks, and thanks for sharing your art.
Back in the day I studied to make computer games, graphics. Learned Maya in school, scrapped that shit and learned Modo instead. Fit me perfectly but that is a long time ago. Some time ago I thought that I wanted to get back to basic modelling again and installed blender. Everything is super confusing and I felt totally noob and can't even make the most basic of things so seeing you able to do this in 30 minutes is really inspiring.
I should really pick up some tutorials. Heard there is a donut that needs to be done again 🙄
The donut is a good one. It's like riding a bike I bet
3D software has changed a lot over the years, blender included. It can be pretty confusing until you get the hang of the hotkeys and navigation. Blender does a lot so finding your way around can be annoying and every action has a shortcut that you either have to already know or Google to find out lol. But you can change the key maps to match other software (like Maya) which might make it easier for you.
I've heard a few people say that the donut tutorial is not that good for beginners. Maybe it would be useful for you because you already have experience with other 3D creation apps, but for people new to 3D, I've seen recommendations for really basic tutorials for low poly modelling with basic texturing and lighting. The idea is that you acquire the core concepts doing something ultra simple, and then move on to other tutorials that more closely match what specific 3D subject you are trying to learn.
Love it! Any chance you're willing to post the deslopped image by itself too?

I used to make stuff with 3DS Max back in the day.
Been wanting to learn Blender for quite a while.
This post might have motivated me to get off my fat arse and try it, we'll see.
I feel like the cheese needs some sub-surface scattering. Would give it those hints of orange/red where the light passes through the thin edges of the cheese.
That said, I still like what you've made 😊
That is super cool. Years ago I followed a a Blender tutorial and made a teddy bear. It looks really life like and was fun to make. Not touched Blender since.


all 37 comments