This week we’re doing the final work on our walk cycle before moving onto our body mechanics shot.
Feedback from George:
I’ve been a bit slow with progress the past couple of weeks, but I’ve managed to make a start on animating. I’ve also found a few more issues with the rig that I’ve been able to fix, one by using the Discord server the creator of the rigs made for users. I also added a Bugs/Issues page to my Google Sheet that I’m using to keep track of the project, mostly so I can note down issues as they come up to look into later and keep track of what I’ve done to fix issues that may come up again and need the same fix. This sheet did prove helpful when I had to redo all of the changes I made to the original rig file due to part of the rig breaking in one of the early rendition scenes. I didn’t notice this issue until part way through animating, but was able to use the sheet I made to quickly reapply the changes without having to try and remember off the top of my head what changes I made.
This week we splined our blocked out walk cycle and started adding details in.
Feedback from George:
Feedback from George:
In Maya, I posed some of the ‘Body Mechanics Rigs‘ (created by Joe Daniels) in the subway train. I then imported these into UE5 so I could set up the cameras there, and export the cameras back into Maya for animating. While convoluted, I found setting up the cameras in Maya less intuitive than in UE5, and I think overall I saved time as setting them up in UE5 was quick and easy when also using my storyboard sketches as reference.
I then set up each key shot (the shots I consider necessary to convey my project idea) in Maya with the Apollo rig referenced in. I had already deform wrapped the clothes to the skin of the rig and hidden the faces beneath the clothes, and saved this as a separate scene to be referenced in.
Before I got ahead of myself, I wanted to test the pipeline I had planed out to import the animation into UE5. From previous research, I decide to export the rig geometry as an alembic. I did a very rough animation for shot 0070, and exported this for the test. However for the materials on the rig, I misunderstood how these would transfer in and found that while the materials did import, they all showed as black. I put this aside to fix later, and I was then able to set up the alembic in the sequencer and see that, other than the materials, the alembic import would work as I planned.
For the materials it was a little convoluted but again, would save time later, to recreate all the materials in UE5, make sure the names of the UE5 materials matched the Shading Group node of the materials in Maya, and (while I couldn’t 100% confirm this was necessary, it works as desired) assigned the materials to the geometry faces rather than the object as a whole. I also had to make sure to assign the hidden faces a unique material that, in UE5, was transparent or “invisible”.
While I was already planning on switching out the subway train materials for my own, seeing the Apollo rig in the level gives me an even bigger urge to do so, as the contrast between realism and stylised stands out far too much to me.
This week we received feedback on our blocking animation of the side step, and learnt how to animate a walk cycle.
We will be using the Walker rig (a ball with legs) for the walk cycle so we can focus on the leading actions, and the follow-through effects of these on the rest of the lower body without worrying about the upper body for now.
Along with the animation plan, reference, and blocking, we were asked to take 3 pose sketches we’ve done this semester and pose the Bony using the sketches as reference. While doing this, we were asked to pay close attention to the weight distribution and balance. (George’s feedback shown as the sketches on top of uploaded images)
I noticed with all three poses I didn’t push the arcs enough, even when both the reference image and my sketches did so in the case of the middle image. With the left image, I think I focused too heavily on matching the pose and not enough on pushing the arcs and balancing the weight distribution, which made me completely miss how far back the dancer is on his grounded foot. For the right image, I missed how far forward the woman’s weight is in relation to her feet and, slightly so, her hips. I realised when thinking back that I was sat quite low when sketching the statue, and didn’t take this into account when studying the angles of her body. While this might have been fine for a general still life study, it might have been better for me to choose a more level eyed pose to study weight distribution from. Good to know for future studies like this!
Side step reference, plan, and blocking animation:
Feedback from George:
This week we looked at rigging in UE5. While I’m planning on prebuilt rigs for Maya and then importing the animation into UE5, this lesson was still super helpful for if I want to make my own rigs for UE5 for future projects, which is something I’d like to try in a future project.
I have done rigging in Maya before, and quickly picked up how similar the general process is to the UE5 rigging process, just with a different interface, names, and (from what I understood) being more node-based.
As previously planned, I edited my sketches into a rough animatic. This also includes all shots for my initial idea, as I wanted to have the full version so that if I do have time to add more shots the layout is already there to work from. I also started filming reference for the key shots where there will be character animation. In some of the shots (mostly those focused on the background characters after the transition to the creepy, bizarre vibe), the characters will be static and staring at the camera, so those won’t need reference. If I have time, I will probably add some camera animation in along with the 2D Animated eyes to give the shot some interest in place of character animation.
Since I’m still quite new to UE5, I wanted to quickly test how easy it would be to replace one of the textures in the subway file with my own. I had the sketch from my storyboard for a rough idea of how I wanted the final shot to look, so I went with replacing the text on the scrolling screen as a test since, if successful, this could also be something I keep for the final render.
Reference sketch:
First I located the material for the screen, then the node for the text texture. I exported the text texture, opened it in Photoshop, and replicated the texture by searching for a similar font (I settled on ‘Dogica’ by Roberto Mocci (link), and manually added the semi-transparent grid on top of the letters.
Original:
My replica:
This week we got feedback on our splined side step animation, and learnt the importance of weight distribution balance in preparation for doing our side step animation. To help us learn how to balance weight distribution, we will be posing the Bony rig using our pose sketches as reference. In general, learning weight distribution is going to be key for any weight shift animations we do in the future.
Feedback from George:
This week Serra showed us some physics in UE5. We learnt how to make objects fracture upon impact, and how to constrain objects.
I bought the Apollo and Artemis rigs and, following the creators instructions, used a wrap deformer to attach the provided outfit to the rig. I’m planning on using Apollo for the main character, and I can use Artemis for one of the background characters. I have other rigs I would be happy with as the other background characters. I did however see on the Discord server for the Apollo and Artemis rigs that the creator, Ramon Arango, is going to be releasing six more rigs soon that look really cool. Depending on the timing of the release and the cost, I might buy these as well, both for the background characters on this project and for future projects.
I also finished my storyboard, and next will be taking pictures of the sketches and editing them into an animatic. I found music I like (link), and while I’m not certain I will use it for the final edit (mostly due to timing, but I could always edit the music also), I can at least use it as a stand-in for now.
This week we were shown how to spline our blocked animation, including cleaning up both the general animation and the animation curves with the curve editor.
I uploaded my shot plan and blocking animation to syncsketch.
Feedback from George:
Last week we had to cut the lesson short a bit, so this week we finished off looking at sequencers, particularly with their uses and how to organise them.
We then moved onto materials, looking at how to make them, and how to make material instances. With materials, it is easy to make changes but can get a bit annoying when you might want to see those changes happening live in the viewport. With material instances, as long as the material it is created from is set up correctly, you can change the parameters live and it also doesn’t change the base material, making it easy to undo/reset parameters.
Serra also showed us how normal maps work, and how they can add depth and detail to objects without changing the geometry or needing to use high resolution assets.
Finally, we learnt how to blend materials together. I realised this will be very useful to achieve the aesthetic look I was hoping for (as shown on my Pinterest board screenshots from Week 2), because combining and laying different materials will probably be more successful to achieve the mixed media, textile look of the references I was inspired by.
Finding rigs:
I didn’t get a response from Hernan Ares regarding his ‘Zio’ rig, so I started looking for alternative rigs. I’m considering buying the Artemis and Apollo rigs created by Ramon Arango, who has a strong background and experience in rigging and animation. While I really like the more unique facial features on these rigs and the level of technical detail, the tracksuit outfits that come with the rigs (and need to be deform wrapped to the skin geo) are a bit too plain for my taste. I’m planning on having a detailed environment with interesting materials, so without modification the tracksuits would stand out against the louder backdrop.
Because the rigs are more on the premium side, I want to wait another week or so before buying in case I find something else. While I would be happy to spend the money for them since they will come in useful a lot in future projects also, I don’t want to buy rigs I don’t need and that could be far more complex/detailed than is necessary for this project if there are better options.
Storyboard/animatic:
I began by quickly sketching a storyboard to visualise my ideas. To maintain momentum, I prioritised speed over detail, knowing I could refine the sketches later. I’m planning to make an animatic later by photographing the storyboard, saving time on digital redrawing.
Some shots were difficult to sketch, so I imported the subway train level from UE5 to Maya, and used the ‘Body Mechanics Rigs’ (created by Joe Daniels) as placeholder characters. This let me experiment with different camera angles and character positioning a bit before adding a shot to my sketches.
I realised early on that my project’s initial scope was too much to complete in the time we have, so I decided to sketch out all my desired shots, then selected the minimum number of shots needed to convey my idea and aesthetic to move forward with. This approach allows me to prioritise the essential elements, complete the project efficiently, and potentially add back more complex shots later if I have the time to do so.
Organisation method:
To help me track all of this and to keep myself as organised as possible, I made a Google Sheet for the project that I’m planning to have several pages for, each with their own purpose to help my workflow.