My name is Philip Dunckley, I am a young games artist from the UK with a lot of enthusiasm for creating. I specialise in Environment Art, hard surface modelling and texturing. I’m a keen, hardworking individual with a love for 3D sculpting and everything to do with it. For me this is not real work, it’s a passion. Today I will be walking you through the process of creating a Beretta Gun.
I’ve always found when I’m working on a personal project that it’s important to have a passion for what I’m creating and to find something that interests me. A few years ago I found a hobby in paintball and airsoft and over time I have amassed a small collection of my favourite airsoft guns, my favourite being my Beretta. With this in mind, I created a piece around the Beretta, for which I had the physical reference for already. If I didn’t have that, I would gather reference or create concepts for an asset, as a starting point for a project.
For me this is not work, it’s a passion
When it comes to building something currently existing and non-conceptualised, no matter what the object may be I never trust my eye alone, it is important to get isometric or even better a blueprint style image as a guide. Ideally, I would want at least a top, side, front and back view of the item to arrange in my scene to build it as accurately as I can; set up and preparation is critical.
Creating The Asset
So where do I start building? It’s the biggest problem I have when starting a project. Typically I find that starting at the part that catches my eye or defines the object as a whole is the best way to start. In this case, it was the slide and so I built the whole thing out from there. I created the low poly mesh to then duplicate it into a new layer and create a high poly mesh, adding edge loops (or retopologizing the entire mesh if necessary) and most importantly following the guides I have previously set.
When UV mapping I started by trying to keep things uniform in world scale and then break the object down to find objects or faces that are small or unlikely to be seen. I optimised these more obscure areas to use less UV space to make more room for the more important components. The decision to take UV space from one component and grant more UV space to another to increase resolutions is a trade-off as texture space is valuable.
High-resolution mesh is where the quality comes from and so the edges must be consistent and clean
I have found hard surface modelling is my strong point and the most important tip****I can give when it comes to hard surfaces, is that the high-resolution mesh is where the quality comes from and so the edges must be consistent and clean. If the edges are wonky, it will show, if they’re too sharp, they won’t have nice edge highlights and if they are not sharp enough, everything looks too soft. When defining edges, you should consider the normal map resolution too; will it render out clean with enough resolution? That doesn’t mean the low poly should be neglected, though, a good rule to follow is to keep an eye on the Silhouette. A good tip for this is to turn on the ‘Use all lights’ option in Maya without lights in the scene and meshes will appear as a black Silhouette. Doing the above is an excellent trick and the aim is to reduce apparent hard edges.
Texturing can be a lot like starting a mesh; I start by rendering out a few texture maps like ambient occlusion, edge maps and normal maps, then Import everything into a 3D viewer (Marmoset) for viewing to get a sense of the object in real time; its size, mass and overall feel. Does it look like it will fit in my hand? Is it comfortable?
Related link; How to create a realistic Canon Reflex Zoom 8-2**
From there I would take it into Photoshop and use all the generated images as masks or added layers to bring out grunge or worn areas. Keep the objects components in mind and how it all moves when adding ageing. It’s always important to consistently check any progress made by saving out the textures and viewing it in 3D.