paper-practice4

DevLog 2

//Building the Look
The concept was in itself more proven now, the palace from the test was very memorable and striking, and most importantly the way there seemed baseline interesting, fun and motivating. But, since i stripped away most surroundings, while photogrammetry has a distinct look, it was still very barebones looking, so next was focussing on both art direction and more overarching design of the entire process, to turn it into a more concise start-to-finish experience.

First i explored possibilities with shaders again, for both the rooms, the objects and the background, toying around in shadergraph, with stack-treating the textures, and much more.

From this came firstly a particle system, idly floating dust-like small white particles in the air throughout the whole space, adding a slight bit of constant movement, also helping depth perception within the space.

Also, when cycling through the rooms with the RoomSpwner, i wanted the rooms to be slightly transparent, so i developed a shader that makes the rooms before placed look slightly lighter, see-through and rasterized (It is a dither-shader i made in Unitys Shadergraph).

Another Shader i worked on was a posterization shader, to make all of the photogrammetry objects more stylized, it takes the texture colours and splits them into steps, reducing the total number and detail of colours, much like the posterization effect in Photoshop. This shader also has an option to adjust the saturation and change the overall colour tint of the texture, further reducing the colours. This created a much more precise and graphical look, which i liked.

This led into the realization that i’d have to look at the style more radical in terms of graphic design and less in terms of game design/level design, to create more striking images. To achieve that, i was looking to work with strong single colors and gradients, to counteract the realistic and muted look of the 3D scanned Objects. So i started working on adding gradient support to the shader, but i soon found a similar, already polished shader called “Quibli” in the Unity Asset Store with a lot more options than i would’ve been able to put in by myself in the given time, and imported it. It offers many options to fine tune how the gradient maps over the object. Trying to reduce the amount of colours as much as possible, i settled after some testing on Red & Blue for the strong contrast and the simplicity of them being prime colours, and white text still being well readable atop both.

//Revamping Palace Structure & Tutorial

To make the palace and the space where the palace would be placed more striking, and -again- graphical, i wanted everything to be visible in one image, to be able to look at the palace in the end and have a very readable overall structure. This meant changing the structure from round-trip to a linear journey. The round trip also did not provide a distinct advantage, since at the end one can easily move within the palace mentally either way in any direction, this interconnectedness wasn’t necessary.

Remembering the ancient greek roots of the history of mnemonics, i placed the inbetween-platforms for the objects on huge pillars. For the background, i used a scan of a mountain/cliff-face to provide more of a place-narrative to the palace. Going on a journey up the mountain!

To decrease the necessity for outside guidance in VR, a tutorial area leading up to the beginning of the palace construction zone is a dedicated space for learning the basic controls. Tutorial poles –basically in world-space text boxes– teach the users to look around, move, move objects and use the RoomSpawn device. The tutorial texts also extend to palace construction, and tell the users after each platform, each step of the way what to do next. Players might still have questions that could be best and most efficiently answered by a facilitator, but this was an attempt to cross some of the way to self-sufficiency from outside help.

//Mnemo As a Process: The Website & Memory IDs

To tie all parts of the method together, i was thinking about ways to connect all the phases better. So far, everything only worked as an experience led and conducted by someone else, could i reshape it in a way that would allow for it to also work self-conducted?

For that, building a website was a logical step, the first phase of memory identification could be a form, the users could fill in, with helpful hints and questions on how to identify. This data would then be taken over to phase 2, which would be the Mnemo VR application, possibly in browser even, where they would build the palace and anchor the memories. The palace and order of rooms, objects and memories would then be automatically sent back to the website to generate a recall sheet for recall training afterwards.

Making use of these connections, i took some time to also mirror the memory IDs (the short phrases to identify the memories by) within the app: They now were small spheres, with the text of the memory ID inside, that could be handled like objects, and were also integrated into the save system. These memory spheres were spawned into the world, after the users would hit a button at the top-most pillar, after having fully constructed and filled their palace. That would officially start the second phase, the anchoring, where now the memory spheres could be placed besides the objects for connecting, without someone form outside needing to read them out.

I knew that i could not integrate these data transfers between the web and VR parts in a working manner within the timeframe (or optimize the unity project for WebVR even!), but since this format would make sense, i started building a website as a vision, a demonstration of what the entire process would look like.

For the site i am using a rather unknown small website building editor called mmm.page, that offers super simple drag and drop “what you see is what you get” tools. The way of working with it feels rather like collaging than classical webdesign and allows for very quick and straight forward setup.

The planning on the website structure was the following:

Frontpage:

Big catchy opener, typo arrangement and collage, potentially video
scrolling down: Pitch and quick explainer on what and for whom Mnemo is in very direct language. Link to starting/trying out the project by oneself
Link to further research (this Thesis)

Mnemo Process Page (Phase 1):