Text to vector jaggedy jagg
So, I've been messing with text to vector, and tried implementing it to my workflow.
When Illustrating sometimes am in need of an assets that create a look, so I tried prompting glasses and 9/10 am getting results that are to jaggy and not usable.. I ended drawing them at the end.
I tried pushing slider to more/less details and did not got expected results
-
Oh yes. Everything this generates is rather sketchy. It can’t create straight lines, accurate circles, nice arcs, it poorly understands symmetry. As Von Glitschka stated, "The ‘Text to Vector’ generative ai feature should be titled ‘Text to Clip-Art."
And a rather poor clipart — "The irony of ‘Text to Vector’ generative AI is the imagery is being cobbled together based on artwork that took skill and craftsmanship to originally create. And the derivative result lacks those same traits."
Imitation, simulacrum. -
Katarina Jovanovic commented
I imagine that there could be implemented some slider loose-tight that would (somehow) regulates the roughness of the shape