Showing posts with label Research. Show all posts
Showing posts with label Research. Show all posts

Thursday, 1 May 2014

Modelling Research

As part of our project brief, Lothar asked us to look into a variety of tools available to us, as well as look into the various techniques used to make high-quality models.

Tileable Textures
Making a tileable texture is not only about making an image which can fit next to itself seamlessly. It also means making textures which don't have any strong defining features, which when tiled, will make the effect obvious. What Doylle suggests is that your photos be taken as a 90 degree angle (or as close as possible), but seeing as I'm going to photograph my textures myself, I shouldn't have an issue with that.

With photograph in hand, I can then use the Offset filter in Photoshop (or do it manually in other programs) to cut the image into 4 pieces, tying each corner to the opposite one on the canvas. Lothar suggested using the clone tool to meld the seams together, which now form a cross in the centre of the photograph. On top of that, however, I found that using the Dodge and Burn tools on a greyscale version of my texture can help to eliminate any extra brightness or darkness which would prevent the texture from tiling seamlessly.

Modular Meshes
A modular mesh is a section of your asset which can be repeated over and over, and is usually coupled with a tileable texture. By making several parts of the asset modular, it reduces the time needed to make the asset, as well as file size if the meshes are to share the same textures. On top of that, if you unwrap the model's UVs before duplicating it, each part will share exactly the same UV map, making texturing much simpler.

The aspects that Thiago Klafke highlight when creating tileable textures are ease of use, composition, shapes and form, and repetition. If the model is too difficult to modularise well, then its likely not suitable. If the composition, shapes or form are too memorable, the model looks lazy. If the repetitions are too obvious, it looks clumsy. When making my Castle, I will keep these aspects in mind, and together will tileable textures, will hopefully make a model I'm happy with, with the aspects I''ve described.

Normal Maps
A normal map is used to make a flat piece of geometry seem high-poly. It uses the RGB channels to give a vector to the plane it is placed on, giving the face fake detail. When viewed from the side, though, the shape and silhouette of the object appears as normal. Lothar suggested using CrazyBump to make normal maps for our models, so I plan to look into the software more. 

Terrain Generation
There are a variety of ways which I could make the terrain for my model. Maya's capabilities include terrain, and I could use the Sculpt Geometry Tool on a flat plane to sculpt mountains and trenches, projecting a texture from above onto the shaped plane. Alternatively, I could also make use of the UDK Terrain generator, which allows for additional effects such as realistic water and simple collision. Online, some recommended tools such as Planetside's Terragen or World Machine, and they look impressive, however they aren't free (or are very limited when free), and so I will most likely experiment in UDK without plugins and Maya to make my terrain.

Monday, 24 March 2014

Castle Research


Beginning with ht efirst thought that came into my head, I started with a broad search of Medieval castles, before narrowing down to English castles. In particular I'm fond of those with a stronger geometry instead of castles deaturing a lot of curves, and so I moved onto looking at more exotic castles with strong corners and straight walls, particularly Morocco.


The design of the battlements of the Morrocan castles and the Red Fort in Delhi particularly interested me, as well as their use of mostly flat surfaces, decorated with different shades of bricks.


Continuing from the Red Fort, I searched for Indian Castles in general, instead of specifying the Red Fort. The Castle in the cliff particularly stood out to me, and made me consider the idea of using the terrain more creatively than I had originally thought. Having previously studied Angkor Wat's architecture, I returned to the ruin also. The use of large towers in the centre of the buildings, as opposed to the outer walls, is an aspect that could be explored further.


I briefly looked at Scandinavia for another different style of architecture, and noted a lot of foliage in the images I found, something that could be used to add detail to the walls of my design.

Looking into games with interesting Architecture I came across the concept art and some screenshots of the Assassin's Creed games, which feature several designs that I would like to include, such as the idea of using a tiered landscape to add a more natural verticality to my castle.


The Moroccan architecture is a theme that I very much would like to follow, and so I decided to further research their designs for walls and doors, discovering their penchant for grant interiors as well as exteriors, something that could be used to add more detail to my finished product. Tunisian designs also stood out to me, as they largely appear as similar to that of Morocco, instead with a wider range of colours used in the abundance of arches and square buildings.

Friday, 24 January 2014

7 Virtues: Brainstorming & Research


I decided to create my character around Charity, as I saw a lot of opportunity for exploration. Particularly, the themes of sacrifice and salvation stood out to me as strong themes, and the idea of nobility was interesting because of the seeming contradiction of charity and lavish wealth. Although in the context nobility refers to the actions of the charitable, it was an aspect that warranted noting.


After brainstorming ideas, I decided to gather images for reference and inspiration, starting from the theme of guardian, something which I thought combined the ideas of salvation and sacrifice - being ready to give your life to give someone salvation. There are obvious religious parallels in that, however I decided to branch elsewhere, looking at nature and the charity that it provides to all inhabitants of the world, even at its own detriment (deforestation, landfill, etc.)

When researching natural clothes for clothing ideas, I came across Nicole Dextras' "Little Green Dress" project, in which Nicole crafted 21 dressed out of only plants and berries to highlight the issue of sustainability in the fashion industry. The range of colours and textures were particularly striking to me, and I plan on using them as inspiration for my own character.

When taking a closer look at Dextras' dresses, many of the berries and leaves have already started to decay, given the time taken to make the dresses. This brought my back to the idea of sacrifice through charity, and that nature, through our sustenance, is being slowly eroded. Alternatively, I could also pursue the idea of man-made products polluting the environment, perhaps incorporating imagery of landfill, deforestation, and the like into my character's clothes or body.

Tuesday, 19 November 2013

Contextual Studies Research

The following is the list of quotes and ideas that I thought relevant for my essay. The primary bullet points are notes made during lectures or reading, and secondary bullet points are notes made with relevance to my essay.

Tuesday, 15 October 2013

Research: Edo Japan

For my first project in Concept Visualisation; Character Creation, I have been given the period of Edo Japan to study. From this period, and based on the Tinder-Box, I will create at least 50 character silhouettes, from which I will create value studies and colour variations. Based on these, I will then design a model sheet (side, front, back, etc. views), character sheets, and finally, a finished character piece, complete with background.

Below is my first pages of research, in which I look at clothing, hairstyles, accessories, and weapons. At this point, I was still unsure about which of the characters of Hans Christian-Andersen's story to base my designs on, and so the research is more generalised.






At this point, I decided to work on the witch as my character, and so I looked more closely at Geishas, headwear, kimono patterns and instruments.