I’ve finally selected an image from my growing collection of stock photos (Photobash) and have started to recreate it in Blender. This is going to be a very back to basics exercise. Model, UV, texture, render. I’ve realized that many of the techniques I’ve been studying lately don’t really apply to my work. Substance Designer, for example, was created to deal with a production issue – creating many similar textures quickly to suit game or film production timetables and budgets. But my work is basically one offs, with no particular constraints.
Another issue is use of displacement maps. Plain old sculpting should do the trick. Keeping my poly count does become an issue, but hopefully my new computer, if I ever get it back and working, will solve that problem. I can still use normal maps to keep things under control.
So I expect to just use the techniques I learned form Leigh van der Byl, with texturing done in Krita instead of Photoshop
I decided to put my new pavement texture to some use. I thought the best way to spread it over a large area was to array the original 2m square plane rather than scaling it. I’ve used a couple of models (collections) from the Brutalist kit I purchased a while ago. Plus some ‘atmosphere’ of course. Trying out different HDRIs is fun. I thought the lighting in this one quite suited the Brutalist theme. I doubt there will be any interest on BA. Perhaps I should add a few people and post to ArtStation.
Different approach here, from another Udemy course. Tiles are sculpted in Blender, then a height map baked for a plane from the tiles in Knald. I tried XNormal first but it only apears to create 8 bit maps whereas Knald can create 16 bit maps which I think work better for displacement. The map is reasonable tileable across a larger area, but the course continues wiith improving it in substance designer, and no doubt creating a normal map for some finer detail.
Meanwhile I’m trying to have my new computer repaired (won’t boot) but the local technician is proving elusive.
So here’s an update. I used the baseColour, roughness and normal maps from a sample concrete material in Substance Designer, with the height map that I baked from the sculpted tiles. The grout is simply a plane with a noise texture into the normal.
I’m back to doing what I do quite well – tutorials. This one is on Substance Designer (haven’t I already done a few of those?) and this is a simple rectangular shape applied as a displacement texture to a cube in Blender. I think by the end of the course I’ll be able to make a rather nice pavement. Not sure if I can actually create a decent image out of that. I’ve already spent quite a lot of time on Substance Designer so maybe I should try to master it. At least it alleviates my frustration with concept art, at least for a while.
I’m supposed to be taking these images and converting to some scene with real elements, but problem is I like them as they are, the shapes, colours, and composition, and can’t really see anything real in the image. Guess I just lack imagination (Duh!).
Anyway, I can enjoy making abstract images but to actually advance with my making pictures that other people might be interested in I’m going back to 3D. Sort of. I’m actually doing yet another course on Substance Designer. I can see a few cool tiled walls coming up any time now.
I’m revisiting a tutorial by Scott Robertson on creating interesting environments. This one seems to have disappeared from Gnomon, but I bought it in the days when one actually downloaded the content instead of consumed it via a subscription.
So this is one fairly abstract combination of components. The idea is to look for some interesting vignette that can be interpreted as an actual scene. I should be able to make something with this.
I think this is the first post from my new computer, an MSI Prestige P100 which is sitting quite elegantly on the bench beside me. Meanwhile I continue to struggle to produce anything worthwhile. I’m revisiting Udemy courses on Krita (and other things) in order to develop some rudimentary skill as a concept artist. I’ve just been looking at making custom brushes, hence the above image using a basic cloud brush. I plan to spend most of my time on painting for a while, just because it’s a bit easier to come up with ideas in 2D than in 3D.
My new computer finally arrived today (technically yesterday, as it’s after midnight) and I’m feeling pretty good about where I am at the moment. Studio is now set up and I feel I’ve got a direction. I’m arting. A bit like that Zen master who used to go fishing but never baited his hook and never caught any fish. He just liked to go fishing.
I was doing yet another online course recently where the tutor advised attending lots of conferences to befriend concept artists (and other people). Finding good concept to realize is obviously a problem for many artists. It’s certainly my greatest challenge. I’m reviewing tutorials I bought many years ago that have more of a concept art focus. At the moment it’s Christian Lorenz Scheurer, a very charming Swiss concept artist and matte painter. I think he was the first one I encountered who focused on shapes and shape language. I think it’s the way forward for me. I’m even coming to terms with how these things are done in Krita. Finding good documentation is the challenge, as it is for most open source projects.
I ordered a Wacom Ergo from Scorptec yesterday, it arrived today. Pity 13IT aren’t so efficient. Anyway I’ve set up the Wacom on the Linux box at the end table and decided to give it a run. My current work is going to require a lot of painting. Textures (pictures mostly) for the church walls, and paintovers for the characters to make them a little more realistic. Here’s the first trial texture, not meant to be final of course. Should be a lot of fun. I’m using the NAS to transfer files from one machine to the other.
So, I rendered out the group of people as a separate image, and painted over the top, with my mouse! It’s a start. I’ll have to set up the graphics tablet again and try to do a better job. However it is proof of concept, as though the concept needs proving yet again. Maybe I should actually clothe the figures to help get the values right in context.