In this year, there will be many modelling projects that I will find hard to tackle, so I made sure to stick to my guns and start planning and preparation as early as possible. Last week was when I first started thinking about modelling a revolver, as I knew that this would most likely be the one I have the most trouble with. I wanted to ensure that I had the best shot at getting this model spot on, so I started my planning and research just a few days after hearing about the project.
I started off by gathering reference that I could use while modelling and texturing. Most of the reference I gathered will be used later in texturing to add more detail to the barrel. Since the revolver I am creating is based off the ones Roland Deschain uses in the Dark Tower series, I have a bit of creative freedom, as the revolvers he uses don't really have a real life counterpart.
I looked at the Wiki page about the sandalwood guns I'm modelling off of, and even read through all the debates to figure out if there was an easier way to find reference for the revolver. However, after coming to the conclusion that the guns are never described in great detail, I decided that using a real revolver as reference and finding intricate designs to add onto it later as normal maps. Some of the larger motifs could probably be modelled, but if I do need to model them, I shall leave this for last.
I paid extra attention to the shapes of the revolvers, to try and get an idea of what polygon primitives to start with. This helped a lot with finding reference for the etching up the barrel of the gun, as I could easily look for items with a similar shape. The silverware and jewellery I found for reference are mainly art nouveau, gothic and neogothic. I decided on these styles because of the fantasy undertones they can give.
Now it was onto the more technical part of planning. Because a revolver is complex to model, it would be foolish to go in without a proper plan. Although I'd already thought about the shapes I would use briefly, I hadn't gone into enough to detail. In the first modelling post of this year, I mentioned how I wanted to go into each modelling project with a clear plan. So, to ensure I create a high quality model, I went into Photoshop to plan.
The first thing I did was find a good image. This is an image of Roland Deschain's guns from google. Since the resolution was good and the guns match the type of thing I was gathering reference for, I brought it into photoshop on a locked layer. I didn't want to accidently paint directly onto the image.
Then, on a new layer, I went over the revolver with a bright colour, and marked out how the geometric could be modelled. I started with marking out the simple things, like if there were any obvious shapes that I could generate from Maya without having to do too much work. For example, there are many pipes and cylinders the I found (like the trigger, around the screws and the chamber). I realised that most of these sections can be modelled separately, but will have to be combined and bridged together. This means I'll have to keep an eye on the number of edges I use, otherwise some components won't match up.
I also took care to consider the faces that will be needed to give certain sections the correct shape. For example, the handle isn't flat on top, as it needs to rest comfortably in a hand. to help get more of a curve protruding, I marked out where the different sections will go and how they will follow the curve evenly. These edges also match up to the metal section connecting the handle to the trigger and the barrel, so this can't be modelled first.
Once I'd planned out all the geometry, I decided to try and work out what parts are connected on the revolver, and what parts aren't. I did this from a manufacturing point of view instead of a modelling one, as most of these components I will model separately.
In the handle, there's a piece of metal tat wraps all the way around this. Depending on how the model needs to be done, this can either be modelled straight into the handle, or it can be modelled separately from the same original geometry as the handle and then separated. Until I get the brief, I will not know how the model is expected to be made.
Planning this out made me feel far more confident. However, I still feel that I need to figure out the details. For example, I'm not entirely sure which part to model first, or which part of the model should come last. I'm leaning towards modelling the indigo section first, which may then give me the scale I need for the chamber. I can then start modelling the sphere of the red section, and then start on the trigger and handle. From there, I'm not entirely sure what to model next, as the red section will be made of multiple parts that have been extruded outwards.
Most likely, I will make myself some sort of reference to use in Maya while modelling, as I don't fancy trying to guess the scale of a fantasy revolver by eye. I cannot wait to finally get the chance to model this, however, as I know it'll be a challenge, as I'm not overly confident in my 3D modelling skills. I hope the project over the course of the this year will help me to hone my modelling skills, as it's definitely useful to have.
Commentaires