Tuesday, 6 May 2014

Overall I am very pleased with what I’ve created. At the beginning of my first year I would certainly not have expected to be capable of doing what I have done.

That being said, after any experience there are often regrets and things you’d change or do different.
One of the first things I regretted was not doing this project using Unity. The project demonstrates that I have a really wide range of applicable skills, something which seems to be more attractive to small developers, and a lot of small development teams want people with Unity experience.

My second regret is, essentially, that I didn’t get around to doing more. I really wanted to look into getting a puzzle implemented, but that would require skills I don’t really have.
I was making assets right up until deadline but I still feel like it’d benefit

I am really happy with my character and its animations. I had no animation experience beyond doing test rigs when I started the project. I’m not gonna say I am the animation overlord now, there’s a reason animation is a three year bachelor. But I think I’ve understood the basics, and it is definitely something I want to experiment with more.

I’m also really happy with the overall look of the level and how much I’ve learned about engine work. Though it was incredibly challenging artistically and technically, I enjoyed the process and I would be really excited to develop more skills in this direction. 

last minute script changes

First things first, this weekend was a bit nerve-wracking. I realised the script I was using, while fine on the computer I’ve been using all this time, was very invasive in that I needed to alter UTGame files for it to work. 

I got help from a lovely guy called Scott Baker. As far as I understand, he used the platformer starter kit provided by UDK and cut out the trimmings and adapted it to what I needed. If you go way back to the start of this project, you’ll see I had a go at this kit at the start of the project but found it too unwieldy to use but all those issues have been fixed.

So basically the problem with the old script was that I couldn’t get it to spawn my character mesh unless I changed something in a UTGame file. UTGame is the default game type for UDK so if I changed those files that might cause problems for someone else down the line. It used scripts to generate a player camera locked to the character and was a bit of a faff to change.  

The new script is self self contained. It also utilises archetypes to generate the camera and the character, making it possible to make some adjustments to things like angle and camera distance while in the editor.

The only downside is that the character does not seem to be locked to its path, so if it bounces against an angled collision it is pushed out of its path. But it’s a problem that’s easy to live with.  

Thursday, 1 May 2014

This screenshot pretty much sums up today's work. Fixed seaming on the ground, repainted the cave, made birch trees and light shafts

The sunbeams is basically a cone-shaped mesh with a translucent material giving off a slowly rotating emissive.

Everything else is pretty self-explanatory. Mesh, paint, export.

Wednesday, 30 April 2014

Puky toad fountain. It's a lot more obvious in real time, but water is streaming from his mouth. Will add some particle effects at the bottom as well to create splashes. The texture is basically a panning normal and a panning mask with a cubemap added. 

Tuesday, 29 April 2014

Finished the bridge yo. 

So Max had a go in my level and started building it out from where I ended it, and he really made me realise how modular my assets are and how much further I can take this thing.

So when it's this long, if you run trough the entire thing as fast as you can, it takes AN ENTIRE MINUTE.


My level ended at abouts the first big tree and then the recycling starts after that. Even if it's just the same assets over and over, I think it's really something I should keep, and maybe even build further, because it really demonstrates the flexibility of the assets.

Here's a close-up of the mineshaft. I mean, it's basically a finished scenery and no new assets made. Pretty sweet. Pretty lazy. But pretty sweet. 

Monday, 28 April 2014

Entering the final stretch, today I sat down and wrote a comprehensive list of things I need to tackle before I am ready to hand, and then started tackling top priorities.

 Everything is coming together pretty nicely, except my eyes. They are feeling pretty square and computery.

Today I made a water shader and a tadpole particle effect. The water shader is basically just reflections distorted by a normal. The tadpoles is an animated particle effect (meaning the texture has all the frames and the uv-space flicks from one part of the texture to the other very fast, like a flipbook animation), so it looks like the tadpoles wiggle along. You can't see it in the screenshot and indeed it's so hard to see in game I sorta wonder why I did it, and I'm thinking about cranking down the water distortion a fair bit for that reason.

I hadn't done any particle stuff before but it was pretty simple to figure out.

Wednesday, 23 April 2014

My level has glowing mushrooms and a giant toad and to be honest that's everything a level will ever need.

It's mostly about tying up loose ends and getting those last thingies in there. I'd anticipated being stressed out of my wits by this point but to be honest I am feeling very relaxed about everything. Maybe I exhausted my stress supply over these three years. Goodbye stress supply.

Monday, 14 April 2014

Rigged the toad today. Instead of mashing up a bip and modifying it to fit a quadruped, I opted for making a really simple custom bone rig. A bip wouldn't allow me to do the animations I need, so it seemed like a good option. As far as rigs go, it's probably not great, but it does what it needs me to do and that's why I made it.

As you can see, she has also acquired some textures.  

It has a bone in its throat which can be inflated to make a croaking animation, and eyes which can move in the sockets.

Thursday, 10 April 2014

kissy toad

Just under 3 000 tris, this mesh is pretty dense, but it seemed reasonable for the size it is displayed at in game. I put a lot of geometry into the throat area, I want to try making an animation where the toad puffs out the throat.

Wednesday, 9 April 2014

Monday, 7 April 2014

Jumping, thumping, shout at something

I sorted out the jump animations today, they are pretty simple. It's basically a set of very short animations which blend according to the jump. The double jumps are just the same set repeated again, but if I have time it's be nice to make an alternate sequence for those as well.

I also had another look at the running animation. I think I've got the pacing down regarding the movement of the feet down; it looks like they are actually kicking off and moving across the ground as opposed to the previous rollerblading-like motion.

There's still something not quite right about it, it looks more like a powerwalk than an ambling jog trough the forest, but obviously I have no experience in this and I find it hard to pin down exactly what's wrong. It's something to do with the arms, will look at it closer tomorrow.

Friday, 4 April 2014

I hope you like green

Had presentations today, received a lot of very useful feedback. After it, I decided to grab the backdrop of an old painting and quickly make it into an alpha mask, to test the parallax backdrops. 

These alphas are awful and have a terrible fringing-problem, but it gets the idea across and even just this quick test just adds so much atmosphere to the level. 

I dropped some point lights in between them as well, I really wanted the effect of light shining trough the forest and wanted to see if it was feasible. It needs work but it looks neat. 

Game screen shot:

Also aren't those trees the cutest god damned fir trees you've ever seen?

Thursday, 3 April 2014

Time to blog a bit!

I've finished the player area landscape bits. Regular landscaping tools wouldn't have done the trick, so it's all modeled geometry.

There are some obvious seaming-issues to address; I textured the bit that goes into the grotto before I'd come to a solution for all of it.

Basically it's all one big piece of geometry that's been cut apart and then reassembled as separate meshes in engine, to make texturing easier and so that the engine won't render the entire object when only a fraction of it is visible.

Now is a good time to talk about my texturing process too. Initially I wanted to texture using watercolours, but with some more thought realised it seemed unfeasible within the time frame, and it wouldn't provide me with the richness of colour which I wanted. Instead, I purchased a set of brushes for photoshop which emulate the look of watercolours really well.

They are the Ultimate Real Watercolors by Kyle T Webster and they are really amazing. Nothing beats the real thing, but as far as emulating it goes, this set does it really well.

Friday, 21 March 2014

Due to the design of the level, lighting it has been quite difficult. It's supposed to look closed, but it obviously isn't physically so. I kept postponing finding solutions for it, but I did some steps towards it today.

I've taken out the half lambert shader which I've been using; it worked well at certain angles, but it didn't fit my purpose. The dominant directional light is top down, and the underground has been lit with some warm, soft spotlights.

Wednesday, 12 March 2014

you can tell by the way I use my walk; I'm a woman's man, no time to talk

I did a walk animation I was finally happy with, and plugged it into UDK. It slides a lot and moves way too slow and tilts over too much. ): (I dunno why this gif is so slow, it loops much faster in game)

Also some more assets woohoo.

Tuesday, 4 March 2014

Building building building.

I've gone away from the texturing with watercolour-thing; once the novelty wore off it proved to be very messy and very time-consuming and I am already pressed for time, and the sort of colours and look I want is just not even attainable with traditional watercolour.

I've been trying to emulate a sort of watercoloury look and texture with my digital texturing and I am happy with the look, but again. You can't replace the real thing. My main worry is that this strays too far from what I described initially in my specification document.

Progress feels really slow right now.

Monday, 24 February 2014

Shader day

Today has been all about shaders for me. Classmates started fiddling with it and it seemed fun and like something that would be good to sort out before I delved too deep into building. I'm glad I did it because I came to results that look so much nicer than standard UDK materials, but I might have to consider a different approach to how I build my models. 

The major failing in standard UDK materials is that the colours are really just black, and it's not very nice. 

I tried on a cell shading material for size, it looks cool but it won't work for my purpose. 

This was a Team Fortress style half lambert shader, it's pretty nice, but the shadows are so hard. Maybe there's a way to fix that. I don't know enough about it.

Mitch set up this shader for Shay expanding on a different lambert shader we had. Shay then showed me how to go about doing it and here we are.

I like this a lot. It lets me control a lot of parameters like the hardness and colour of the rim lights, diffuse intensity, and most importantly it lets me control the shadow colour. 

I will need normals for this to work proper so the idea about leaving those out is pretty much a no go. It might also be good to make the grass as a separate object and make it separate from the rock, in order to adapt material properties so they'll look more right. 

Friday, 21 February 2014

I wasted a lot of today trying to clean up files and scripts. The package I was working from was a test package and was named as such, and figuring I was starting to build the level proper, that wasn't really applicable. So I moved my assets over to a new package more appropriately named, deleted assets no longer needed, and so on.

Obviously this messed up essential custom scripts, especially in regards to the player pawn, so I had to go trough all the affected scripts and change file paths. Then I tried to tidy up the scripts a bit so that it wouldn't be depending on so many scripts outside my game folder, but I ended up getting nowhere and had to revert back to square one. For some reason I can't get my own scripts to load my mesh, even when I follow tutorials which tell me how to do so. It works fine as it is, I just thought it would be nicer if it was cleaner set up, and mistakenly thought it would be an easy thing to do.

Anyway. Much time wasted and no progress made.

Also I made a rock.
I think my next step should be getting my own skybox in; I can't really get a proper sense of the colours until then.

Wednesday, 19 February 2014

I slapped some colour on my whitebox today, mostly to test if the colours would work with my character and to see how the light works with my general colour schemes. 

I am pretty pleased with it, it's definitely something I can take further. 

Also finally finished my diffuse map. I've been picking away at it in bits and pieces over the past week, glad to see it done. It's not very detailed, the character is very small on the screen after all. I'm not sure it'll need normals, it has all the detailing in the diffuse. It'll probably need specs to control that gloss UDK slaps on it though. Ew. 

Tuesday, 18 February 2014

whitebox iteration 2

I spent a bit of today building a new whitebox according to the concept I painted last week, so that I could test the colours on it with the diffuse of the character. As I was testing it in UDK before putting colours on, I realised my scale is completely out of whack and started scaling and and tweaking and away went the day...

Now what I've got is a fourth the size of the corresponding area in the old whitebox, but it works much better scale-wise and there's a much better sense of location and where you are and such. I'll have to give the rest of the environment the same treatment, but it's not as important since those parts are mostly quite straight forward and have little verticality to them. 

I'll be the first to admit that last week was incredibly unproductive. I was threading water a lot, I have a lot of tasks that need finishing and I have a hard time settling down and giving any one task the attention it needs, and I was really just generally a bit burnt out with it all, I think. I'll just trudge on and try to break down and select my tasks better from here on. Better planning.

Tuesday, 11 February 2014

I got bad with the blogging last week, as I felt (and still feel) there hasn't been much to blog about. But the longer I let it lay the worse picking it up the habit will be.

I've been working on the diffuse a bit. I'm using Photoshop and 3D Coat and switching back and forth between them. I've never proper used 3D coat before, but it's been very easy to get a handle on it. Especially the ability to seamlessly toss the texture back and forth between 3D Coat and Photoshop with no faff at all is just fantastic and super useful.

3D coat also has a functionality where you can export a screenshot from 3D Coat to Photoshop, paint on top of the screenshot in Photoshop, and then 3D Coat will take the screenshot and project your painting onto the model. I found it really helped me in the initial stage, especially for the face.

I had an initial pass at attempting to texture the character with watercolours, but it didn't work out very well. I will attempt emulating it digitally instead, once I've got the colours down.

I'm going to start building the tube station, so I've been trying to plan it, working on top of the whitebox. I realise that some spaces in the whitebox are just way too vast, like that cave, and I'm going to adjust it a bit. 

The main reason for my silence is that I'm working on a 16 page comic for a western anthology to be published back in my home country this summer. I needed to get two pages done by today and it sucked up a lot of my time, but from here on, I should be able to combine working on the two a lot more effectively. Here are a couple of panels.

- - - 

Wednesday, 5 February 2014

the same picture over and over

After the feedback from Monday I've been working on trying to put together an image that describes what I want in terms of colour and layering in my level. This is very static, but it served my purpose well enough.

So I did this first but I wasn't entirely pleased with it. 

Receiving some feedback from Jolyon, we established that the vertical trees-thing I was trying to adopt from illustrators like Bauer is just not going to work very well. The character is also not sticking out enough against the backdrop.

He also told me to have a look at some concepts by Eran Hilleli for a side-scrolling game named Lady of the Sword, where depth is handled quite cleverly.

I grayscaled the concepts so that I'd have a better idea of what's going on without being tricked by hues.

The foreground elements have the highest amount of contrast, with elements having less and less contrast the further away they are. Backround elements further away are simply block colours only a few values lower than the sky colour. It's a neat way of achieving atmospheric perspective.

The middle ground where the player walks seems to be brighter than the fore- and background elements immediately close to it as well, especially where there are platforming elements. It's obviously important for the player to quickly know where to jump to.

Breaking this down really helped me get a grip on what was wrong with my own values. Here's the image completely desaturated. There's not a lot of variety in the foreground- and player path areas.

Especially the foreground is just looking really flat here.

I did a second pass, upping the contrast in the foreground grass layer and the path while desaturating all the tree layers. That's just a lot more depth already. 

Looking at this and being pretty pleased with it, I've laid down some guidelines for me to continue into further concepting. 

This is working a lot better in terms of depth. It doesn't address the composition-issues with the trees.

Even with just some really quick, flat colours overlayed, it works a lot better than my first pass.

One fun observation: When I designed the character I consciously tried to avoid gendering it because I wanted the player to sort of be able to identify regardless of gender. A thing I've noticed is that men talking about the character tend to refer to it as he, and girls tend to refer to it as she. It's not true 100% of the time but it happens often enough. I think that's really neat. 

Other than thaaat I've been working a bit on the character's diffuse. To be honest I feel like the past two days have been a bit unproductive and I'm not at my ideal level of activity. Boo hiss.