[WIP] Fish Shooter - Hard Surface Model

Toureiro

Hi all. 

After years of doing everything from indie games, teaching game development and working on virtual environment productions, I developed a serious itch to go back to my roots and focus on 3d modeling.

I'm starting this project in my spare time, inspired by this incredibly wonderful concept of Garing3


Besides looking super fun to model, I'm foreseeing some classic challenges when it comes to translating this concept to 3D.

The idea it to turn it into a game-ready asset for a third person game.
(Do you think I should define which game at this stage, so the style of it shapes the modeling?)

LAYOUT

First I invested some time in the Layout phase, trying to align things, seeing where things break when putting simple 3d shapes in front of the concept.


I'm using blender for this one. Since I don't have a model sheet, I prefer to use the Background Images feature of the cameras. This way, the image stays locked in place as I move the camera around and change the focal length.

It's a more familiar workflow for me, coming from Maya and Max.


I also like to place the camera in the first collection, allowing me to quickly toggle the reference on and off with Shift + 1.


I could have spent more time in this phase, but since I'm pretty sure I'll need to make adaptations to the concept to make it work in 3D, I move on to blocking it out as soon and I tested a few configurations and am comfortable with this as a base. 

BLOCKOUT

The first thing I realized was that the barrel couldn't be a perfect circle, as the horizontal cylinder on top (what do you think it should be called?) would be too wide.

This cylinder needed to be wider than the body, but not so wide that it deviates from the concept. However, it also can't be as narrow as the concept, or it won't stick out from the main body.

I'm still using separate pieces, trying to figure out how they can work together:

At this point I'm seriously tempted to try out the newly released Conjure SDF addon 


So this is where I'm at now:

Notably, I decided to rotate the top cylinder. It's a slight deviation from the concept, but I think it visually reduces the perceived width while still ensuring it sticks out from the body. I rotated the side panel for similar reasons, as it would protrude too far to the sides in the back view.

I'm using a subdiv workflow at this stage, as I'm most familiar with it, and using temporary creases to hold the shape. I usually swap them with control loops when It's time to work on the highpoly but I'm considering keeping them and trying out the Zbrush workflow. Still not sure. Feedback on this would be apreciated.

Also, should I avoid a subdiv workflow at this stage, since it can be slower?

Have a nice day!

iam717

Not exactly a weapon "modeler", so can't say much but this question:

Also, should I avoid a subdiv workflow at this stage, since it can be slower?

Is answered to me by this top of the post:

I developed a serious itch to go back to my roots and focus on 3d modeling.

Subd is a part of modeling since you wanted to go back to modeling I'd say you are doing what you wanted by going back to the roots.

Toureiro

Had some free time this weekend.

The front view of the concept is my main reference, and I liked the shape of the backside from that view. However, trying to cajole the model to match that reference was a major pain in my backside. Ultimatly I decided sticking to the front view for the back of the weapon would result in a less faithful model overall.

I also created a mockup of a third person view to gage detail sizes and proportions. The bolts on the back attachment were the biggest changes, as they looked too small in this view (even though they match the ref).

GAME VIEW

PANNELING

I'm also deciding whether I do real paneling or use floaters.

TEST RENDERS

Here's the progress thus far. Shading is messed up in parts due to temporary creases.

sacboi

Neat looking model and concept also really liking those fat bevels  :thumbs up:

"...I'm considering keeping them and trying out the Zbrush workflow. Still not sure. Feedback on this would be apreciated.

Also, should I avoid a subdiv workflow at this stage, since it can be slower?"

Yeah zbrush, I'd say go for it probably add that extra dimension of realism up close i.e. surface imperfections dings - dents....etc and as for your high poly, well a subd workflow at least from my pov wouldn't be to much of a hassle iterating, since it looks as though there's an overall low mesh density and/or complexity anyway.

EDIT:

For paneling imo hard surface decals are worthwhile mentioning here but floating geo is just as valid in this use case aswell.