Portfolio/Production – Benjamin Corcoran

Non Narrated Production Piece

Narrated Video

Purpose of the Production Piece

The purpose of this piece is to act as a learning resource in a hypothetical scenario where digital VR takes a more prevalent role in the classroom in conventional learning institutions, (Schools, Colleges and Universities) with the focus of this piece being geared towards young key stage learners around primary school/secondary school age.

With each new generation becoming more adept to digital mediums, attention spans for young students are becoming maladapted to the conventional learning style of pen and paper learning and having to sit at a desk for hours each day, emerging technologies should be introduced to enhance teaching and to keep learners engaged, this is the goal.

The reasoning for using a historical context for this piece is more anecdotal, from my experiences and observations from my own school learning, often students I observed in my history lessons would sleep, cause trouble or simply not care when information was being conveyed, however there were certain instances where the classroom was engaged and participating when technology was coupled with the learning, rather than reading from books and papers. And so, with the points given earlier about newer generations and technology, and my own anecdotal experience, i feel it is best for learning resources to take advantage of the emerging tech that is rapidly evolving VR in particular.

The video below is an exmape of how i imagine this resource taking shape as an example, the points made in this video exprress how i feel this technology and medium can be utilised.

Production of the piece.

For the production of the emerging tech piece, i’ve utilised Blender over Maya, my reasoning being that i am more proficient with Blender than Maya due to utilising it more for hobbies and such so i have a decent grasp of its keyfarming and rendering pipeline among other skillsets.

Production Week 1 – 18th November – 24th November

My first preliminary steps for the piece is deciding on where to source texture assets for the production, for pretty much all 3D assets in the scene they will mostly be primitive shapes, roman structures such as defensive frontier forts were usually quite simple in design, mostly employing squared battlements and towers, see Hadrians wall as an example, so i can pretty much model them myself by mainly extrusions and resizing faces. For the texturing i used Quixel Megascans, i will most likely be using this for all texturing in the project.

I used a basic plane to act as the ground, i thought about subdividing and sculpting to add height variation however in Roman forts the ground was usually level within the complex anyway for construction purposes so i left it default, i made use of a dry grass texture from megascans to texture the ground, this is usually done last when everyhting is modelled but i needed the contrast to help me with placement of my meshes as everything was white.

For the fort walls themselves i actually constructed one mesh and then cloned copies around the ground mesh to form the perimeter, i think a better approach on reflection would have been to use one cube object and punch a bevel through the center and then extrude the battlements shown in the image above.

Afer the walls i constructed the gate towers for each side of the fort.

In terms of the user experience and what they wil see, i made some changes to how the viewer will experience the video, the camera will be stationary in the center of the world, and the fort will cover all angles so the user can’t see the bottom of the Hdri i wil be using later, i always kept in mind the user audience i was hypothetically making this project for, mainly young keystage learners, who may get disoriented at movement in VR while they are stationary so i changed to an idle camera.

In terms of technical challenges faced this week i had issues with the cloned meshes sharing the same texture and i didn’t remeber why until i realised that cloned objects still share the same material when cloned, so i just added a new material and deleted the old, reapplying textures in just a couple of seconds using blenders node wrangler tool.

Production Week 2 – 3rd December – 8th December

This week I worked on creating textures for the flashcards in the scene, the flashcards are how i wil convey facts to the viewer about the site at each period of the video, for sourcing the info i utilised English Heritages Site info and their sources and gathered key points about each stage of the town to inform my research, the facts i gathered i wrote down on a black canvas in photoshop before exporting as jpegs to blender, there i applied the images as materials.

From my memory of our Emerging tech module presentations, humans can only remember short chunks of information in short term memory before it is processed, so for each card i stuck to three line paragraphs, so the viewer can quickly read and understand a key fact in each card.

I also confirmed the render settings for the project, i decided on keeping the project at 1920 x 1080 for each frame, each at 1024 samples, the reasoning for this is that the total amount of frames for the video is planned to be around 5500, which is a lot of frames i will need to render in total and given most of my work is done on my home machine which has inferior raster performance, that amount will take a handful of days to complete and there is a chance i will need to re render my video is something goes wrong or something is out of place

The idea for the video is gradually info cards with images wil appear around the viewer doing a complete 360, so i’m trying to keep it somewhat light in terms of keyframed objects so i can render quickly.

Production Week 3 – 21st December – 26th December

This week i added some cut outs of roman soldiers to the scene to set dress the scene a little better, i had issues getting the transparency to work intially, the images were exported from photoshop as png, however, when i imported the image into blender and hooked it up to the alpha channel, instead of the background being masked, theentire image disappeared instead.

I looked around the blender forums and found a solution, which involved a multiply node being plugged into the alpha channel, which seemed to work in eliminating the background.

So far the scene for the first area is complete, at this stage i am now keyframing the scene and once it has rendered i will move on to the second scene. For the transitions i have used a black sphere over the camera and have keyframed the alpha value to transition from 1 to 0, each info flashcard for the scene is 10 seconds apart from each other, so the reader has time to parse what is on screen.

I did a test frame to check if the settings i made were correct, the sample rate i set at 1024 is a slight concern at 1080p but i’ve had to find a compromise for quality and time to render, issues can be retroactively accounted for should there be any problem.

Production Week 4 – 27th December – 30th December

The render for the first section is completed, it took just under 2 days to render, i plan to render the second section right after this one and then import the frames together in blenders video editing suite. I rendered in total for the first section 3600 frames, the second section is shorter so it shouldn’t take more time, while the video was rendering i worked on the second sections area to get ahead start. It is set a few centuries later and is based on the twon itself rather than the early fort.

The two blocks are the granaries of the town, I’ve started with these two as they are one of the important landmarks of the section, it’s important that i start blocking out the important landmarks of the section first before adding other details, for these buildings i’m planning to use some third party meshes for the columns at the front of the buildingsas they are an important structural feature of these buildings and i want to depict them more accurately.

The street level outside the granaries is higher than the base of the buildings due to centuries of street reapir even when the town was actiev so it was important that i depict this by raising the stanegate street level

Production Week 5 – 6th January – 12th January

This week I’ve worked on finishing the second section, in my initial plan back in november i talked about focusing on just the main stanegate street with the market, fountain and granaries.

Image Reference courtesy of Englidh Heritage

The fountain is a major part of the town as it provided water for several uses so i’d like to frame this in the center with the granaries and market on either side of it like how they are actually constructed in the survey above.

Many elements of the scene outside of the camera view i’ve left incomplete as they aren’t being rendered so it would be a waste of time to construct, the market however is incomplete by design as this building pictured above never finished construction in reality.

To help achieve a more accurate depiction i’ve based the fountain area off an illustration done by english heritage, the ground textures being changed helped quite a bit and added some much needed contrast to the scene. The viewer is planted in the center of the street looking at three directions, one towards the granaries, one direction facing the fountain and one facing right towards the entrance to the militray garrison and market. I’ve imported some stone plinths from Megascans and placed them at each side of the fountain to get a closer depiciton of the real one.

I retextured the columns i acquired from sketchfab as they came with no shaders, i added some red to give the scene more added contrast, to compound this i’ve added in the same roman soldier cut outs and also some civilians to fill out the streets to assist in the depiciton of a civilian settlement, both sourced and referenced.

I filled out the scene lastly with the info cards for the viewer to see, i created them in photoshop using the same consistent resolution, for this one i’ve inlcuded the excavation photos from the early 20th century to show how these buildings and features looked when they were first uncovered along with some of my own photographs.

Production Week 5 – 13th January – 19th January

This week i finished off by doing some post processing in adobe premiere with my video, for the frames on the last section i processed 1800 frames which is 1 minute. In adobe i added some music over the video as some ambience as the video had no audio stimuli, i wanted to add some crowd chatter audio but there wereno quality sound files that were free. I noticed from video testing that some of the cards in the sections were hard to read due to some youtube compression and quality, so i compensated for this by adding custom closed captions to the video so informtion can be understood.

When uploading using the recommended video format on the emerging tech modules page, the video had a few errors and 360 video didn’t work, the workaround i found was using youtubes guide on how to set up 360 video instead, they recommended a python script to inject the VR metadata into the video itself which worked for me.