I decided to try my hands at creating a sword with the PBR workflow. Here is what I have, I want to make sure I am not working in a vacuum so I have to post some renders to see what you guys here on polycount think. So with that said feedback would be awesome if you want to give it
What engine are you using to to display the model? The materials aren't working to well for me, the blade looks too rough to be metal, and the black parts just look like the sword was dipped in black paint.
I am using marmoset. I am going for the rough worn out metallic material (not shiny). The black areas are supposed to be the same base gray metal but painted black with chipped areas that indicate the underlying color.
Good start!
I agree with the material definition. Even for worn metal, you want the non-worn parts to be shiny, to create sexy roughness contrasts - in my opinion.
I also think you could reduce the polycount on the handle by 80% if you are going for efficiency
If the black parts of the blade are supposed to be painted black, why are the fairly significant gashes going through it not changing its color to the same gray as the outer part of the blade?
Certain scratches and dents in the sculpt also make it seem almost more hollow, as opposed to a nice strong blade. As an example the first dead-on shot, right underneath the first large gash that crosses between outer/inner blade starting from the top, that dent under that gash looks like something I'd see on a car door, not a sword, and there are other similar looking dents throughout.
Agree with the other comments that you need some nice shiny bits in there to sell it as metal, worn or not.
Spoon: Thanks for the feedback. I generally try not to get too caught up with topology. I generally just make sure it's clean and it looks like I know what I am doing. Most of the stuff in my portfolio are personal work so I tend not to squeeze out maximum efficiency since it makes the whole process more enjoyable for me. In this particular case it made for much easier normal map generation.
Two Listen: Thanks for the feedback. It's definitely something I'll look into further in terms of the kinds of damages normally seen on weaponry. I did add those dents in order to get some variation, but perhaps I went a bit too far since it looks like it may be hallow beneath.
Thanks everyone for the awesome feedback! I really appreciate it, here is an updated version.
Replies
I agree with the material definition. Even for worn metal, you want the non-worn parts to be shiny, to create sexy roughness contrasts - in my opinion.
I also think you could reduce the polycount on the handle by 80% if you are going for efficiency
Looking forward to see where you take it
Certain scratches and dents in the sculpt also make it seem almost more hollow, as opposed to a nice strong blade. As an example the first dead-on shot, right underneath the first large gash that crosses between outer/inner blade starting from the top, that dent under that gash looks like something I'd see on a car door, not a sword, and there are other similar looking dents throughout.
Agree with the other comments that you need some nice shiny bits in there to sell it as metal, worn or not.
Two Listen: Thanks for the feedback. It's definitely something I'll look into further in terms of the kinds of damages normally seen on weaponry. I did add those dents in order to get some variation, but perhaps I went a bit too far since it looks like it may be hallow beneath.
Thanks everyone for the awesome feedback! I really appreciate it, here is an updated version.