Winter bench

Do you ever get those projects where no matter what you do, nothing turns out right? Yeah, this one was like that.

I think I said last week that I was deciding on what to do for a Christmas competition. I finally settled on a large outdoor scene that would capture a city park in the middle of winter. Everything would be covered in snow. You would look down a path decorated with benches, trees and lamp posts, and you would see the centre piece at the end of the path: a large statue of Santa Claus majestically perched on top of a reindeer. A glimpse at a society that has traded mighty heroes for a fat guy with a full sack. 

But I'm not gonna do it.

And why is that, you ask? Because my heart is not in it. I'm trying to participate just for the sake of participating. And that's not right.

The bench scene didn't make it easier for me, either. I was trying to work iteratively: this week was meant to give me the bench, a street light and a nice tree. Seeing now that I only have the bench tells you a lot about how well that went.

In principle, there is nothing inherently difficult about a park bench. As long as you don't start late, because you keep changing your mind about which design to pick. Which is precisely what I decided to do. You'd think I was running this show for the first time. 

I was a bit nervous about getting the proportions right, because I only had a three-quarter reference, so I started with a simple blockout. From there, I split the shape into two parts: the planks (an arrayed cube) and the frame. For the later I used curves (I can't really imagine doing it any other way, the shapes are too flowy for standard box modelling). 

With the base frame finished, I converted the curves to a mesh and started sculpting the curvy bits. Not the central motif, mind you, just the top and bottom swirls. Once I finished practicing on the large set, I did a new blockout for the central ornaments, booleaned everything together and sculpted it into a more pleasing shape. I kept it separate from the main frame to have a better access to it.

Then it was time for retopology. And I seriously underestimated it. Even with RetopoFlow helping along the way, there was too much to do. (And the add-on didn't always want to help, either. I got one crash and four freezes that could only be resolved by killing the Blender process. Luckily, the autosave system didn't bail on me and I didn't lose any work.)

There is not much to say about the UV unwrapping and texturing; just the usual grunt work. What was more interesting was the ground. And when I say "interesting" I really mean "maddening". I was running out of time, as usual, so I put two planes on top of each other: one meant for the ground, the other for the snow cover. With fractal displacement added in the subdivision step, nothing too fancy. I quickly textured the mess in SP, brought it back to Blender and then proceeded to fight it for two hours. 

The more I moved it around, the more it was falling appart.  Every change in camera angle meant new issues popping up. And I changed the angle bazillion times. By the end of it I was so sick of the scene that I just wanted to delete it all. If I wasn't committed to the whole one-project-a-week thing, I probably would have.

I'm sorry to finish on such a sour note. But that's become a trend lately, hasn't it? I'm not picking my projects wisely - with the exception of the cardboard city, I can't find anything enjoyable to work on. That's definitely not sustainable, so I'll have to think hard about how to deal with it. I think I might finally turn to my large collection of concept art and have a go at something nice there. Wish me luck.