3D Modeling – Utah Teapot

In this 3D session, we learned about the symmetry and mirror functions and Maya and modelled the Utah Teapot.

Before we got to the main part of the lesson, however, we addressed the right way to delete edge loops, since a lot of people, including myself, have been deleting them incorrectly and causing problems for themselves later in the modeling process.

First, we created a cube (of course) and then added an edge loop in the middle, selecting all of the edges then created. If we didn’t want them, most people, including myself, would instinctively hit delete at this stage, but we were then shown why this wouldn’t work properly. In vertex mode, we could clearly see that the verts remained even though the edges were deleted, so all of the edges of the cube were effectively split in two. This is the infamous case of “floating verts” which can be incredibly annoying to 3D artists. To properly delete the edges, we selected them all, went to ‘edit mesh’ and selected ‘delete edge/vertex’, and this gets rid of both the edges and their verts.

 

 

We also looked into an alternative method to filling up holes. Using ‘fill hole’ is one option, but in this lesson, we looked at ‘bridge’. Once everyone had a cube with a hole in it, done by inserting edge loops and deleting a middle face, we selected the top and bottom edges of the hole and went to ‘edit mesh’ and ‘bridge’. This instantly filled the hole, much like the ‘fill hole’ function, but allows for more accurate fixes because of the ability to select edges and then apply.

 

 

 

After these quick tutorials, we moved on to the main part of the lesson, which began with symmetry and how it can be used in Maya.

Symmetry – the quality of being made up of exactly similar parts facing each other or around an axis.

This is the correct definition for symmetry in 3D, which was actually far more simple than I thought. I kept trying to search for some mathematical, complex meaning since symmetry is a concept introduced to me through maths, but I am grateful to have a more simple, clear meaning that I can remember and refer to much more easily. Symmetry exists in society in more ways than just in the classroom, however, (obviously) and has been regarded as one of the closest states to perfection throughout history. Here is an example of a well-known symmetrical building, the Taj Mahal.

 

 

Demo – Symmetry Function

 

This time, we started with a sphere and went straight to the tool kit. In the top part (not the buttons) there was a Symmetry option which I had never noticed before, with a drop-down that allows for the selection of different axis. If, for example, you chose ‘object X’, the symmetry would be based on the x-axis. We tried this out by clicking the very same option to see what it would do to the sphere. The difference was not very visible, but we were shown that we could affect both sides equally and at the same time. Extrusions in face mode being the go-to, I went with that first to test it out and saw that everything was indeed repeated exactly on the other side of the sphere.

 

 

 

This already seemed very handy and time-saving when creating symmetrical objects, since there is no need to spend time trying to match the actions already done on one side. However, an arguably more useful tool was what we learned about next when we undid our actions, hit space and went into the front view. We were told to turn the symmetry off and delete half of the sphere by selecting as close to the centre line as possible in face mode, leaving only one half of the sphere.

 

 

Then, we selected the box next to the ‘mirror’ option in the ‘mesh’ dropdown, which made the window of settings for the mirror function appear. I wasn’t surprised to see a bunch of buttons, levers and buttons to select, since this tool seemed like it would be more complex than the symmetry one. Because I could see that it offered a lot more customizability, I knew I was correct.

 

 

I looked at the bottom left corner of the workspace and saw that the x-axis was where I wanted the shape to be mirrored. So, I chose this direction in ‘mirror axis’ and then chose ‘+’ in ‘mirror direction, because the sphere would then be mirrored in the positive area of the x-axis (to the right) instead of the negative area, where the sphere half would be the other way round. After selecting ‘merge border vertices’, the mirroring happened, and the sphere was full yet again. The symmetry function was definitely faster, but I could see how this would be useful too. Both tools perform a similar purpose but they are slightly different and give the artist more freedom depending on what they need and what kind of workflow they prefer.

 

 

After this, we also looked at the ‘world’ option of the mirror function, by deleting half of the sphere again but this time moving it to the left or right and changing the mirror option to ‘world’. A dupe of the shape appeared, but it was also noticeably the same distance away from the centre of the grid as the original. World mirrors the space as well as the shapes, hence the name and why the distance was duplicated too.

 

 

This is very useful when modelling arms for example; one is complete, and a much faster and more accurate method than simply duplicating would be to mirror the shape and distance so that the exact same arm appears in the exact same position on the other side of the body. I then inquired as to how to select the shapes separately, since I had noticed that every time I clicked one, both were automatically selected. The solution to this was quick and simple, by going to ‘mesh’ and clicking ‘separate. I was sure to take note of this because if I ever model a character, I might want to change certain features to make them asymmetrical, such as a missing finger on one hand. This would require modelling only on one of the shapes. I could now see how the mirror tool offers more flexibility, and why it is so preferable.

 

The Utah Teapot

 

This is a particularly famous teapot and is quite well known in the 3D community, so it would be in our best interest to also know of it thanks to some quick research. It is an inside joke amongst 3D artists, being the first complex model to be used to test 3D algorithms. It is said to be an ideal object for tests, since it is not too complicated, but also not as simple as a primitive shape for example. As a class, we also watched an informational video about the teapot and saw examples in famous films such as Toy Story, where the teapot was snuck in and wouldn’t be picked up on by the general audience.

 

 

Modelling Task

 

Predictably based on the above research, our task was to create a teapot in Maya using the following skills:

• Using the extrude tool
• Creating a curve
• Extruding along a path
• Inserting edge loops

I began by importing the images given to us and using ‘x-ray’ and ‘wireframe on shaded’ to start modelling and simultaneously being able to see the orthographic.

 

 

Then, using both front view and perspective mode, I began to model the base of the teapot, extruding bit by bit to create a nice curve. This part didn’t take too long as the extrusion process was like second nature to me by this point. Where the steps before were complicated and tricky to get right, they are now simple to understand and very easy to successfully do. Pressing ‘G’ and using the arrow or red cube to scale were the only things I really did from the bottom to the top and soon enough I was moving on to the lid.

 

 

Initially, I thought of extruding, but for some reason, that didn’t sit right with me. I had the inexplicable urge to create the lid separately. Perhaps I was truly thinking like a modeler and trying to build a teapot in 3D as one would in real life, or perhaps I thought this would be easier than trying to get the ridge of the lid through extrusion, although that isn’t the case and I definitely opted for the harder route. I also later realised that I had deviated from the task since we were supposed to use extrusion for the lid as well. I think that my extruding abilities are quite clear, and I still used the skill for the lid, just not connected to the base. It would have been easy to go back and do this, so I am glad I went for something a little trickier, to gain more experience and engage my mind, since making a separate lid actually also inspired me to hollow out the teapot, to make a hole where it would fit.

Making the lid wasn’t too difficult, although I did have a little stumble with the edge loops at some point. I simply went to the top view and created a cylinder to match the circle that represented the lid. I then used extrusion to build it up and form the shape of the knob, again, extruding bit by bit to achieve a curve, which would later be much smoother with SMP applied.

 

 

 

 

The trouble I had was when I extruded from the bottom of the lid to make it able to sit neatly in the teapot. I must have been careless and rushed because there were extra that I didn’t want and had to delete, but I fixed them using ‘delete edge/vertex before I could get a screenshot. I am actually very proud of how I modelled the lid because of the knob and the bottom extrusion. It’s really fascinating to see a simple cylinder turn into something like that thanks to a few simple steps that you take.

Once I had completed the lid, I moved it down towards the teapot and started thinking about how I could make a hold for it to neatly. The x-ray and wireframe on shaded options helped immensely with these next steps since they allows me to see through the entire model; how the lid fit into it, how big the hole would have to be, and how curved I was making it. My pace slowed down towards the beginning of making the hole, since I didn’t have a lot of experience extruding inwards, only outwards. I also struggled slightly with making the hole exactly big enough to fit the lid, and again, being able to see through the frame helped a lot, since I could tumble downwards and look at my extrusion from the bottom, to match the faces up. I had to wrap my mind around the steps to making the hole round too, even though they were the exact same, just because I was extruding in a different direction and tunnelling inside a shape. It was the first time I had ever done it, and it was definitely an interesting experience.

 

 

As you can see from the screenshot above, my first instinct was to go down and then out, but I knew this was wrong right away. I wanted the hole of the teapot to match the curve of the walls as if you would actually be able to hold tea inside. I don’t know why I was so intent on getting this correct since if I had simply made a small hole at the top for the lid, that would suffice. It would not make a difference to the outside appearance of the pot either way. But I did it anyway because it was enjoyable and made the model even more interesting to build on. Having a working teapot with a lid that fits inside is much more fun!

 

 

None of this, however, compared to my struggle with the handle and spout. It felt as if I had waged a war against that handle since I made so many attempts to create it myself and none of them worked. I reluctantly had to ask for help at the end and follow the demo, which was frustrating because I had managed fine by myself for the other parts of the teapot and it would have been incredibly gratifying to complete the whole thing by myself. Unfortunately, I wasn’t quite at that level yet and had to ask for a tutorial in how it could be done. Firstly, here are my attempts:

Initially, I went to extrusion. I thought that if I could simple insert an edge loop to make the faces near the location of the handle smaller, select them and extrude, I would eventually have a nice handle. This seemed like a very tedious method since you have to extrude bit by bit in order to make something curved, and I remember the teacher saying that this is exactly what we want to avoid and we learned extruding along a curve to save time. Unfortunately, the tool we used for extruding along a curve didn’t make curved lines, only straight ones. Plus, it could only be done don’t the grid itself and I wasn’t sure how to follow along with the handle and make it accurate. So, I tried my method first.

 

 

 

This method was working fine at the beginning, but then it all started to go wrong. Even though I was extruding in small parts, it was becoming increasingly difficult to match the curve of the handle, because the faces that I extruded from weren’t straight. They were curved since they were part of the base of the teapot, which meant that their extrusions would also be curved, or at least, slanted, as you can see above. Soon, I reached a point where the handle was getting too inaccurate and I couldn’t continue. It was very annoying to delete all that work, however small, and go back to start again, but I wanted the handle to be as good as my lid, so it was necessary.

 

 

My second attempt, looking back now, was even more comical and unlikely to work. Since extrusion didn’t work, I thought that I’d try out the same method I used for creating the lid: using a separate primitive and attaching it afterwards. I suppose I should have used a cylinder and tried to extrude it in a curve, but at the time, I didn’t think about this and went for the doughnut instead. I generated the doughnut-shaped primitive and tried to somehow shape it into the handle. This was doomed to fail from the start, and after trying to scale and stretch it, I knew that it wouldn’t work, and deleted it.

 

 

At this point, I was in despair with this handle. I knew that if I tried extruding along a curve, it would take multiple attempts to try to match the curve of the handle with straight lines, and it would take far too long. In hindsight, if I had just explored a little, I would have found the correct tool for this easily enough, but I was too exasperated to try anything else of my own accord for a third time and asked for some help. We were, luckily, given a demo, and shown how to make the handle and spout correctly.

 

Handle & Spout Demo

 

Firstly, to tackle the handle, we went into the front view and clicked the ‘curves and surfaces’ button in the top bar, where the ‘EP curve tool’ was then selected. This is what allows for a curved line to be made, not the straight, jagged one we created last time. Using that, we drew a line through the centre of the handle by clicking every so often, making sure to keep the distance between the verts quite small, and roughly the same size, to achieve an accurate curve, which happened automatically since the tool was designed to create curves. There was no need to be super meticulous too, just roughly correct, since we could always go back in after and in vertex mode, move the verts to adjust the curve.

 

 

Next, similarly to when we did the extruding along a curve lesson, we generated a cylinder and rotated it 90° along the z-axis, moving it in place so that the curve was just going inside. With a quick jump to perspective mode, we selected all of the faces on the front-facing side of the cylinder (touching the line) and then shift selected the curve and extruded.

 

 

As expected, the cylinder awkwardly extruded to the end of the line, which we fixed by adding more subdivisions and making a smooth handle. I always enjoy seeing how the greater value of subdivisions changes the shape since it’s quite satisfying seeing it become smoother or match the line that it has been extruded along. I was still slightly annoyed that I hadn’t figure this out previously, since the method was quite simple and pretty much the exact same as our ‘extruding along a curve’ lesson.

 

 

Whilst waiting for everyone to catch up with these steps, I decided to try and jump ahead and do the spout. I believed that we would be using the same method – it seemed logical considering that the spout was also a curved object like the handle. I was wrong yet again, because after extruding the cylinder along the curve I created, I tried to shrink it down and scale it inwards near the top and it didn’t work. It was the same thickness all the way through and I didn’t know to how make it gradually become smaller. The geometry and edges were also very weird, and it was obvious that I was in the wrong lane.

 

 

For the fourth time this lesson, I had to delete a part of the teapot and start again. I wasn’t angry though, since it was all part of the learning experience and helped me gain new skills, such as using the ‘EP curve’ tool. I waited for the demo to continue and took a quick moment to reflect on this, and look at my teapot once again.

When the tutorial started up again for the spout, we entered the top view-port instead of the front one and created a cube instead of a cylinder, which was placed next to the outside edge of the spout. This surprised me since the cylinder seemed the better option, being curved like the spout. In vertex mode, we then left clicked and dragged through one side, moving it to the edge so that the cube was a thin pane.

 

 

Like before, we popped over to the 3D view for a second to select the front face of the cube and then back into the top view. This enabled us to start extruding, pulling out the front face and matching it to the edges of the spout. We used the little cube on top of the red arrow, not the arrow itself and made sure to extrude in little parts so that a curve could be achieved. Of course, this would later be properly done with SMP anyway, but a clear curve is still needed. For the tip of the spout, we extruded yet again and this time scaled in almost fully so that it was closed off and looked slightly more rounded.

 

 

We had now shaped the cube into a spout from the top, but it still needed to be matched from the side. So of course, our next view was the side one where we selected the cube/spout in object mode and moved it down so that the beginning roughly matched the width of the spout. Since the cube was already extruded, it was now time to shape it using the verts, which we were shown how to do with the first couple and then left to complete ourselves.

I drag selected the top vert, move it into place along the top line of the spout, and then did the same thing for the directly opposite bottom vert. I could have done all of the top verts first and then all of the bottom ones, but I chose this method instead because it was more clear to me and it felt like I was truly building up from the side. When moving the verts, I had to estimate the right distance between each one to allow for a flow/direction with the lines. They were not supposed to be completely vertical. I also made sure to drag-select the vert I was moving every single time since I knew that if I simply clicked, it would move only the vert on that side and not both.

 

 

Once I had finished matching the snout, I went back into perspective mode to see my work, and I was really happy with it! It was actually looking like a snout! I knew how to create a hole (select the face, extrude inwards) but I was running out of time, and I really wanted to get some good renders, so I skipped it and moved on. After selecting the orthographics and clicking ‘H’ to make them invisible, I could see my teapot in full and I admired what I had modelled.

 

 

 

For the rendering aspect of the lesson, I will be putting down my notes word for word and adding my opinion of the process and renders at the end, since I followed the instructions pretty exactly and had no problems to comment on.

• Make a poly plan, assign a new material, add blinn. Darken the colour.
• Go to Arnold, lights and select ‘skydome light’.
• Find a Kitchen HDRI image and save it.
• Select the teapot, assign a new material, go to Arnold in the window and click ‘aistandardsurface’ where in presets, you can choose ceramic, chrome or car_paint_metallic.
• Select one and change the colour if you wish.
• Select ‘skydome’, click the checkered flat at the end of ‘colour’.
• In the window that pops up, click ‘file’ and then the image name box. Find the saved kitchen picture and select it.

Doing all of this allowed me to reach this stage with my teapot, and I can confidently say that I enjoyed every second! Being able to see my model in an environment with a realistic material and interesting reflective surface was so satisfying.

 

 

 

I also got the chance to play with colour and experiment with Maya’s rendering and material capabilities. I don’t remember exactly how, but my friend showed me how to pull up a window where I could experiment with the look of the teapot more in depth. I started adding materials on top of each other, mixing different combinations together and seeing what kind of result it created. It was so cool to see, and I really love the results.

 

 

 

 

 

– Final Render –

 

 

This lesson was nothing short of a rollercoaster for me. There was a great mix of frustration and anger, joy, awe and pride. I kept taking wrong turns and having to repeat steps, and I lost the time necessary to complete the teapot fully, such as add the hole in the snout. However, I learned a lot of new techniques and tips and I am really happy with how the teapot turned out! My favourite part was adding the materials and background but I can say that I enjoyed the whole process from start to finish, even when I was making mistakes. Modeling is not easy, but this is what makes it so great when you can overcome obstacles and see your finished work in a render.

 

Quick Tips that I have learned this lesson

If the window that allows for extra subdivisions had disappeared, press ‘T’ to make it reappear.
To work on one thing at a time (one part of the model), select it and click ‘isolate select’ which has a mouse icon selecting a square.