Author's Notes : Technical in The Discontinuum | World Anvil

Author's Notes : Technical

About image quality   I've looked for the best way to present the images and text in this site, but every method involves some compromise. Primarily, I wanted to allow for easy navigation within a story format and settled on a standard design that puts each picture in a uniformly sized header above the associated writing, with simple forwards and backwards links that should make linear reading reasonably simple. One minor negative consequence of this choice is that the source pictures are automatically scaled to fit the uniform header format of the articles. Mostly this doesn't matter very much but its worth bearing in mind that when I originally rendered them they were intended to be displayed at slightly smaller sizes and sometimes the up scaling does reduce the quality. This is particularly noticable for portrait images, which being scaled most along the shorter horizontal dimension are increased in size the most. With the older hardware I owned at the time I created these images I had to compromise on "time to render" and size of image, and since this was well before modern high DPI displays there was no need to create them at a resolution that would have prevented this problem.   In an ideal world I could re-render them from the original files and for certain images where this was easy, I have done. Often, however, there was also some post work involved after rendering, making it a very long and tedious process to create higher resolution versions for every single image. More seriously, I don't even have all of the original source files (some were lost in a hardware accident a long time ago before my backup policies were improved) so it is now impossible to recreate them. Finally also, life is too short and I'd rather create new content than obsess over "improving" the older images, just for a quirk of World Anvil presentation.   It is worth noting, though, that if the imperfections of image upscaling bother you, you CAN still see them as nature intended at their original scale in a couple of ways. Firstly Chrome (and I expect other browsers) will allow you to open an image in a new tab from the right click context menu. Here it is shown at its true resolution. Secondly, at the end of each illustrated story below the final image, I display a gallery of thumbnails for the entire story. By clicking on any of these you are taken to a simple screen display of the art work and from there a link allows you to see the original resolution as well.   I also do have even higher resolution bit map versions for some of my art, rendered to a 3000 by 2000 pixel aspect ratio which looks pin sharp on my Surface Book HD display and is definitely the best way to see those specific ones. Even where this level of quality exists, however, I have World Anvil site constraints on maximum image size and limits to the total content size across the whole site which I need to be careful not to breach, so sadly I'm unable to share the ultra high resolution ones here.   About the software   Most of the images which predate the building of this site were originally made with two primary pieces of art creation software. To illustrate people and animals, including clothing and posing them, I have used a program called Poser. This was about the only decent solution for figures when I started but in recent years DAZ Studio offers a worthy alternative and I am experimenting with it. However, Poser still interoperates better with my favourite single piece of software, the landscape and rendering program, Vue. I can easily pull figures from Poser into a scene built in Vue where I have better control over landscape creation, lighting, material crafting, atmosphere tuning, ecosystem generation and all the other goodies Vue does so well. I love Vue because it is relatively easy to set something up quickly but there is a lot of power to do interesting things with the settings when you want to try something different. Over the years I've invested a lot of money and time in it. Unfortunately Vue has changed hands and its new owners prefer to licence the latest versions on a subscription basis. I have yet to justify the cost of this since it is very expensive for a hobbiest like myself, even at the cheaper end of the subscription model. I'm also not convinced there is anything that exciting in the newer releases, so I'm happy to work with my existing licenced version for now.   I also use Photoshop for a bit of post processing but not all that heavily, and sometimes I use a 2D image generation program called Ultra Fractal which is a great little piece of software for producing fun abstract art. It's possible to turn this output into terrains in Vue as well, which is something I've played with on more than one occasion although I could hardly describe it as part of my usual workflow.   There are a handful of images here dating back to my early days using Bryce, the piece of software that actually introduced me into the whole digital art hobby. I learned a lot from using Bryce, even if I eventually outgrew it. I have also dabbled with Terragen as a landscape generator. Artistically I think it can outstrip Vue and produces some of the most stunning photorealistic landscapes in the new version, but because it is not strong on importing figures and is harder to use than Vue, I have yet to use it as much as I think I might in future. We'll see.   Finally, I should mention my very recent discovery of AI art, just since 2022. The Art Breeder site offers an interesting approach to creating original images by merging source pictures and tweaking conceptual parameters for them using an AI model which has built an "understanding" of the visual syntax of the images and is able to make credible and intriguing derivatives. I find this fascinating. It doesn't offer anything like the same degree of control over images as can be achieved with the "traditional" digital art software just discussed, but it is fast and fun and particularly well suited to generating portraits. The technique for using it is more like beachcombing, looking for a bright pebble to select from a huge ocean of solution spaces. I've used Art Breeder to add images for the main characters in "Nine Meditations in the Temple of Chromatic Enlightemnent" in my Bubble universe and for a number of character based articles in this universe as well (e.g. the Anglo-Franco Royal Family in Earth Baroque) and extensively throughout the Fey Court. Most recently of all, I have started playing with images generated from online AI tools called "Disco Diffusion" and "Night Cafe". These are AI art at its purest, with images which can be generated entirely from text prompts alone in a process which is more like commissioning art than creating it. At the time of writing there are only a handful of images like these on the site (e.g. the "paintings" of cities and aliens from the World of the Long Sleep, which have helped me to give a visual lift to a lot of short text articles) but I'll certainly use more of them in future.


Cover image: The Discontinuum by DMFW with Midjourney

Comments

Please Login in order to comment!