Windmill Project

Modelling

 

Following last weeks research, this is the rough concept I had in my head at the start, I imagined a lonesome windmill at the top of a hill, with a small house next to it.  It’s pretty stereotypical, a wood frame around a stone brick structure, maybe in the countryside of a Norwegian country, quite cold and secluded.

From here I started modelling the house, I’m going to gloss over some of the techniques I used, as they were mentioned in the previous weekly updates, but it pretty much consisted off the multi-cut tool and a bunch of trial and error.

 


House

I started with a cube, scaled it to the rough dimensions I wanted and added some cuts. I pulled the vertices at the top inwards, to create the slanted top, where the roof would be. I then added a beam along the top for support. The beam is yet another cube, scaled and cut, manually creating the desired contour.

 

 

I added some other beams around the side, to support the structure further, I tried to imagine how it would be constructed in real life, and then I added the roof, using the same techniques mentioned previously.

 

 

 

 

 

 

 

I cleaned up my mesh, checking for any n-gons or anything else that would make life difficult. I tend to do this after completing most sections, before moving on. I attempted to use the mirror tool to create the support for the roof but during clean up this brought up a bunch of messed up geometry, so I ended up doing it manually.

 

I was pretty happy with the general shape at this point, I made some cuts along the middle to shrink the lower half a little, to give a more stylised silhouette, but this ended up very subtle. I then grouped and named all my objects, to keep it all clean and organised for when it starts to get more crowded.

 

 

I looked at some actual 3d models, for ideas on how to approach the next part.

 

 

 

 

 

 

 


Windmill

I started with a cylinder as a base, scaling it to the rough size I wanted relative to my house. Since I used the cylinder shape, to avoid mesh problems (n-gons) I had to select the top and bottom, bevel them, then use the multi-cut tool to make sure it was all polygons. (I go into more detail on this process in my previous weekly challenge posts.)

 

 

 

 

I added another cylinder with less divisions, to create a more structured look for the roof, and scaled the vertices on the top and bottom to get the desired shape, again I cleaned up any n-gons during this process.

The photos aren’t the clearest on what I did here, as I was using the orthographic views. Basically I added another cylinder, bevelled and removed the faces on the top and bottom, and put this hollowed out, flattened cylinder at the centre of the windmill, to emulate a wooden walkway. I duplicated this and scaled it down at the thinnest part of the windmill, to add a wooden support. I also modelled small supports for the walkway, using angled cuboids, removing the faces that weren’t visible.

I wanted to add further supports, wooden beams from the floor to the centre support, but I found this difficult as I had to model wood along a curve. I attempted to do this with multiple rectangular objects and snapping them together, consulting the forums on how to do this, but this resulted in bunch of mesh problems for me. I then realised I could just select the face and extrude it along the contours of the windmill. I duplicated this beam and placed it three more times around the base of the windmill.

I then started on the blades, extruding a rectangle from the roof, and creating another cylinder. I used 8 divisions for the cylinder, as I thought it would give it a more wooden and stiff shape to the structure, trying to avoid perfections so you could imagine this windmill was manmade. I took the middle beam of the blade, and scaled the middle vertices on one axis, so it looks as if the wood is going through the beam from the roof. I also added some more shape to the front of rectangle, making it look like the rotor.

I used the curve tool on two more cylinders, to create a more organic shape for the outer beams of the blade, and added 7 small cylinders connecting the three wooden cylinders. Again, I ensured it was all made up of polygons.

 

 

 

 

I duplicated the blades, and started to construct what connected the blades to the windmill’s roof. This took a few tries, as it kept coming up as having more than 4 faces on certain sides, but I eventually got it to work and I came up with this mini house-shaped structure for the top of the roof. At this point, I didn’t know if I should just keep the model simple and start my UVS, or continue to add detail.

 

 

 

 

So I watched a lot of YouTube tutorials on stylised modelling, and how to UV/Texture stylised objects, and decided I would need to add more detail before I started any of the texture work.

 

 

I added some rectangular beams to the roof, using the same methods as before. To save time I did it once and duplicated it. I divided 360 by 8 to work out how much degrees I had to rotate each rectangle by when duplicating, as the roof had 8 divisions.

I considered adding roof tiles, but I thought it would take away from the style I was going for. I added further detail to the windmill, and played with the idea of having extruding bricks for added depth, something I saw on sketchfab

 

I watched a tutorial for clean booleans to do the house’s windows, but it kept messing up my topology, so I just added more cuts with the multi-cut tool and extruded/indented the faces. I then added some more detail to the house before starting to UV.


UV / Textures

 

  

As you can see I didn’t have the best start with my UVs and texturing, I was very new to substance painter and didn’t realise you could export the whole scene, I thought it was one object at a time. So I would UV one object then work on it in substance. I spent about a day on the above, researching, running into a lot of issues with my workflow, textures and even my pc hardware at one point. This was kind of my baptism through fire, before really starting on my textures.

 

I assigned everything its own material, and started on the UVs, I’m not going to include them all in here as its quite repetitive, and this blog post is getting long. I would generally use planar projection when applicable but if it was a more complex shape, I would use automatic and then manually stitch/cut the edges.

I didn’t realise that the UV maps of objects within the same material would overlap. So I ended up having to redo (layout) them by material. I also didn’t realise I had to put them into the correct (relative to one another) orientation if I wanted the wood grain to travel the right way. So I had to redo it yet again.

 

A technique I used a lot on my textures, Inspired by a YouTube video, ‘Creating Ghibli-Style Textures in Substance Painter’ by Stylized Station. I wanted cell shaded, hand painted aesthetic to my textures. I would use fill layers with black masks, using the blur slope and blur directional filter to emulate brush strokes.

 

Baking always resulted in some weird Ambient Occlusion and I’m still not sure why, but I managed to find a work around later. I used a combination of hand painting, combining existing materials, and fill layers/procedurals to get the outcome I wanted on my textures. If the texture was similar to another, like the different woods, I would make my own smart material from the first group, and edit it to my liking.

I started to venture off from my concept art, the dark stone brick and wood didn’t translate as well as I hoped. I had to go back into maya, as I wanted the extruding stone bricks from the windmill to match the stone bricks from the texture. I also felt like the scene was missing something, so I added a ground.

 

I made a grass texture I liked using the method I mentioned previously, but it was missing something so I added stones to it. I used fill layers with black masks, the ‘Cells 4’ procedural with the balance turned right down and a blur filter to get this effect.

 

When I brought the windmill texture into Maya to further adjust my stone bricks, I kind of got carried away modelling. The grass looked empty and it was just floating, with no bottom face.

So I added some mushrooms using techniques I got from Michael and Dermott, the curve tool and the soft selection on vertices. I didn’t create a sphere, but instead made a cube and used the smooth mesh tool, something Alec mentioned a while back. I also added a door to the back of the windmill, as my sister asked how people get into it which I didn’t even consider.

      

I made some final adjustments, adding imperfections to the windmill and a door to the back, then brought it all back into substance. I managed to figure out how to enable emissive materials using the forums, allowing me to have objects emit light. I used this on the windows, as well as the mushrooms, inspired by the blue mushrooms from the first Fable game.

 

When adding the emissive materials in the texture set settings, I noticed a section to add ambient occlusion, allowing me to fix the issue I had with the back of my windmill, setting the ambient occlusion I made to replace the baked one.

I exported my mesh, messed with the lighting and, choosing to go with environment lighting, finally uploaded it to sketchfab!

Sadly I couldn’t upload it with the full 2048 textures (file too large for free account), so I had to downscale the resolution to 1024. https://sketchfab.com/matthewshannon


Reflection

Just looking at the 3D Digital Literacy section of my blog you can see my modelling development from the start of the year. I was surprised to see the amount of crossover with even just the first two model challenges and the challenges I faced on this windmill, when all else fails the multi-cut tool is always there to save the day. I have barely scratched the surface of Maya, but using just the multi-cut and a few other basic tools I was able to create a full scene that I envisioned in my head, when previously I struggled to model a table leg from reference. The talks from experienced workers in the industry, seeing their workflow, listening to their feedback has already gave me insight into the possibilities within Maya, and the level of progression possible. I tackled numerous different challenges in this assignment and managed to overcome them all. In just a couple months, I was able to go from having no knowledge in a foreign software to modelling a full stylised fantasy scene. This assignment alone has taught me a lot of new skills and I’m excited to see how much further I can progress.

 

 

 

Week 03 – Last Weekly Update/Windmill Project

Windmill Project

I’m changing the format of the blog from this point on, as it makes more sense to do it per assignment instead of weekly updates, I made a start on my windmill project by scrolling through sketchfab for hours and adding cool 3d models to my collections. I also collected some reference images of windmills or styles I wanted to replicate. I knew I didn’t want to do a modern day windmill, and wanted to go for something more stylised and medieval/fantasy.

          

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Week 02 – Modelling

Week 2 Challenge – Table

For this weeks challenge we had to model a table, using reference images taken at a more awkward position than the vase. I started this out the same way I did the vase, using a cylinder for the table leg. I bevelled the bottom/top vertices to avoid n-gons, then scaled it to the rough dimensions then I followed the contours, using the multi-cut tool to match them.

 

 

 

 

I then went back to the reference image in photoshop, to better get a grasp of the proportions, and notice the significant shapes in each leg. Using this new reference I made, I went back to my table leg, and added more divisions to get the shapes more accurate, and made some other adjustments, mostly changing where the shapes I outlined were, relative to each other.

 

 

At this point I was debating calling the leg finished, just a simplified version of the reference image, or going back and adding more detail. I decided to add the details, as my vase project was quite simple, so it would be good to get experience with something more complex. I wasn’t sure the best way to approach this part, so I just went back in with my multi-cut tool, adding more divisions for these details. I then added a cube into the scene, and scaled it up to a rectangle, to match the upper part of the leg.

 

 

I duplicated my finished legs and, using the distance tool, I placed them at a distance from each other I estimated using my own kitchen table relative to its table legs. I added another cube to the scene and put it in the centre of the legs, then scaled it to match the counter top in the reference. At this point Cathair sent a new reference image he found, which was helpful to compare my table to, and further clean it up to match this.

At this point all that was left to do was to add the underneath wood, connecting each leg. I went into top view and made cuts where I think this would come from and then I used the extrude tool on these faces.

I think my finished product pretty closely resembled the references and I’m happy with how it came out. Next time I would probably try and use less subdivisions, as the legs ended up pretty complicated.

 

Week 01 – 3D Digital Literacy Intro

Week 1 Challenge – Vase / Jug

This week, we had to watch the videos provided and try to model a vase / jug of our own. Our first step was to find a vase we wanted to use as reference using https://pixabay.com/. I decided on this image because I liked the shapes, I focused on the middle one.

After setting my project and my scene up, I set up my reference images using the orthographic views – front and side. I aligned the bottom of the vase with the grid, and put my vase in the direct centre.

 

To start modelling, I inserted a cylinder and scaled it roughly to the same proportions of the vase. I was introduced to the multi cut tool, probably the most useful (or at least used) tool when modelling.

I used the mutli cut tool, adding divisions (edge loops) along the main contours of the vase, the thinnest and thickest parts, and scaled it to match my reference. From there I would add another edge loop in the middle of these points using ctrl middle click, to better match the shape of my reference image, and have a less jarring transition.

 

I then had to change the top and bottoms of my model. as using the cylinder as my base made this an n-gon. This is to help avoid errors with the mesh during exporting. To do this I selected the vertex in the middle of the face and bevelled it, then connected the remaining vertices alone the bottom/top in a straight line, with one horizontal line through the middle.

 

Then at the top we used the extrusion tool in order to hollow out our vase. The extrusion tool lets us manipulate this face, moving it down within the vase. Using G to repeat this extrusion, I moved/scaled the faces down at each major change in contour from inside.

I encountered my first problem in this section, something I did caused Maya to unexpectedly crash. I didn’t lose a lot of progress, maybe 15 minutes, but from then on I learned to save a lot more often.

 

When I tried to render an image of my final vase, it made the whole scene black even after adding lights to my scene, I guess I’ll learn why in the future, for now I just took screenshots in my viewport to show how my vase ended up. Next time I would probably have chosen to do a more complex vase. I like how the vase came out but it is very simple, and without the flowers it could be hard to identify.

 

Final Animation

 

Full Group’s Animation: https://www.youtube.com/watch?v=E_OzwunBYYk&feature=youtu.be&ab_channel=EJAnimations

 

My part of the animation serves it purpose but it didn’t come out as good as I hoped it would, I think I was too ambitious for my first time, animating at 24fps and with the character running towards and then off camera. I had to cut out the cultist at the end to save on time, and my character isn’t exactly consistent with the others, she also doesn’t have much of a face. Overall, I would say its a solid attempt, and when partnered with the other parts it feels more complete.

Animation Studio – Portfolio

Week 1 – Form and Shape

Week 2 – Composition & Perspective

Week 3 – Tone / Value

Week 4 – Colour

Week 5 – Character

Week 6 – Studies Advice – One to One tutorials

Week 7 – Animation Introduction

 

Week 8 – Storyboarding & Animating with forms

GIF Version (if progress bar blocks him)

Week 9 – Animatics & Secondary Actions

 

Week 11 – Progress update on animated run

 

Week 12 – Final Animation

 

Full Group’s Animation: https://www.youtube.com/watch?v=E_OzwunBYYk&feature=youtu.be&ab_channel=EJAnimations

 

Week 11 – Progress Update

ORIGINAL ANIMATIC

Process – Working from the animatic in Krita

Running past camera – rough animation and timings

I will change the background to better fit perspective of running, easier than changing each frame individually + 4th stride needs to cover more distance. Obviously it all need cleaned up a little, and i need to animate the hairs movement. I was planning on having the cultist character walk around the corner and dramatically look towards where she ran to, to save time – but I’m not sure how this would impact my transition to Jess.

I researched a lot trying to create this run. watching YouTube videos, reading online, recording myself running about my house as feminine as I could. I have a page of ‘maths’, estimating the distance covered and how fast the average person would run it.

We’ve been unable to contact Cathair, who was animating the first 6 seconds, establishing the scene, so I guess we’re starting with me. In order to have it loop with the ending I will try to animate something at the beginning if I have time to but if not I’ll just fade from black, so it somewhat loops with Harry’s.

I was promised full marks if I released the embarrassing reference videos of me running about my house, but I am never posting those to the internet. Not because of how I run though, I’ve got the stride of a gazelle. A beautiful, beautiful gazelle person. My body achieves a perfect symmetry. It’s that long, lean muscle I’ve worked so hard to achieve.

Here are a couple stills from one of them though :/

 


Visual Reference

Week 10 – Progress Update

 

I spent a lot of time messing around with the perspective, I chose a very hard scene to animate for my first real time animating because I am stupid. I spent this week trying to figure out the correct perspective of a character running both towards and past a camera, as well as changing the background a little.

We discussed on discord and decided to stick with a monochromatic grayscale colour scheme, using my backgrounds as a reference. EJ wanted to do the audio for our animation, and link it all together, but Cain and I decided to do our individual sound effects, with EJ adding an ambient backing track to it all, so it would link together well.

I used this to help: https://www.youtube.com/watch?v=F1Ewm3c8hT0&ab_channel=EdTadeo

Week 9 – Animatics

This week we made an animatic based on our storyboards, showing the key points and timing.

This is the 6 seconds I worked on:

I think I will slow everything down a bit, build more suspense – but i tried to shove it into the 6 second time slot. I might have the character turn his head as if he is listening to the footsteps, before running after her, as the cultists original idea had them being blind. Then again to showcase the blind cultists idea, he could run into the wall at the end instead of running off camera, adds more variety to the animations instead of having just two identical running paths.

 

We then combined all our pieces , to see how it would flow into one another, Cathair’s scene should be before the first scene, establishing the character and context, linking the beginning and end causing it to loop. He’s not been able to make it to class recently, so if he isn’t able to do it, I’ll try and add something at the beginning to loop with the ending,


Visual Reference

https://www.artstation.com/artwork/LoVB5