Following the brief, I utilised the story board provided, I made some minor adjustments including the layout and animation of props whilst still maintaining the essence of the original.
There wasn’t too much planning to be done regarding design as majority of the props were provided to me. The only object to create is the treasure chest, whose design was mainly inspired by the chests in tutorials I watched. The stylisation of the chest had to reflect that of the rest of the scene.The purpose of this project is to create an advertisement for a game, I decided to make this advertisement with the game being a mobile app of the board game.
Here are some reference images.
As I am new to animation and glitches in the software, this was the most frustrating and difficult step in the pipeline for me. The program 3Ds Max that I used repeatedly froze at first, making it impossible to make progress. The first model of the chest that I managed to make took hours, and lots of mistakes. I used basic boxes and modified them through using the extruding and inset tools mainly to alter the boxes appearance. This chest ended up having too many flaws that created problems both at this stage and into the next stages as well, due to my amateur techniques. I then went onto make a second model, with the help and advice from friends and lecturers this model was successfully built to a good standard. I used the techniques shown in both class and in online tutorials, although the chest isn’t identical to either.
My First chest model.
The basic version of my final chest.
This was another difficult stage in the pipeline for me, as a front panel on the treasure chest didn’t relax properly, like show in the tutorial which I found easy to understand. The method I was following involves breaking corner seams and then using the relaxing tool to make them flat so that all the UV shave enough space as well as snapping matching vertices on matching panels. It is also economical as it allows more pieces to be connected (which makes it easier to tell what panel is what). The method I was trying to follow wasn’t working with this panel, even after flipping and mirroring it. I spent 3 to 6 hours trying to work out this problem, by re modelling it, using a different assortment of tools and then just sorting out the other pieces. This meant that I repeated some steps multiple, which was extremely frustrating but it I like to think of it as practice and that it helped me develop skills in some parts of the 3D people. After showing my lecturer my model it was discovered that there were duplicate vertices that were causing my problem, after their removal it worked. Although I followed a series of tutorials, I made some adjustments of my own e.g. the layout as well as having a separate UV map for the lid and base. Would have been a possibly enjoyable process if technology/ my modelling skills were better.
My UV Maps, PNG files.
It turned out to be a more time-consuming process than I thought it would be, but it was probably the most enjoyable process.
I used the mouse to do all the painting in Photoshop, which is reflected in the quality of the texture. Due to the design and UV map for the chest it meant that some parts of the tutorial I watched were irrelevant and/or a bit confusing due to the difference in the layout. The layout I created was efficient and made it easy for me to paint most surfaces. The metal was difficult particularly the lock, the faces on these UV pieces where a bit confusing but after referring back to the UV unwrap on 3Ds Max, I was able to tell what part was what.
Textures painted in Photoshop.
This step in the 3D pipeline wasn’t necessary in this project as the crabs came already rigged. The only minor thing of rigging I had to do was linking the lid and chest (the box is the parent of the lid) as well as the particle dispenser.
The pivot point of the chest lid, allowing it to open.
I began animating during the stress of UV mapping, which did have some adverse effects which did arise in some problems in scene assembly. I animated the chest falling and moving side to side as it fell and hit the ground, as well as the lid closing. I also animated 2 different crabs, each with their own little sequence. I did this using auto key. I attempted using the curve editor, but it got confusing and the results I got weren’t what was desired so I didn’t use it. Although through not using the curve editor I think it made it harder to make smooth movements and camera movements, this came down to my lack of experience. The background of palms and the grass are all animated to squish and squeeze, which I did through cloning from an original. I created a sequence of movements (on all axis’s) until there was a decent amount. I then copied and chopped it up to animate all the different grasses and palms, all their movements are therefore they are very similar to each other. But this lead to an unusual side effect in the grass, around ¾ through the length of the ad, the clones would merge back with the originals. I fixed this problem by deleting all movements after they started to merge back and then copied some of their previous movements. The coin disperser was easy to use insert although the deflection area attached to it followed the movements of the chest meaning I had problems with coins being stuck in sky and in other random places in mid-air. I then went through and adjusted the positioning of it which fixed it. Animation is one of the most exciting parts of the 3D pipeline as it brings everything to life, although it is very fiddly.
The positioning of props and cameras, depicts early in scene, the palm trees, crabs and chest in particular show how they are animated to appear in the scene.
Constructing the scene was simple. Although I had some problems in using some of the objects I had animated before scene assembly. This was because of their placement, it lead to problems in the crabs particularly, as they were too low to be seen in the scene or would behind other props. This meant that some of them, are not included in the final ad. Setting up the cameras was a bit fiddly to get it to a good standard. I watched my animation and took notes on what frame I wanted to see and finish a certain shot, e.g. crab close-up or chest closing. I then made cameras and aimed them in what felt like a very time consuming way, as I didn’t use the curve editor, as it was difficult to do what I wanted.
The images above demonstrate some of the stages in scene assembly including; a beginning of a scene and the positioning of the camera as well as its view.
At first it was a bit confusing because I was watching the wrong tutorial for the renderer I was using, as their settings are different. It ended up being very simple with Quicksilver the renderer I used, it was as simple as putting in a Standard Skylight which gave all the objects some depth, which had a peach coloured tint to match the sky. The addition of a mr omni light produced drop shadows for all the props in the scene. I did adjust the skylight to exclude the sky dome in the scene to avoid any shadows on it, as well as making the sky dome self-illuminate.
This is what the lighting looks like with just a skylight.
Here is a direct scene comparison, first one again shows the light produced by a skylight. Whilst the second picture presents mr Area omni light. The shadowing produced by the both of them is very different, both are essential.
I used Quicksilver Hardware Renderer, recommended by my lecturers, it was extremely fast and produced some great results. I struggled a bit at first as I tried to follow a tutorial for a different rendering software. Once I watched the right tutorial the process was very easy. Although I still had complications, which involved certain items that shouldn’t have been rendered, being rendered. As well as lighting issues, due to me making the lights un-renderable, as I thought if I made them renderable that the light itself (the cage/ wireframe shape) would be seen. This turned out to false and caused mainly black rendered images.
I didn’t have the need to separate any of my images into simple layer to make it easier to render, there fore I didn’t do any composting in that sense. The way that the animation from different camera angles is rendered into images is similar to compositing. It is very simple if you organise your rendered images into folders, labelled for different angles or scenes.
I used Adobe Premiere to edit my advertisement, which was easy enough to use after I sorted out an importing issue. Each image was being brought in individually, each went for 5 seconds each which made it excessively long, 10 minutes for less than 20% of the animation. If imported via “file” it enables you to import your images in a sequence, which made them play at the right speed, as well as making the layout of it neater as all the images from the same camera angle were bundled together in a a single block. From this point it was very simple. Transitioning between some shots I found difficult, to make look good without adding in tacky looking effects, so I left it plain.
References for images;
Most of the images are screenshots of my own work
However the treasure chests under Pre Production aren’t mine. The largest picture is a screen shot from a tutorial I watched by Elle. The other one is from google: