Genius Gallery: A 3D artist's LD58 postmortem - Part 2
Part 2
If you missed Part 1 of my postmortem you can read it here.
(Once again, if reading isn’t your thing and you just want to see the pretty pictures, scroll down 😉)
By the time I had finished with the Battle Models it was already Saturday night and we were a good 30 hours into the jam. The environment kit and first 10 or so exhibit assets were finished. It wasn't a bad start, but time was ticking away and I still had the massive still-mostly-unfinished list staring at me. Thankfully by then the list was finalized and nicely sorted into groups, which was a huge help to me. This made it much easier to parse so I could eat away at it in chunks.
Since I had already spent the entire day at that point doing nothing but modeling, I really wanted something -anything- that would give me a break from staring at Maya's interface for a little bit. So I decided to start on the only exhibit assets on the list that would require actual texture work instead of using the texture atlases I made earlier.
Books & Scripts 
I quickly modeled a map/chart that would work for the blueprint, treasure map, and star chart assets. Next I made a simple open book model. I also realized I'd want to make a quick display stand for these that would make more sense than the other exhibit bases I already had. After a quick UV layout for each of the two main models I could once again open up my beloved Substance Painter.
There's something about the texturing process that I've always loved more than just about any other aspect of creating a 3D asset. Modeling can certainly be enjoyable, but for me it's ultimately dependent on how interesting of an object it is and how complex the topology needs to be. Then of course there's unwrapping UV's, which I can't imagine anyone actually enjoys. It's a necessary evil as far as I'm concerned. And hey, if you happen to find it enjoyable... well good, but also, who hurt you 🤨? On the other hand, texturing is where I truly feel like I'm making art. It's the process where everything comes together and takes shape. The icing on the cake so to speak.
Scripts texture atlas (left) & Books texture atlas (right)
Unfortunately texturing is also another place where I get way too wrapped up in what I'm doing and time melts away. I often find myself spending time on little details that no one will ever see or appreciate. Then again, I feel like part of creating art has to be for yourself. If you're not happy with what you make and you can't let yourself enjoy the process then what are you really doing it for?
My standard procedure when texturing multiple assets that use the same atlas is to bring them into painter as a group and texture them all at once, using folders with alpha masks to be able to work on one asset at a time while retaining the ability to export it as a single texture. I started off with the three scripts. For the blueprint I used a 3D voronoi fractal noise generator to get a quick blueprint style grid, then masked off the areas I didn't need. For the star chart I couldn't think of a quick way to make anything resembling a star chart without doing it by hand, and then I remembered that these things have existed for centuries so surely there would be something I could utilize for the texture that is free to use. And sure enough I quickly found a star chart from the 1800's and used it as a brush mask (public domain FTW!). The treasure map I ended up just drawing by hand—because I mean come on—who doesn't want to make their own treasure map?
The books were slightly more time consuming, but there were also six of them. After choosing different colors for the bindings that somewhat fit the theme for each of the books I went to Photoshop and make some text blocks using the tried and true 'lorem ipsum' with different fonts. Once again I knew this would never be seen, but dammit, I was in my zone!
Some of the brush alphas I made for texturing the books
A few little symbols and scribbles later the books were all textured. I exported the atlases, brought them into Unity, set their respective materials and applied them to the models and I was done. Just like that I had 9 more exhibit assets ready to push to the git repo, and it was only.. oh god, it was 4 A.M.! Well, time flies when you're having fun. But we were now just a couple hours away from the mid-point of the jam and I had only completed 18 exhibit assets. Sure, that's not a small number, but it was still just a mere fraction of the list.
Proof from the Git repo history that I was beginning to lose my mind
I knew I really had to kick things into gear. But approaching 5 in the morning is not the time to kick anything into gear unless you sleep in a race car bed. Despite not owning a race car bed, I still decided that it was time for sleep. I set an alarm for 3 hours, which I consider to be the bare minimum to remain functional for the next day.
Plants 
Shortly after 8 in the morning I woke up from my ~~restful night of sleep~~ extended nap. After a quick shower and a few gulps of an energy drink I was ready to get back at it. After a quick scan of the list I immediate knew what I wanted to tackle next: plants! Not only were there a ton of them, a few of them would be simple re-colors. Plus I love any chance I get to make foliage. Every game jam I'm secretly hoping that the theme or idea our team chooses will be filled with excuses for me to create a ton of foliage. And finally, my time had come! Unfortunately they wouldn't be used in a fun way to propagate an environment with to make a lush forest or dense jungle, but I'll take what I can get.
Besides, in theory they shouldn't take terribly long to make. Just sweep some bendy splines, make a petal or two and rotate them around a central axis, slap a little stigma in the center and baby you got a ~~stew~~ plant going! But no! This was my excuse to create some truly killer plant assets. I had to make up for all the lost chances of jams past and make super detailed flowers, individually modeled after their real-world counterparts and all with a stylized flair that made them uniquely mine.
So sure, I spent a bit more time making the plants than I needed to. But I also lost an hour to a team meeting (where I struggled to focus because my brain was in full plant-creation mode) so it wasn't entirely my fault. And right when I was back in the groove I hit the one 'plant' in the list that I glossed over. Enter: the Mandrake.
Technically a plant, I guess.
I feel like I need to make one thing clear. I am not a character artist. While yes, I am a 3D artist, and yes I have made characters in the past, and yes they've usually turned out alright, I simply do not consider myself a character artist. I'm more of an environment, props, and interactables artist that is occasionally forced to create a character every now and then and I dread it every single time. It's just not my thing, especially if it's humanoid. Sure, I may have taken an anatomy class in college but that doesn't qualify me to be a surgeon. Character modeling is no different. It's a very specific area of study and people who are really good at doing it typically do just that. It's their thing. It's not my thing.
Now, could I have slapped a few elongated spheres together, haphazardly tossed some leaves on top and called it a day? Absolutely! Should I have done that? Well, the jury is still out on that one. But that's exactly what I started doing, and when I was so disappointed with how it looked I exported it and used it for a base mesh. It was that time again... the time during a game jam where I prove myself a liar and open up zBrush after swearing "I'm not going to use zBrush this time" at the start of the jam. I was so close to succeeding, but this stupid little potato monster sold my soul to the devil once again (yes I know it's not a potato).
So yeah, there went another whole hour on a single asset. But dammit if he isn't a cute little bastard. Of course I had absolutely no time for a proper high to low poly bake and retopo, so even after some decimation master magic he still weighs in at a hefty 7.4k tris. But I could at least take solace in knowing that a single slightly sub-optimized mesh wasn't going to affect performance, especially when the entire project only uses 5 texture maps across less than a dozen total materials.
So after finishing up the mandrake, I knocked out the last few plants pretty speedily as I didn't want to waste any more time. Thankfully a couple were just recolors. When I finally got everything in engine and pushed to the repo it was closing in on 5 P.M and I had another 11 exhibit assets ready. And it wasn't even close to being dark out yet!
Now that I was done with the plants I could move on to other things. I decided to knock out something easy that could be re-used, so I made a ship in a bottle by quickly modeling a bottle and sticking my pirate ship model from the naval battle diorama inside. I also exported a version of the bottle by itself standing up so that it could be used for the 'lightning in a bottle' asset. It would just need some VFX made, but hey that was finally something I didn't have to do on the art end of things!
I was locked at this point and just started going full steam ahead into the next category: Beast objects.
Beast Objects 
I had already made the kraken tentacle from before the list got organized, so hey, I was already one asset down! Then it was a phoenix feather. A feather, that's it? Easy. DONE. A megalodon tooth? That's basically just a fat triangle. No problem at all! A dragon egg? Sure, that's just a stretched out sphere with some of the edge loops extruded to make it look kind of scaley. In hindsight I probably should have moved some of the vertices down to make it more scale-shaped but oh well. I was cooking and just banging out models at this point, so much so that I made a fancy little box for it to rest in, just because! A dragon claw? No prob—A DRAGON CLAW?! Dammit, not again... :sob:
Well at least it wasn't a full creature, just a hand, er... claw. Everyone knows hands are the easiest! (That's a joke for all you non-artists) Well at least zBrush was already open. After a truly hilariously poor attempt at modeling a base mesh of a claw I took it in zBrush and worked what little sculpting magic I had left in me. I think I spent around 40 minutes on it, which isn't terrible, but it was more time than I wanted to spend on a single asset right then.
After getting everything imported into Unity and pushing it all to the git repo I now had another 6 assets done, and it was actually only 8 P.M. this time. The night was young, I was up to 35 exhibit assets now, and was ready to burn the midnight oil!
End of Part 2
There's still plenty more to talk about and a even more assets to show so stay tuned for Part 3. Until then, please drop by our game page and give Genius Gallery a play!
Now off to play and rate more jam entries! 🎉