Unravelling a bit

So this week has left me a bit behind schedule once again, and I am starting to worry about the quality of my final hand in, because whilst I think I'll get it finished, I don't think it'll be the standard I want it to be if I keep going at this pace.

Unwrapping the model was fairly simple, but I spent quite a bit of time doing it, because I think my unwrapping for texturing could certainly be improved. By this I mean, I wanted to spend more time thinking about seams, the shape of the objects I'm going to texture, and how to unwrap and pack them accordingly, to make texturing fast, efficient and clean. Right now I think a big gap in my knowledge is understanding that flow between model-unwrap and texture. I know I can do all these things individually to a relatively high standard, (or rather I think I can), however I don't think I always connect them in the best way. For example I can get my checkers even when unwrapping, but the way I pack it, or where I place my seams, means texturing is confusing and long-winded (e.g. trying to do detail and having it warped slightly).

Something I want to start doing to combat this is start making quick small assets, unwrapping and texturing them as practice. However I still have a lot of work to do that has a higher priority, so being realistic with myself, I'm going to set it as 'something that'd be nice to get done,' but only after the more pressing work is finished.

Anyway, back to what I've actually done this week. I recieved feedback on my unwrap in Visual design, because I wanted to work out what the most effficient way to pack it was, since I have a lot of very similar shapes (pipes), that are different enough from each other they can't be overlaid perfectly, I wanted to be efficient without compromising texture quality. Luckily substance is very visual ( as you can texture straight onto the model) so fixing seams will be much quicker in there. This meant that as long as I aligned all my pipes the same way (so the seams were all along the top edges for example, I wouldn't get obvious seams, and whilst there'd be some stretching (from different sized pipes being stretched to match each other in height), as it's metal with not obvious detail on it, it shouldn't affect the texture too much. I also straightened all the pipe unwraps, as it wouldn't affect the texture too much, and would make it easier to texture.

So here is my unwrap:

On reflection I should have ensured 'UV mapping' was enabled for when I used the path deform modifier. As it would have saved me time with the unwrap.

I spent more time packing this than unwrapping it. Personally I find hardsurface stuff the most stressful to pack as I know there's always a way I could have rotated this, and put this here to make it more tightly packed. However I wanted to find a happy balance between laying everything out for ease of texturing, and having it packed tightly. For example trying to keep all my toaster islands together, and all my glass pieces together etc to make things neater, and easier to keep track of. That being said, looking back onto my unwrap, I know I could have packed it tighter, but I was running out of time and I knew I needed to start baking and texturing soon, so I had to call it done.

Baking was another consideration whilst unwrapping/packing as obviously faces that have different details/AO information on them need to be seperated on the unwrap, and I needed to take into consideration the level of detail on the bake. So things that might not take up a lot of space on the model, but that have high amounts of normal detail on them, needed to be made bigger on the template.

I also realised that before I could texture my model I needed to know what colours and materials I was going to be texturing, as I hadn't defined that yet on my design.

The only materials I really needed to define were what metals or metal my pipes would be, and to consider what other parts of the design, if any, would also be metal. To do my material iterations then I used substance painter. I simply loaded pre-set materials onto my low-poly model in substance, took a screenshot and put them all into photoshop as layers on top of each other. I then used masks to try out different metal combinations, and occasionally used levels just to push things out a bit to get a sense for value too.

SO here are my final material iterations:

As my turret has a retro-esque style to it, and is loosely inspired by victorian style lab equipment, I ended up choosing the brass and copper combination, as it was not only the most popular, but the metals worked well with the retro/victorian style. As can be seen on my moodboard, brass was a metal that was used for victorian lab equipment a lot, and copper is highly conductive so it makes sense practically for a gun that runs off electric charge.

I then created a series of swatches from my bacteria petri dish image for my colour scheme, like so:

Using these swatches and my chosen material combination, I then did some quick colour iterations using a colour layer over the top of my turret. Here they are:

Funnily enough most people tended to go for the blue, orange and green colour combinations. Now portal is one of my favourite games, and I hadn't intentionally tried to do a turret for it, and at first I was hesitant to use the orange and blue colour scheme, as I was worried it would look like I was just copying an existing product. However Mike rightly pointed out that Valve doesn't own that colour combination, and that this turret is a portfolio piece. I decided then that why not tailor my turret to fit into the world of aperture science? It's a science gun, it has a retro feel to it, and colours that would slot it nicely into the old aperture laboratories. This would also make the promotional art much more fun for me, as I could really play around with the idea that it's an early aperture turret, and incorporate portal's dark humour into my 'advertising' of the gun.

So hey, the MFC Turret is now an early aperture experiment. Perhaps made on the fly to deal with an experiment gone awry?

So onto baking normals. I had to cram this into the end of the week, so I'm fully anticipating having to go back over them and edit the hell out of them/redo them, but at least I've reached the point where I can bake, and that means I can get feedback on it in Monday's lesson. Therefore I won't say that much on baking this week, as there isn't much to say really.

I had a bit of a hiccup with my low poly and high poly not aligning perfectly, not sure how as I modelled them both with the same pivot and had my low poly at 0,0,0, so they should have matched up 100%. But hey, it happened and I fixed it, but thats part of the reason why I couldn't bake properly this week.

Finally this week I rented out a DSLR and took pictures of textures round town. Unfortunetly there wasn't much that was directly right for my work, as finding brass and copper bare around town or your house is trickier than it might first seem. There weren't any clear textures for them in the museum either! However as I'm texturing in substance anyway, I wanted the photo's more for reference than anything. I wanted to understand how metal erodes, gets damaged, how wood wears etc, which is more what the photo's are for.

Over the weekend I've put together a moodboard for my promotional concept. I already know there are 3 ways I want to present the final concept; blueprint, science illustration and a beauty shot (render from z-brush probably) so thats the kind of stuff I've looked at for inspiration. Here is the moodboard:

Anyway, In Critical studies this week we critiqued a film. Whilst it wasn't personally my kind of film (far too slow) I could appreciate the beauty of the shots in it, and how well composition was used to create a visually interesting still. Something this session did highlight to me was how much I need to brush up on my knowledge of composition and my use of analytical language for art. Whilst I could make fairly insightful remarks about the stills, and knew the basics of what was going on, I didn't feel as confident with my opinions as I would have liked to. I also learnt the significance of thinking about how your work will be viewed. It's something that hadn't ever crossed my mind before, but of course if your work is going to be viewed on a 16:9 frame ration, then creating your work with that in mind is going to give you the best possible outcome.

So it's time for the action plan:

• Fix/re-bake normal map following feedback on monday

• Do some value and final colour iterations to come up with a finalised concept.

• Start and finish textures (at least get them all done, and can then spend week after editing them).

• Make glass shader in engine

• Start looking into how to make fireballs in engine, how to make the liquid move and how to make a force-field shield for the gun.

Alright so thats the plan for next week. I still have a long way to go with this turret, but even if I don't make it to the standard I want it to be for hand in, I can still work on it in my own time to make it portfolio worthy. Hopefully next week will run smoothly, but as always, there's bound to be something that goes wrong. Lets just hope it's something that won't cost too much time to fix.