Introduction

If you’ve just picked up this book and are ready to get straight into the modeling, I’d recommend that you read the introduction first. It’s short, but explains how the book and website work together, contains useful links, and it also sets the scene and background story for our characters. Have you read it? Great! Let’s get started.

What We Are Building and Why

At the start of any project, we first need to decide what it is we’re going to create. This may seem like a bit of a simple or insignificant thing to say, but it’s not. Far too many people set out on creative projects without knowing what they’re going to do and what they want the end result to be. This can waste an awful lot of time and be quite de-motivating.

To start off, make a few notes on what it is you’re going to do and what the end result will be. In the case of our Robert and Robot tutorials, we will be creating a number of assets to build up a scene. The end result will be a render. The style will be similar to something that Pixar may create in terms of texture style, the film “Up” could be a good reference point for color and texture, though our assets will be lower detail. We will be aiming the complexity at the higher end of game assets but not quite as detailed as the motion picture assets.

Reference Materials Including Mood Boards

The next step is to start to gather reference materials and to create some mood boards. The reference material can be sketches, images found on the Internet, photos, snapshots from films, or anything else that you like the look, tone, texture, or form of.

To create a mood board, just collect all the images in a folder on your PC or print them out and either stick them to a large foam/cardboard board or arrange them digitally in Photoshop. Next, put the boards somewhere prominent next to your PC or set them as your screensaver or desktop background. Alternatively, if you have a dual screen setup, you can have your reference materials on your second screen – most professionals do this.

Brief

If this isn’t a personal piece of work, and you’re doing it for a college project or professional contract, you should have a brief. If you don’t have a brief, always ask for one, and try to ask as many questions as possible, so that you can pull one together yourself.

If you’re working to a brief, it’s extremely important to ask as many questions as you can before you start, rather than during the project. If you complete this part of the process well, it can save you a lot of time. There are a number of reasons for asking all of the questions up front – here are a few of the more important ones…

•   You’ll be able to scope to project and ensure that you have enough time and resources to deliver it on time. Always ensure you know the deadline.

•   You’ll be able to commit to dates (or not), knowing whether you should actually start the work at all. Remember to include reviews, work-in-progress deliveries, and a final delivery.

•   You will be able to determine the polygon count, texture resolution specifications, and what types of textures are required (Bump, Specular, Normal maps, etc). Never start a job without knowing these things.

•   Are different versions of the model required? Levels of detail, for example (also known as LODs). These can take at least as long to create again if four or five LODs are required. This can also be a very costly mistake if you overlook them.

•   Does anything need to be animated, rigged, or have particles included? Maybe you can only deliver the modeling part of this build.

Most importantly, if you are doing work for someone else and charging for it, agree on a price up front. This can be a fixed price, or if you are working on an hourly rate, estimate a final price, so that you don’t put the customer in a difficult position when you submit an invoice five times more than they are expecting.

Even if this is a personal project without a brief, it’s a good idea to go through a similar Q&A process as the following one to ensure you are clear with what you are doing. A lot of professionals make this mistake, so try not to get into the habit.

Below is an example of a simple brief…

Now, for the uninitiated, this may look like a pretty good brief. It tells us what we need to build, the poly count and also the price we’ll be paid for the work.

This is in fact a very poor brief as there is lots of information missing.

When presented with a brief like this we should first ask a lot of questions. Here are some of the initial ones that I would recommend in this situation and why…

Which Noddy are you referring to?

We could at this point assume that our client means Noddy from the children’s books. But it could also mean Noddy from the animated series. Are there any differences between the two? There could also be any number of other Noddy characters. Let’s assume the brief refers to the Noddy on the animated series, this leads us on to our next question…

When you say similar, what do you mean? A rough likeness, an exact copy or something else?

This should tell us whether we need to design a car to build or whether we need to obtain reference of the actual car required. In this instance, I would ask the client to provide me with reference material or a concept of the car they want built. Always ask for orthographic drawings so you can see the top, side, front, back, and maybe even underneath the vehicle.

If the client doesn’t have a sketch, or orthographics, you can always offer to produce them at an extra cost. Make sure this is added to the brief with the costs defined, clearly keeping them separate from the modeling task. You’d keep this priced separately so that your bid won’t appear more expensive than another one, and it helps everyone’s understanding of what’s involved when it’s all broken down. Once you have the orthographics or reference material, you should know exactly what you’re building, so on to the next question…

When you say 15,000 polys (short for polygons), do you mean four-sided polys or could you possibly mean triangles?

It’s quite common to get a brief stating poly count when it actually refers to the triangle count. It may seem a small point, but is well worth checking as your model will be well off the stated detail if you fail to check. No one will mind if you ask questions like this, and could save both you and the client lots of unnecessary conflict.

When would you ideally like delivery of the final asset?

Although the client has specified that this is a fixed price job, so they don’t mind how long you spend on it, they could want it in three days time, and it could be a two-week build for you to complete. If this is the case, then you’ll have to politely contact the client and say that it’s not possible for you to complete the asset in that timescale, but you could complete it in 10 days if all feedback is prompt and there are no major changes in the brief. This covers two important issues. First, you are letting the client know that you need feedback to be prompt on your submissions. You may ask that the client provides feedback within 24 hours for all submissions. If they are late, it will make you late, and you wouldn’t want to be penalized for something that isn’t your fault.

Also you are stating that you’ll only be able to complete on time if the brief doesn’t change. Briefs have a habit of changing, and goalposts move quite often on jobs like this, so you want to make sure that if there are major changes, they will be needed to be added as extra tasks, which you will estimate an additional cost for should they come up. Agreeing a day rate for your time is a good way to solve this.

What are the delivery and sign-off steps and schedule?

A good delivery plan could be completed poly model, completed textured model, then final model. If you agree to review at each point, this could save a lot of rework at the end. You could also break the $1000 fee up into three parts, to ensure that you get paid for each signed-off step if you can’t get the final delivery signed off. It would be useful to set specific delivery dates for each step too, so that the client can have someone available to review. As a client, there’s nothing worse than taking delivery of a load of outsourced work that you weren’t expecting, especially if you’re really busy.

What format will you want the final model in?

Max, Maya, Zbrush? It’s important to get this right, especially the version of software to use.

What is the budget for the texture maps?

How many texture maps of what size are required? Do we need normal maps, specular, anything else?

What scale would you like me to work to?

Again, not a major issue but one you’ll need to know, 1 unit = 1 cm maybe?

Payment terms?

Full payment on sign-off within 30 days is common. If this isn’t on your estimate, you may have difficulty getting some people to pay on time. Even with terms like this, it can be difficult.

Anyway, I think you get the idea, ask lots of questions of your client, or define exactly what you are planning to do for yourself.

Concept Art

OK, so if we’re doing a personal project and we don’t have a client brief, we’re going to need some concept art. This can be photographic reference, sketches, or orthographic illustrations. As we have the concept art for our project already, we can move on. If you don’t have clear and detailed concepts for your project, don’t start modeling. It’s a lot easier to make changes in 2D and on sketches, than it is halfway through your 3D build. A lot of artists fail to do this – don’t fall into the same trap. It’s a worthwhile investment of time to make all your decisions up front, and believe me, it really does pay off.

Making a Plan and Estimating Your Work

Now we know exactly what we want to build and we have our concept art and specifications, before we jump in with the modeling we should really make a plan. Now I realize that you probably don’t plan out your work much at this point, but you should definitely get into the habit.

The most important reason for this is so that you have a good idea of what you’re getting yourself into. All too often I see artists starting out on a project and running into trouble because they took on too much work, or they didn’t set deadlines to help them focus on specific goals or milestones leading them to completion.

Now this doesn’t have to be a mega Hansoft or Microsoft Project Plan, it can be a simple list of tasks with estimates against them in a notepad.

First of all, you need to break down the project into components. In our case, we will break down our scene into separate assets (Robot, Character, House, Workshop, Tree House, Garden assets, etc.)

For each asset, we now need to break the work down into smaller tasks, adding rough estimates for each one…

House

So for the house asset we have a rough estimate of 15 days. Continue to do this for all of the other assets in your scene. Don’t forget to include all of the details like the trash can, the fence, any plants, or shrubs, absolutely everything.

Once you have the full list of estimates, add up all the totals and it will give you an idea of the scale of the project you are about to embark on. Now, if you’re just building a single asset, you shouldn’t really have any problems, but if you are building a scene it’s essential that you go through this process.

The final part of this is to honestly work out how much time each day you can spend on this work.

Let’s say that we have eight main assets that take around 15 days each to complete, and 10 smaller models that will take 3 days each to complete. This will give us a rough total of 150 days (eight main assets totaling 130 days added to 10 smaller ones totaling 30 days).

If we are estimating a day means 8 hours, the grand total is 1200 hours. As you can see this has become a major undertaking. Especially if you consider that you may only be working on this for 3 or 4 hours a day. If this is the case, it’s going to take you 300 days without a break to complete this – almost a year!

Now, the good thing about this plan is that you are clear about the amount of work it’s going to take to finish the scene without modeling anything. So if you decide that you don’t want to invest almost a year on this, then you can redesign and de-scope to something a bit more manageable.

You might decide that you can’t see all the assets in the scene at once depending on where the camera is, so you may only need to build half of them to get the first completed render done. You can always do a few other projects and come back to this one if you want to at a later date.

You may decide to reuse assets too, so instead of building three different trees, you may decide to just scale and re-color one tree to fill the scene.

Taking this method, you’ll be able to complete more work, and not have so many unfinished projects in your portfolio. This will enable you to bulk up your portfolio with quality work, and none of your effort will be wasted.

Finally, once you have your estimates, add a column next to each one of the date you expect to have each model or component completed by. You can also set up reminders to keep you focused – five days left to complete the house as a meeting request in Microsoft Outlook, or Lotus Notes maybe.

This will keep you motivated, and it will also show you whether you are on track or not. If you’re on track or even ahead of schedule, you may want to ease off a little, but if you are behind schedule, it’s a good pointer that you either need to work harder or de-scope the work once more.

If you get good at estimating your work and keeping a record of how long assets took you to make, you’ll be a lot better at pricing your work as a freelancer, or you’ll deliver a lot more time for your lead artist or art director.

Now we have the basics of working out what you’re going to build and also a simple plan, let’s get on with building the first asset for the scene.