CAGD 320

Blacksmith Micro Scene

May 10th, 2019 - Finale

So this is it...this is what it feels like to be "done" with a project. I put done in quotes because I don't think that a work is ever really "finished". We can always go back and tweak the little things that make it even more outstanding (well, I don't know that I would necessarily call this project outstanding, but you get what I mean!). Anyhow, this project has been an adventure. Here are just a few numbers to put it into perspective:

• 4 different software programs
• 6 blog posts
• 14 project iterations
• 5 weeks
• 10 class periods
• 30+ hours of work
• ...and too many crying and frustration sessions to count...

To start this project, I did what every good artist does: I looked for concept artwork. I find that making Pinterest boards is super helpful for me to just gather a bunch of ideas and then put them all together to make one cohesive thing. Funny enough, though, I ended up finding a scene that I really liked and used one piece of art to make it all happen.

Once I got that down, I started in Maya. There were a bunch of assets that were provided, but I was not a super huge fan of the look of them and they did not match the scene that I was making. Thus, I decided that I was going to be modeling. This presented lots of problems that I did not take into account. Number one was that after taking basically a semester off, my brain was not ready to model. It had troubles remembering how to do things. Then, on top of that, I was having some serious issues with n-gons. I will have you know, now, that I am pretty dang excellent at n-gon problem solving.

After I thought I was done with Maya, I was never actually done with Maya. I had done all my UVs and made my ID maps, but I still found myself frequently going back into Maya because I goofed somewhere. If you have never used Substance Painter before, know that it is very touchy and will show you all the places that you goofed when you were working in Maya. After about 80 iterations of my Substance projects, I was able to get everything textured. I have found in Painter that I am not very good at making my materials from scratch like we did with the Front Loader. I am, however, good at taking smart materials and adding on top of them to make the textures that I want. I was able to make the texture for my archway from a smart material, but it was so heavily edited that it is honestly my own at this point. It was also nice to work in Substance Designer for texturing. I will be honest, the transfer from Designer to Painter is a little displeasing due to the lack of the "tessellation" feature in Painter, but I found Designer to be so much easier to navigate compared to Painter (don't ask me why). Once everything was textured (and textured 100 more times due to my goofiness), it was time to get into Marmoset.

Getting into Marmoset was interesting. We were provided some videos to watch about it and I think that they are useful, but it was a lot of messing with settings to get the look that I like. The narrative to my scene really began when I started lighting. Through the lighting, I wanted to capture that early morning scene for our Blacksmith here. I wanted the lighting to be a little softer and just be something playful. I didn't want anything too mysterious or foreboding, just something pleasant to look at. I really appreciated my ground texture for that too because it brought some life and vegetation to the scene.

I think I definitely learned a lot of dos and do nots with this project. I didn't necessarily have a time management problem, pat on the back, me. The biggest thing for me was making sure that I knew how to problem solve. I was running into problems right and left, whether it would have been a place I goofed in Maya, a Painter issue, or just my mental capacity. At some points in this project, I really had to learn to walk away from the computer and do something else. I knew at some points that if I had continued to work, it would not have turned out like I would have wanted.

If I were to do it again, I would have wished that I could work with other people. In school it is hard because you are learning everything and your projects are your own. Out in the industry, you are really only doing one part of this project: modeling, texturing, UVing, rendering, etc. I think that for me it is fun to work in a group because you always get a new perspective on things too. I also think that I could have probably added more to my scene. I was so concerned that I would not have enough time to texture if I didn't get out of Maya, and I made some critical errors. Overall, I am pleased with how it turned out, I just wished that I would have put a little more effort into some areas (which I won't tell you so that you can appreciate the scene without seeing my goofs).

That's all for this one! Until next semester!

May 8th, 2019 - Week 5

This week was a lot...and the deadline is coming up and I'm getting stressed out. Yes, rant done. This week was a week of texturing..and problem solving.

So anyhow, when Connor gave me the brilliant idea to make certain textures for things and then just stack the things for Maya to figure out later, it was a wonderful idea, but completely screwed up my mental capacity for my brain...so yea. The first part of this week was figuring out how to reconfigure my texture maps to work for the thing that I was trying to do. Once I was able to get all of those figured out, then the real work began...texturing.

I started putting things into Substance Painter and I began to just figure out what I wanted things to look like. I found that making smart materials look like I want them to look like was the best way to attack things initially. I was looking a bunch at the reference picture and also thinking about what I really wanted things to look like. Eventually, I have been able to get that stylized look that I was going for, but in my opinion, I do not think that it looks all that good...

This was my pride and joy (with the help of smart materials, but they're heavily edited...promise)

I have also been working inside of Substance Designer to get some custom textures. My friends and I at the beginning of this project found a giant bank of Designer tutorials that we could use for this project and I saw one that looked really cool that I thought that I would use for my ground texture. The tutorial itself was only 14 minutes and I got really excited for the look of it. Then I started to do the tutorial...and that was an adventure.

Here is how the texture turned out

The texture ended up looking okay, but once I applied it to the ground, it was a little bit less cool...anyhow, I ended up taking the cobble out of that texture as well to use it else where, so I guess it was not a complete bust. Other than that, I also ended up editing my brick texture a little bit so that I could use it in my scene. We are still putting on the finishing touches for texturing, but I was a little over it by the time I got to the structure.

...and then I took the cobble out of the texture to utilize it elsewhere. (I edited it some more to look a little cooler)

Beyond that, here is kind of where we are at with the textures. We will get in Marmoset before Friday...it may be a rocky journey...BUT LET'S GET THIS BREAD.

April 30th, 2019 - Week 4

Remember when I said that I had a goal to get stuff done? Yea...that did not happen...I tried really hard, but let's just say that I had an eventful week on the other side of things.

I was able to get a little bit of work done, even if it was not what I wanted at the end of it all. I was able to FINALLY finish UVing literally everything. I am a little sad though, because once I got into a rhythm with it, it was something that I found to actually be super simple. After that, it was fixing the little things. I realized that I had MORE N-GONS that I had to handle; let's just say thank goodness for the multi-cut tool or I would be in deep trouble with n-gons.

When I filled this hole, I didn't realize that it was a giant n-gon...

After that, it was an adventure of figuring out what I wanted on my texture sheets. I also came to the conclusion that I have a LOT of texture space...and I mean...a LOT of texture space. All in all, I really only occupy the three 2k maps that we were alotted and the two 1k maps. I was able to get my outliner mostly organized now too, so that makes me feel a little bit better about things. I sorted things by the texture map that they were on and inside the maps, I tried to put like items together (i.e. the house parts, the weapons, ...).

Here is everything all laid out (disclaimer: I'm not using UDIM tiles)

Then, the real trouble began...I had to remember how to use Substance Painter. I will be super completely honest, I pretty much completely forgot how the program works, even though it's easier than Substance Designer. Don't ask me why, but Substance Designer makes so much more sense to me than Substance Painter and I think that it is because I am basically making the thing from scratch. Couldn't tell you though. I ended up throwing some smart materials on my sword with the ID map that I put on it, but it did not really look like I wanted...so I deleted the file.

I ended up trying to do something else and that something else was looking through the tutorials for Substance Designer on YouTube to find a good ground texture. When I found one that I liked, I was a little bit over it....so then I decided to close the file for the week.

Overall, not a super productive week, but at least I got a few little things done here and there. Crunch time is here though...so there will probably be a spike in work next week...

April 24th, 2019 - Week 3

This week was a lot of UVing...and when I say a lot...I mean A LOT. Why? Well, because I am not smart...yea. Remember how I talked about how I may have bit off a little more than I can chew? Yea, well, that continues into this week. Where the scene that I have created I am happy with, it is a lot of work on the back end. Blocking out is just the beginning. Hey, we will see how things turn out.

Anyhow, UVing was not actually too terrible. After being introduced to the 3D UV Cut and Sew tool in the Dagger Tutorial, life in UVing has been a much quicker process. I knew what I was doing for the most part when it comes to UVing, but I was not able to do it easily in the 2D view. Being able to but in the 3D view is super helpful. The only thing that made it daunting was the pure amount of UVing that I was doing. I spent about four hours on Monday doing just UVs, but I have literally one more object to UV and probably a couple things to clean up (I will tell you why later) and then I am set to texture! The one object that caused me the most stress when it came to UVs was the staircase that I made coming out of the side of the house. For some reason, when I was making it, I probably made a double extrusion, which caused some other faces that created edges that would not go away...complicated explanation, but you can see what I am talking about below. In order to fix this, it took a lot of face deleting, some vertex merging, and then some hole filling. At the end of it all, I was able to fix my problem :)

Bad Stairs

Fixed Stairs

The other issue that I was running into a lot, which is where my clean up is going to be in, is that I created a lot of n-gons...like A LOT of n-gons. How? No clue. I think that a lot of it had to do with when I was using the bevel. Instead of beveling an entire object, I was beveling the edges and magically it created a lot of n-gons. This happened specifically in my weapon rack and I realized that it was also in my staircase and the little stump that I made. I was able to use the multi-cut tool to fix that right up, but nonetheless, it was still something that I had to fix.

I fixed the sides of this...n-gon city...

...and I fixed the part around the twig...also n-gon city...

Next week, I am hoping to get some of the textures done with. I have a floor texture that I would really like to go after making in Substance Designer, so the goal is to at least get that in. I will probably make a base wood texture as well that I can change up the color and use all over the place since a lot of my scene is wood and the other thing that I am hoping to get done is some kind of cobblestone texture.

That is it for this week! TTFN!

April 17th, 2019 - Week 2

This week was definitely an adventure...I feel like I sort of bit off more than I can chew, but hey, I am sticking with it. I made a commitment to my idea and I am going for it. Too late to turn back now. HAHA.

So, in my process, I may have abandoned the "squished, dumpling" idea in my scene, but I am still going for a more stylized look, that will probably come with texturing, however. I was a little worried this week about the poly count that we were given, but upon modeling lots and adding assets, then checking...I realized that I wouldn't have much of a problem. I have been working off of my reference image in order to make the composition like the art that I found to use. I found that this was more of a weapon-making blacksmith and they seemed to make weapons of all types: swords, spears, shields, axes. Apparently, they have a lot of customers....but they could just be making a mass amount of weapons and that works too. 

Anyhow, in the process of this week, circling back, I realized that this was going to be a lot of work on my end. Leave it to me to always take an assignment and make it more complicated....but hey! I love it, anyway...so HA. The bulk of the work started with the structure that I wanted to go for. I liked the structures that were provided, but they did not exactly match what I wanted...therefore, I made my own. Challenge #1 was the ROOF. Why? No idea. I definitely worked on it for an hour and a half and couldn't figure it out, so I asked an outside source...aka Connor, who allowed me to make the process take 20 minutes. Won and done.

Challenge #2 was creating some unique weapons to put into my armada. There was a sword in the asset collection, but that one looked a little bit worn down. I figured, if this blacksmith is making weapons, I can imagine that a lot of them would be pretty new. Thus, I made my own sword. Next, I wanted to make another weapon that was not provided: an axe. This one was a little trickier than a sword, but I made the axe before the sword...so the sword was comparatively easier.

The thing that I did not think about after modeling these complex models was....bum bum bum(!!)....UVing. Each weapon that I began with posed its own unique challenges. The biggest issue I had were the sides of each of the tops of the weapons. Eventually, with the help of the 3D Cut and Sew tool, I was able to fanangle the UVs a little bit and make a variety of (needed) small pieces out of the sides in order to make some kind of sense out the shapes.

That is it for this week, I made a little more progress in the blockout and I have been UVing as I go, but we will see where we end up next week. TTFN, friends!

April 10th, 2019 - Week 1

For this assignment, we are to compose and texture a micro blacksmith scene. In my head, I immediately thought of it like a dollhouse and I think that that's sort of what it is...huh. Anyhow, for my scene, I have been playing around with a million and one ideas...it's possible that it might change after my initial selection, so stay tuned for that I guess.

In my scene, I still wanted to go for that typical Medieval Blacksmith kind of vibe, but I really want it to end up looking stylized. Since we are provided most of the assets for the scene, it is going to be up to me to make the things look stylized. My plan for this is to draw a lot of inspiration from two different sources:

1. John Loren, an illustrator/online artist - I really like his art style and I think that he has a really interesting viewpoint on things
2. Pixar's animated short, Bao - I like that everything is so squished and looks like it is meant for dumpling-like people...that is the style I want to go for in the end

I have been developing a Pinterest board as well in order to collect my thoughts on assets, style, and color scheme. From this, I have found a kind of composition that I am going for from a piece that someone posted on ArtStation. I think that another reason that it caught my eye was because it matched the art style of John Loren as well. You can see this image below. 

I will continue working on my blockout and then more to come on texturing and "making things look pretty" (in the words of Jen). :)

Knife

March 15th, 2019

In this assignment, we had a couple of tasks. First of all, we were to take a super high poly model and retopologize it to make a low poly model, texture the model, and render it in Unity. The biggest task in this assignment was the retopologizing, mostly because it was something new! Through in class lecture, we learned how to use the Quad Draw Tool in Maya. The tool utilized a series of vertices to create the parts of a quad. When the tool recognized there was a quad, you could fill it in and create a mesh out of it. Once the mesh was made, it was then up to the relaxer (with the help of good ol' soft select) to mold the mesh to match the high poly mesh. Once you get used to the tool, it is super helpful and works wonders.

After exporting the mesh from Maya, it was then the job of Substance to aid in texturing. I ran into some problems initially because I was not thinking correctly when it came to my UV shells. Before exporting the model as an .fbx, I had forgotten to orient all of my UV shells the same way (thank you, Alex, for catching my error). Once I got that going, it was making sure to edit the smart materials to give it my own flare and taste for looks. Then, away it went to Unity!

Using the Quad Draw Tool for the Face Knife

Putting models in Unity has never been something super difficult, the only thing is the added bonus challenges that come with Unity. Model importing and making materials is generally straight forward, the part that I struggle with is making sure that I have all of my maps correctly exported from Substance Painter. Something probably should have clicked in my head when I only got 3 exported maps from Substance, but eh [insert shoulder shrug here].

Using the Quad Draw Tool for the Native Knife

After getting everything plugged in and the model looked pretty good, it was the other thing that we sort of got to learn last semester, but got more adept with this semester: post-processing in Unity. The biggest problem for me with this was that I was having some issues with the imported package into Unity. Through some consultation with Jenn and Connor, we found that it was a couple of scripts that were causing a problem. The solution: deleting the scripts because we were not really messing with them...

Overall, this assignment was an easy one, but also one that allowed us to learn some new things. The Quad Draw tool will definitely come in handy when wanting to easily retopologize things and with the knowledge that we already have in Substance, it makes our models look that much better than they were before. :)

Front Loader

February 25th, 2019

Greetings from the construction site...aka...my desk. Anyhow, making this lil thing (or monster thing, depending on how you look at it, was its own little challenge...for a number of reasons, but hey, it be like that sometimes. In the process of creating this, it was a lot of layering. Layers on layers of a really big cake that we did not really get to eat. Layers were super important, however, because they gave you the ability to make little changes here and there. Each later was like a bookmark that you could place and then come back to and tweak at a later date.

ID Mapping

Ambient Occlusion

Before those adjustments, however, it was super important to bake your maps onto the model. Maps can be baked on in a variety of ways, but some of the main ones that we really used were ambient occlusion and the ID maps. Ambient occlusion works in such a way that it allows you to add those depth details that you look for in real life rendering. When you bake that map on the model, you are essentially adding all the little nooks and crannies into your work. The ID map was a color based map that you could not really see, but was used in order to apply your textures in a convenient way. Through the ID map, you could see colors that you assigned to the model in Maya for the purpose of making one texture per color. It is like when you color code a binder, but in this case, it is a model. 

These two maps were super important in the process, but the maps inside Substance have different adjustments as well. The Ambient Occlusion map for example has a couple of parameters that are super important to mess with. Making sure decrease the occluder distance is of vital importance to make sure that you are not projecting the shadows of the ambient occlusion on parts of the model that you do not want them to be projected on. This causes some weird issues that make the reality of your model a little bit wacky. You will also want to check out the common tab as well to see how far the ray is being cast as well (though this will usually be of importance when you are baking a high poly model onto a low poly one). The ID map has specific parameters as well. One really important aspect of the ID map is the option called "Match > By Mesh Name". Through this option, you can have the model reference specifically to the Maya file that you were working with. The parameter will reference due to the name that you put on the parts of your model and match the parts of the two different models. This sounds a little bit complex, but think of it this way: you have two jars, one labeled "Flour" and one labeled "Sugar". When someone else goes grocery shopping for you and puts away your flour and sugar, you can assume/hope that they put your flour and sugar into the proper jars (Substance is the person that went grocery shopping for you). 

When you are texturing in Substance, there is another thing that you need to look out for when it comes to texturing and that is the usage of multiple UV tiles. Why is this important? Well, let's think in terms of our Front Loader here. When you are using multiple tiles, it allows you to have variety between different parts of the model that are exactly the same, like the tires on the front loader. All of the tires share the same model, but they have different tiles and, therefore, take a different iteration of the texture.

Tire Tiling

All in all, this tutorial was super helpful for reviewing the vital parts of Substance and allowing us to expedite the workflow of our texturing, especially when texturing something as big and complex as this. I think that if there were less texture sets, it would have definitely aided in the time of the workflow, but things like instantiating, ID maps, and smart materials make the process a little bit less of a bear.

Dagger

February 17th, 2019

Maya Render

Unity Render

Models Courtesy of Frank Pereira

In order to create this, we followed the tutorials that were given by Frank Pereira and Jennifer Underwood. The model was already made and our job was to texture it and make it look, well, like an actual dagger. When we initially opened the Maya file, the only thing that was in the Maya scene was a fully modeled dagger that was smoothed. After that, we put the model back into the "1" setting that made it the low poly model. In order to expedite the process of modeling and texturing, we UVed the model before we duplicated the dagger for the next step. All of the models did not necessarily need to be UVed because all of the things were happening to the low poly model. Just to cover yourself, it is helpful to UV all the models just in case you needed to do something later to the model. We then created two different duplicates of the dagger: one was a high poly version of the dagger and one was a color ID. The high poly was a model that was used in order to bake the high poly detail on the low poly model in Substance Painter. The color ID was used in order to put the textures on only particular pieces of the dagger in Substance Painter. Before putting all of our models into Substance Painter, it was important to "zero" them out in the Maya scene because when they are put into both Substance and Unity, the models are placed on top of the low poly based on the location in the Maya scene. After we had everything textured and mapped out properly in Substance Painter on our dagger, it was time to export everything. Since we wanted to render the model in Unity and in Arnold, we needed to have two different exports of maps. Arnold and Unity require two different kinds of maps because Arnold has a metalness map and a roughness map whereas Unity only has one for the two maps called a metallic roughness map. When we went back into the Maya scene, we had a bunch of materials already on the model that were Lambert shaders. Since we were rendering the scene using Arnold, which is not Maya's original render, we needed to use a shader that would show up in the Arnold rendering system, in this case aiStandardSurface. Additionally, when we exported the maps in Substance Painter, we exported them for the aiStandardSurface shader. When all was said and done and we finished with Frank's tutorials and had everything rendered out, we noticed that there was an issue in our Maya scene: the UVs were a little bit turned around. One would just think to themselves: oh, just turn the UVs around. NOPE! The issue with that is that when you do it, the entirety of the map is already laid out and the textures do not move with the UVs. It is like if you were to cut out shapes in dough with cookie cutters and then wanted to move the cookie cutters around the dough in the parts that you already had cut out. Once all of that was fixed, we were able to properly render it in Maya and reimport the model into Unity. When we were putting it all into Unity, it was helpful to put the model into one mesh because if we had not, we would need to apply the same material to multiple pieces. With complex models, that can be irritating and tedious, thus, make the one mesh and have the one texture map with all of your materials. The texture map is another thing to denote as well because when everything is in one map, there is less of a render time when you are functioning in a game setting.

Overall, this tutorial was an easy one to accomplish, but a lot of steps that were important to learn. At the end of it all, I have a few more skills with Substance under my belt and a few more ideas of how textures work when you export them for Maya and Unity.