Week 9 – Experiments

So as I said I made some different assets. In fact, I started making a bush and ended up working on it for quite a while.

I also scanned my own textures this time to be completely free to do whatever I like with them.

texture
My homemade cardboard texture scan

When I tried using them, the colours somehow did not seem right so I went on to edit the colour to be similar to the one before.

I continued to make a pattern for the texture because why not!

I still wanted to see how I should deal with the sides of the cardboard as they still are stretched.

However it looked even worse since the flutes completely disappear from this side. I tried researching a bit to see what solutions are out there.

image-3

study
Study sketches

I used this tutorial to learn about blending textures but since my assets are supposed to be for a game the whole thing would work differently.

While looking for a solution, this video was linked in one of the forums:

at about 12:00 minutes the person talks about putting textures on top of each other in the standard Unity shader. This seems to put the textures on the whole model and not a part of it so the individual parts would still have to be blended directly onto the texture.

Painting the different angles onto the texture came up a lot during my research so, I have tried to solve the stretching this way. After thinking it through this way makes the most sense, instead of blending it gradually as that is not really how real life cardboard works.

The problem now is to coordinate the folds along the mesh and make the texture look seamless. First step is to unwrap the sides completely, or maybe separate them?

Unwrapping the uv’s and texturing them with one texture makes this so much neater, not accurate though.

So… I ended up watching videos about how cardboard is made and meanwhile I tried to figure out what to do.

After consulting with Ryan, I felt more and more curious what would happen if I just modelled the whole cardboard inner structure and just cut it up to fit the model. This would mean perfect deformations and no need for complicated texturing. On the other hand the polygon count would go up slightly for each tree.  This method would also not be viable for the smaller assets since they will not have as much detail. I could use both texturing and modelling the flutes together, but first I need to decide whether there are any more advantages to adding the extra detail.

After all the viewer might not get that close to the asset anyway.

As a short break I made some mock up scenes that I imagine the assets to be potentially placed in.

When arranging the scene I wanted to somehow combat the symmetrical nature of the assets by placing them at different angles, which I think also makes the scene natural and more interesting.

Advertisements

Week8.5- Getting There!

This time I feel like I made some progress!

Continuing on from last week’s failures this week is quite a nice surprise for me. I did not want to do anything as I was scared it will not work out in the end. I was afraid that I have not done much, without realising that I have unconsciously been looking for the best practice in this specific asset creation.

Last week I have promised myself to push on and fix that UV map properly. Here is the result:

A partly working UV map. I chose to have planar mapping from the top as I felt that this is the way that cardboard ruffles( apparently called flutes) naturally line up. To some extent I was right in thinking so. The cardboard is cut therefore the ruffles do not change direction, instead they look deformed depending on which angle they are cut at. I also later decided to add a thin outline which I feel makes the textures stand out more and have a softer blend together with no sharp edges.

I have proceeded with the rough texture and quickly downloaded some real life cardboard textures to get a better feel of the deformations.

Here we can see that at most angles there is deformation and I think this turned out well, however at extreme angles like on most of the images below the distortion is too much.

Even If the overall idea is working, the texture looks stretched in those areas. The possible way to fix it is to have a separate texture for these extremities and try and blend them so that the texture does not look forced and inaccurate.

Since I UV mapped I wanted to see the whole tree in action once again:

I made a slightly differently shaped tree to see what the differences would be, I also made the cardboard texture repeat more to give a more accurate fit with the sides.

Overall, I feel much better about this project now that I can see the real potential look of the final assets. I noticed that this time I made the tree much quicker, I might be actually getting a hang of it.

Next steps: acquire my very own cardboard textures and play around with them.

Make  differently shaped trees or completely different asset using the techniques.

 

Week 8 – Gallery Studies

So, we were supposed to go to McManus Gallery in groups. Along with a few other people we went early and discovered that most galleries were closed and as lucky as we were none of them really interested us at the time. We went ahead to Dundee Contemporary Arts Gallery since it was close as well.

There was an exhibition on as usual. This time we could see Katy Dove’s artwork.

Many of her work has this simplistic feel to it, pretty much abstract.  The thing that I enjoyed throughout her work was the use of colour. The ones that caught my eye however were two photographs, side by side. They hung right outside in the foyer, almost hidden.

katy-dove
Untitled,2013,Digital Photographs by Katy Dove

kd1kd2kd3

What I think would be useful to do as a response is to see how the grayscale values that are bothering me actually affect the focal point( although it is not clear to me at all to begin with) . I am gong to make the whole image grey and try and pinpoint darkest and lightest areas to see whether there is any balance.

kd5

When I made the whole image black and white and increased the contrast the dark areas really stood out. In the first Image the first thing my eyes focused on was the centre however it was difficult to keep me eyes there as they drifted downwards to the left. This one is quite a diagonal composition and in the end I do not think there is a specific focal point judging from these.

The second one is dark on the left and light on the right. With opposite elements in each. Because the ribbon swirls are more visible on a dark background my eyes drifted towards that spot and to the child silhouette afterwards.

In the end I did not find any specific focal points in either image, however I thought I would be interesting to see what my thoughts are on the colour version in comparison.kd4

Interestingly, very similar results were revealed, however this time I felt the brighter/warm yellow and red colours stood out more therefore influencing where my eyes will land first, next the purples and in the end green and blue.

Overall, I think the day went well, I got to see some thought provoking images where I honestly did not expect to. It is a wonder how our view of something changes the longer we look at it.We end up analysing other artist’s work and get some insights into their thought process, see how they solved problems and see how they did not. This kind of exercise is always an interesting experience.

Week 7 – Trials and Errors( a lot of them)

This week is the “Feedback” week at University. Although not much feedback was given, I decided to push on and start Serious work.

This week has been intense, a lot of things happened including several dramatic fights with, every artist’s beloved software, Maya.

After we finally made up, we watched some tv and shared some peace food. Then, the work begun all over again. Until…

…My brain stopped working, apparently!

Here are some examples of how I made the same thing over and over and made the same exact mistakes as well.

Take 1 – Horizontal edge loops

First I made the trunk and extruded it in a shape of a cardboard intersecting. Then I extruded the leafy parts, subdivided and shaped the tree. At this stage I used the scale tool which I thought would be handy to use, however I noticed that this deformed the thickness of the individual parts and this is no good.

Oh well, it is a learning process. I guess?

Take 2 – Vertical edge loops

If horizontal loops do not work out, do vertical? Not really.

At first everything went great but when I tried adding detail( additional edge loops) everything started to get increasingly difficult to manage.

The specific shape that the tree had did not allow to place straight edge loops so there were a lot of deformations.

Take 3 – Flat “plane” approach

In between the frustration I decided to go back to the trusty technique for making symmetrical objects, aka make one side, then duplicate it. After all, this seems to be the best option.

First I made the one side and UV mapped it for later( here it turns out I forgot to map the sides). Later I duplicated and merged the objects, this gave me a better view of the tree shape.

pole_0000_14
step4- try out a texture

Here I finally pushed through and managed to make a mock up texture. I used the colours from my colour studies, the result on the model is not bad, but not great either.  Since I did not map the sides and it would take too long to do it at this stage, I will try that out next time. What I will have to do as well is tho experiment with the outlines and textures further to get a really good idea of the style and finally create more than one tree. Hopefully I will not make the same mistakes.

In the end I learned some things, here are some notes to myself:

  • A whole day in Maya is productive, until you realise you have done something wrong ages ago.
  • It is okay even if you did 🙂
  • Extruding and scaling do not work very well on symmetrical objects.
  • It is better to move the edges themselves using the move tool.
  • —>UV map EVERYTHING not just the front view <—
  • side, top, bottom views exist as well.
  • If something is wrong and you do not know what, do it again, but slightly differently
  • The more you do it the faster you do it!

That, ladies and gentlemen, is the process.

Week 6 – Colour studies

After spending a while trying to figure out topology, I realised that I am still unclear about the colour palette I want to use for my assets.

Because I was unclear I decided to do a lot of colour samples, with different tree trunk and crown combinations. As I was adjusting the colours, there were too many that I liked!

Sadly I had to narrow them down to get a better idea of matching colours. I will need to further refine them to somehow extract a colour palette.

This slideshow requires JavaScript.

Interestingly  I did not like the pure green hues, I think the hints of blue and brown(even purple) really enhance the overall look and feel of the trees.

This slideshow requires JavaScript.

When I was painting the tree trunks, I noticed a mid range contrast seems to be the most effective , whether it is lighter trunks and darker leaves or darker trunks and lighter leaves. In addition colours from the same colour group( warm/cold) often fit together better, although there are colours from edges of these groups that look pretty good with their opposites.

I also noticed that most of the combinations that I liked are quite muted, yet unusual colours, which remind me of old cartoons like The Flintstones.

The Flintstones art style is rich in textures, ranging from pure watercolour paint textures(leaves) to hand painted textures(wood). The interesting thing in there kind of watercolour works is how they decided to place the outlines. In both images here, it is mostly the man made structures that got the emphasis. In the image below the rocks are also outlined, possibly it is simply to ensure they are viewed as the foreground.

Overall, I think in this case the textures go well with the choice of colour however, I am not sure whether this approach is right in my case.

While searching for some sleeping beauty environments I found this gem, that I feel is a better example:

Since I want to focus on the environments for this project I am not going to talk about story or animation(although they are impressive). Throughout the video I can see the environment has some textural inconsistencies, however everything works together because of the effective use of colour. The palette, again, is pretty muted. Although there are parts that stand out, it is done in places that empathise and compliment the scene;which in turn makes the setting more beautiful and magical.

I feel like this example would be really useful for me to take example from since the main setting is in a forest area, very much similar to what I have in mind. It also gives off a magical feeling that I would like to achieve.

Now the next thing to do is try out the colours on the models!