The difference between doing projects on this course and working in
an actual development studio when creating concepts is that I am working solely
on my own and not being instructed by an art director or being restricted by
any executive meddling. This is good for the creative freedom but it makes
things more difficult when it comes to finalising ideas. Other than that
though, the process is more or less the same.
The first step once the brief is set is to start collecting
inspiration and researching the subject matter. The level of importance
research into the specific details whatever you’re designing must have depends
on how realistic the concept is meant to be. Most people on this course like to
use Pinterest to collect reference but it can be quite helpful to save
reference images you think you will want to use again.
Once you’ve started to get some good ideas you can start
with silhouettes, thumbnail sketches, or both. It is important to come up with
as many designs as time will allow, but also consider that you will start to
run out of good ideas eventually . Krome Studios’ Adam Nichols says “I find the
first idea is the most generic idea and has been done before , but by
idea 6 you are rehashing ideas or running low on time”. However, it
really depends on what you are designing. I have a habit of getting stuck on
ideas fairly quickly and then regretting it later when I realise it could’ve
been better. You won’t regret your decisions if you properly weigh up all your
options so before you decide on favourites do as many sketches as quickly as
possible.
In this interview, Bioware concept artist Matt Rhodes explains that you’ll only need to do a couple of designs for something unimportant like a bartender, but more important characters can go through hundreds of designs and iterations.
If you were working in a studio you’d have shown whoever is
directing you your ideas and been given advice on what ideas to develop before
continuing development- many artists like to size their thumbnails differently
to try and highlight designs they like to the director. After doing as many
iterations as possible you would usually create presentable production artwork
with the final design. This may be in-situ or on its own depending on the brief
and purpose of the artwork. Sometimes artists choose to create a model of the
design -especially with vehicles and other mechanical subjects- and paint over
them. This of course takes extra time but allows for the final piece to be
planned properly beforehand, ensures there won’t be any errors in perspective,
and allows you to do multiple angles accurately and much faster. Sometimes the
model can even be used as a base by the 3D artists when they make the asset. If
you choose not to create a simple model, you can also make the 3D team’s life
easier by providing orthographic drawings.
Personally, if given the time I like being able to model
things first before painting them, though that is in part due to my current
lack of drawing skill. Comparing the final pieces from my vehicle project last
year I didn’t model first and the one this year which I did, I can see how
the template helped an awful lot in allowing me to accurately portray more
detail.
I think I still need more practice in going through the
design pipeline as I have a tendency to rush ahead or put too much detail in
too early. The thumbnail stage is the most important one, so it requires a lot
of deep thought and there is no room for perfectionism.
Other references-
RSI forum