Final Year Project - Week Fourteen: (1st May - 7th May)
- Ryan Harlee Jones
- May 8, 2017
- 10 min read
Final Year Project - Week Fourteen: (1st May - 7th May)
Monday:
The Monday began with me taking another look at the lighting within the scene, I had to get the scene to a reasonable standard before working on anything else, that being because the lighting couldn't be avoided. After speaking to more students about tips and tricks, I found a few settings which allowed me to manipulate the lighting differently. Un-ticking a box labelled "Use Inverse Squared Fallout" allowed me to have more control over the lights, particularly allowing me to use and edit the "Light Falloff Exponant". This helped in the sense that I was able to manually change how strong the falloff of the light was, this meant that I could edit the bulb light field that the point lights were giving off in the environment, I was able to use a combination of this new value and some lower intensity point lights to help blend in the lighting around the environment.
After long discussion with some of my friends in the labs with me, we pin pointed that the floor still needed something to help it come through with the rest of environment. The glowing symbols on the floor still seemed too flat even though rough normal maps had been manually created from the Nvida Photoshop plugin and applied. We decided that a different kind of texture applied to the glowing sections of the symbols would help draw attention to the correct points of the environment. I had personally been debating using crystal in my environment, originally I thought about using coloured crystal on the symbol images. I decided that a crystal material would be a good decision considering the colours that were already being used for the floor symbols. Firstly I decided to try and sculpt the crystal detail onto the floor panel and bring it across to the low poly asset as a normal map. Although, considering the time that was left and the other things that had to be accomplished, this approach with my skill level didn't work out. After multiple attempts I decided to take a different approach, and hand paint the relevant normal detail after the texture had been created.
Hand painting the crystal texture once I had got one of the symbols right was simple, the hard part was figuring out how to initially go about approaching it. The first few attempts were mediocre at best, and didn't look correct when viewed in the environment. Eventually after a discussion with other students in the labs, I managed to paint something which I was pleased with. The issue being caused before hand was that I was painting the light detail how I would a brick wall, although as mentioned by one of the other students, with crystal the light would be placed differently, as it is a transparent material and the light would leak through the other side of the crystal at the thinner sections of the material.
Fig 01. Crystal Floor Texture (Jones, 2017)
Tuesday:
Tuesday I got straight to work on the normal map detail for the crystal texture. This was also fairly simple, although time consuming. Considering I was mainly drawing in the detail for the pointed sections of the crystal, the process was fairly easy to implement. The normal painting graph which I had found online and the nature of the crystal, allowed me to segment the edges into the respective colour groups. Although the end result did add some extra detail that was needed, I feel as though the sections inside the crystal edges probably would have also benefited from some extra colour detail. Examples I had seen had left a lot of the sections other than edges the neutral normal map colour, although I had been focusing on very specific normal map examples, such as ones for brick walls.
Fig 02. Floor Texture Normal (Jones, 2017)
After the floor had been tweaked and the asset had been changed to better suit the environment, I started to evaluate the remainder of the scene. Since this was the stage before the trailer creation, I needed to pin point the last few things that would help polish the scene. I had been debating the creation of a fog particle effect, because the preset fog in the engine did not function the way that I wanted it too. The fog that I wanted to make use of needed to be subtle, and act not as a general fog in the room, but an atmospheric fog that would drift around the floor of the environment.
The fog was initially difficult to create, although eventually I managed to find a tutorial on the creation of a material that could be used, along with the particle effect settings which I varied from slightly. The result was close to what I was aiming for, and considering the other tasks I had to accomplish, worked for my needs.
The largest issue during the creation of the fog effect was how much processing power the computer used if I made the particle effect spawn to many particles, so there was some time spent making sure that the fog didn't cause running issues when the environment was loaded up.
Fig 03. Fog Effect In-engine (Jones, 2017)
One of the things that still needed to be accomplished was animating the monolith asset in the centre of the environment. I decided that this was one of the things that I needed to implement because it would help the scene seem alive and dynamic compared to dead and static.
The issue that I found was that with animating via Matinee which is the in engine animation panel, if I wanted to animate the individual shapes of the monolith asset, I needed to actually make them separate assets unless I wanted to animate in a different manner.
This meant that I needed to break the asset into the respective parts, and then reconfigure the uv's and textures to work correctly again.
Fig 04,05. Monolith Diffuse Examples (Jones, 2017)
The process of animating what I wanted was much easier to accomplish that I expected. A tutorial online allowed me to watch how they animated a simple moving platform in certain directions. I used the logic the tutorial gave to test if I could do the same with rotation coordinates. In the end, this worked, and allowed me to animate the objects to rotate in various directions, while also moving up and down slowly to give a hover effect. The part that I spent some time getting right was the looping of the animation, getting the animation to loop correctly meant that I had to change the interpolation of the animation keys to relevant types. To get the animation to function correctly when the play button was pressed, I also needed to just in the level blueprints make sure that when play was pressed, that the animation chosen would play from that moment, along with making sure the loop button was ticked so that it wouldn't play once and then stop.
Fig 06. Matinee Panel (Jones, 2017)
After that had been accomplished, I decided that I had to finalise what I had and move onto creating the last few pieces I needed for the hand in. I started with the stills that I had promised to create I placed camera's around the scene where I felt as though a relevant image could be taken, then using the in engine high resolution screenshot command I gathered the image to make use of. After that I then decided upon a few between the stills that would show the environment off in enough sections, while also looking interesting.
Fig 07,08. In-engine Still Image Examples (Jones, 2017)
The remainder of the day was spent placing more cameras in the scene in anticipation of creating the trailer. Recording was accomplished by using OBS to capture what happened on one of the monitors. From there, I took manual control of the camera and moved it where I wanted for the video, this allowed me to get all the sections I needed and I could edit it later. This was done with a variety of shots, and then they were all placed into a Premiere Pro file ready to be used the next day.
Wednesday:
The trailer was the focus for the Wednesday, because apart from building the remainder of the disk information, the project had almost been accomplished. I had recorded a wide variety of video angles from the day before, and spent the day syncing up the video to sections of the music. I started by listening to the music track that I had chosen the day before and thinking in my mind how I could make use of the areas I had recorded. This meant that the creation process of the trailer was not linear, although this meant that I could built around the foundation I had initially built, allowing me to fill in gaps, or debate switching pieces to make the trailer flow better overall. The trailer didn't have many issues apart from one or two of the recorded files had menu's peeking in from the corners, this was easily solved by resizing or re-recording portions of the trailer.
Fig 09. Premiere Pro File (Jones, 2017)
The creation process of the disks was the next thing on the list, the folder layout had already been given to us as a template to fill out. This meant that I just had to abide to the checklist that I had been given, along with what I had said I would hand in via the proposal. Gathering the various pieces I needed was not difficult considering I had categorised the majority of my files quite well on my personal hard drive. Although, with the Max files I was asked to hand in as prove of me creating the assets, I needed to tidy them up, since they had been used to the point of having duplicates of the same thing and random boxes and assets created for testing or measurements. I decided to send one file in which was where I did a lot of the experimenting, although I had to group files, and then hide them so that the poly count would not crash the computer it was viewed on, a read me was placed beside this as warning. But I also had a file with just the assets used in the scene lined up in their respective groups for solo viewing, this way I was sending in all the binaries I needed in respect to the assets. The other things such as textures were just grouped into folders for viewing.
Fig 10. Asset 3DS Max File (Jones, 2017)
The last thing I decided to do was the creation of the marmoset renders that I had promised of the assets used. I decided to group assets together for viewing, such as the floor pieces, lower level sections, etc. This took some time, although in the end allowed people to see what the whole environment consisted of. The renders were worthwhile also as they would be useful for my portfolio to go alongside the trailer and environment stills.
Fig 11,12. Marmoset Toolbag Render Examples (Jones, 2017)
Thursday:
The Thursday morning I burned the disks, and double checked that everything relevant was on them. I tested the zipped unreal file to make sure that it worked, and also made sure to place read me files wherever was needed to further help the marking of the project easier. After making sure the disks had the appropriate information on the front and handing in, I went home to sleep, because so that I could fit in other things I needed to accomplish such as the poster for Expotees, I had made sure to work through the night, hand in on the morning, and then refresh myself for the following day.
Friday & Saturday:
The Friday and Saturday consisted of me creating the poster for Expotees, the deadline for the poster was much sooner than the report hand-in that I had to accomplish. The poster was much more difficult to create than I originally had anticipated, although compared to accomplishing the poster in a few hours, it did mean that I spent a lot of time thinking about how the poster was going to be designed.
The first struggle was debating if the poster was going to consist of only the most recent projects stills, or a mixture of what I had accomplished. After talking to tutors and students I decided to make the poster half of each, as I wouldn't want my poster to give away everything about what I was going to show them.
The other issue was how to lay out the poster, I had a lot of trouble deciding what images complimented each other, and avoiding the poster looking like a mash of various colours. I solved this by finalising on a diagonal split like layout, with some extra shading over the top to change some of the saturation and tones in the images on display.
The last issue I had was the actual images used. I was originally set on using a selection of images, although I decided to change them because they were group projects and consisted of a lot of work that wasn't mine. I decided to take new screen shots of the group projects, showcasing some specific areas that I had an influence in, this way I wasn't essentially faking work on the poster, and if asked about them I wouldn't have the awkward answer of what I had accomplished in the images.
Fig 13. Expo Poster Image (Jones, 2017)
Summary:
The end of the project was very hectic, although I had accomplished a large amount within the last few days of the deadline. There was plenty of things I wished I could have spent time on, although when a deadline comes close, individuals have to prioritise the best things to spend time on. I feel as though I wasn't happy throughout the majority of the project with the quality of work I produced, but as the same time I surrounded myself with some of the hardest working members in the university and compared myself to them daily. By the end of the project when I had the trailer and stills finished, I came to feel pride for what I had managed which was a relief to my system after endlessly worrying about what I was accomplishing.
The thing which I am most disappointed with myself in this project, would be the lack of sculpted normal map implementation which had to be changed due to how the project progressed and time, and the fact I didn't manage to upload the finished environment to Sketchfab because of the lack of documentation on how to bring lighting, special effects and such across to the program. I hadn't originally realised that there was support for programs like Unity, but not currently for Unreal, so considering the time that was left, I couldn't spend my last few days researching how to implement so much.
Although, I'm aware that I also have a lot to be proud of, I now understand the workflow and basic of at least five different new programs, and managed to create something of a reasonable quality with all. I learnt a large amount of new applicable skills, and even now understand the difficulty of creating your own scene from scratch, along with the relevant pitfalls which can come up. The actual end product of the environment also functions the way it is meant too, and is a success considering that it could still be improved vastly.
This project was successful in my eyes, even if I judge myself against others for quality. Although, at this stage I have learnt a wide variety of things which can be used in future projects. This project especially proved that I shouldn't let myself get wrapped up in what other people are creating in comparison to mine, as we all have our different growing paces, and others might only be focusing on learning one type of thing. It's a trap that I need to be aware of, but if anything, I need to do it within reason, comparing isn't a problem as it helps me to strive to get better, but when it is ruining my health and attitude to my work, I'm letting it control me in comparison to using it as a constructive tool.
-RyanHarleeJones
Commenti