Devblog 14: Month Two

It has been two months of full time development.  Ten months remain.  The clock is ticking, but the light at the end of the tunnel grows brighter!  Right now I’m deep in vector mathematics.  Which is about as fun as it sounds.  Regardless, code integration is going well.  In my last two jobs as a software professional I could never give decent estimates, so I’m not going to try here.

Instead, I’m going to talk about art design and process.  So, how does one go about moving ideas from brain to screen?  The first step is finding an artist.  In my case, with limited budget, I can’t afford to employ an artist with the skills I need.  I can however afford to pay a freelancer.  This will still be very expensive, but it limits potential cost and risk.

There are many websites which allow you to meet freelance artists.  Generic spaces, like Fiverr, were recommended by a few people.  However, I decided upon specialist websites, starting with Polycount.  Polycount’s “Artists Looking for Work” forum allows you to post an advert, and then scrutinise each message based upon the artist’s portfolio.  Later I used ArtStation, which requires you to search for artists and browse their often intimidating portfolios, but you can specify search criteria which is really helpful.

I found an artist on Polycount whose portfolio I liked the most.  This was from about twenty applicants who messaged me.  It’s very important to accurately describe what you need, from the style (preferably with examples) to the quantity of work, price range, etc.  It’s also important to find an artist with a portfolio demonstrating the kind of things you want.  One applicant had the most excellent portfolio of human faces… when I’d asked for robots and animals.

Where they live is another consideration, as this can influence cost either way, and legal concerns.  The artist in question was Malaysian, and I used Transferwise to pay them.  Different people will ask for different ways to be paid, whichever is most convenient for them.

Concept art is a gradual process, and so will be expensive.  It requires iterations of designs.  First, you must describe a clear vision of what you want, and from this the artist creates an estimate.  The more professional, the more likely they are to ask for exchange of contracts and some of the cost up front.  Contracts are good, but only if they’re from a jurisdiction accessible to your lawyer.  Asking for references before signing contract is also a good idea.

After the first draft is done, the problem is trying to figure out and then describe changes.  You must have a strong idea from the start, the last thing you and the artist want is badly defined designs, as this just stresses people out and wastes money.  If this isn’t possible, make it clear that you really don’t know what something should look like, but are happy to take the time to explore different ideas.  Professionals will be able to provide good ideas, and you need to do your homework: at the very least to know what you absolutely don’t want.

The first artist developed rough designs.  They were good, and really helpful.  However, they did not have the skill required for the final rendering, which must have absolutely no visual ambiguity.  A good question worth considering is what makes good art, but poor concept art?  When a digital artist creates something in three dimensions, and there’s visual ambiguity (for example, only the front is shown) this is a big problem.

So I had to find a new (and more expensive) artist.  I used ArtStation, messaging artists from various places who I thought may be appropriate.  In the end I paid three artists to do exploration sketches, and from this had something to compare.  Though I already had a favourite, it was best to be absolutely sure.  It’s basically a job interview at this point, you want to see how they work; not only the quality of the art, but how long it takes and how well you communicate with each other.

After I was sure I had found the right person, we went about exchanging contract and having Skype calls to clarify things.  There is a lot of work to do, but we both agreed to take it slow, and began with just one batch of concepts.  This works really well, I’m in no hurry to acquire art assets.  I can use programmer art (read: cube soldiers) until the art is done!  Everyone appreciates being given the space and time to do a job properly, so I was keen to avoid any undue pressure; which is especially bad for creative work.  If the contractor needs time, give them time.  Delays are sometimes inevitable.

The conclusion of all this is that concept art for units is almost complete.  Next step is final rendering, which will be done gradually to make sure everything is perfect.  I also need to request concept work for weapons and environments.  That will come later.

In the meantime, I am more confident than ever in the code and art.  Existential terror is at an all time low.  And I’m still excited that I will eventually have something to show you!  I suspect this will suddenly arrive all at once, when programming and artistic labours finally meet.

Devblog 13: Month One

It has been a month since full time work began on 13 January.  Finally, light flickers at the end of the tunnel of technical challenges.  A lot has happened in this time, but it has been overwhelmingly technical, so I won’t bore you with details.  I always find time spent trying to understand someone else’s code to be one of the more frustrating moments in software development.  Hours spent reading doesn’t feel productive, when fingers could be tapping away at keys.  Especially if I’m feeling overwhelmed by the scale of the task at hand.

There is new concept art, but it’s not quite ready to show on the development blog.  I didn’t think I’d be quite so stressed doing what I’ve always wanted to do, but mercifully this isn’t constant.  At least, as the technical solution slowly takes form there’s less anxiety hanging around.  During the early stages of any ambitious or ambiguous work self doubt is at its most distracting.  Especially when you aren’t working with anyone most days, so having friends to talk to who are programmers and can give pro tips is indispensable.

Month Two should certainly feel busier, as there’s less mystery about what is left to do to achieve a working deterministic (multiplayable) system.  Preparing Lockstep and Pathfinding libraries for integration has been no joke.  Integration will consume the coming month, along with finalising character and environment concept sketches.

After all this time, the game design feels less flimsy, it has been years of arguing with myself.  Though, there are still some questions to resolve, but the core of what will make this game unique and fun is absolutely obvious.  Ideas are like wine, best left to age in the cellar of the mind.

Devblog 12: Rebirth

This week has been the first week of full time game development in a long time.  It still feels weird, but little steps have been made on every front, from code, to concept art, to business admin.

There won’t be anything to show for a while yet, but now I am lucky enough to be able to work full time on the project and acquire freelance artists.

This is a momentous occasion for me, and so I’m feeling somewhere between anxiety, excitement, and terror.  The project has been a labour of love, on and off, for a few years now.  Finally, I have the resources to be able to realise it.  Can’t wait to have something more to share with you!