Thank you for that! It´s so amazing and frightening at the same time what AI can do. Iám playing around with runcat too - didn`t understand some things - but let´s see.
why frightening/ in this example, it’s simply a ne rendering engine/ tool.
no one freaked out when keyshot came around.. give it a go…i think you may find some interesting opportunities. My biggest complain so far is the pay as you go / credits used business model for this type of stuff.
Vizcom is definitely worth looking into as well. It’s very quickly becoming a standard tool for product
I love Vizcom, when I saw your runchat screenshots I thought you were using vizcom—I didn’t know the workspaces had similarities. I’ll try it, thanks Kyle!
I mean, how can this tool know where to place the mudd on your car? Why does it create these gaps and rivets in this way? It added a wheel and glas at the right place - in your input image there is´nt! The image looks great that is fact! And yes, AI is just a tool.
There is an interesting workflow here where you can kind of send it and see what happens, ( I can this the FAFO phase) then you can go back and refine specific areas and details with remarkable detail. I’m still learning, and am far from an expert in this stuff, but there is a LOT of power here when you get a handle on it.
The AI recognizes the environment and maps it onto surfaces, just as it does for lighting and reflections. It also understands where specific elements are most likely to appear. It’s important to include in the prompt that, in this case, the second image is the correct one. This allows everything to be connected effectively.
In the same way it handles light and reflections, the AI analyzes the object, identifies what is statistically common for that type, detects what’s missing, and fills in the gaps accordingly—placing elements where they are most likely to appear. In this case, it’s likely emulating a Fangio-era car, as suggested by features like the glass.
