Do this before I forget

This is a postmortem for Twisted Theater. I'm writing this partly as a mental dump but mostly as a remembrance on this process as I will probably reference this article in the future. PS. I will be interchangeably Mr/Dr Hyde name.

After a big project it's important to air out the room while things are fresh. I mention things are fresh because I just woke up from a 5 hour nap after staying up 24 hours to complete Twisted Tower. Now, from the top. Joppee (https://dgjoppe.itch.io/) and I, thought to build a longer development game outside of GameJam time restrictions. That didn't go so well and after many months not much had developed. 5 months go by and no real progress on my end. There were too many things to learn and develop. Somewhere along the way I suggested another GameJam and out from the ashes rose ScottBros, a Marioesque knock off. Despite the janky animation I think I did a good job with level layouts. It wasn't too hard nor too easy and had some fun Mario easter eggs. Looking back at it now, I'd call this a success. This was middle of December. By March we agreed that our long dev project should be axed. This brought both relief and opportunity, no sooner did we axe one project we picked up another. PS1 GameJam style sounded perfect. It was a style I'd like to try and we both appreciate that era of games. 

Joppe had suggested to do a semi  horror game, inspired by another game called Twisted Tower. I played the demo, it was fun, I think it was by a single dev. I agreed to the idea and we went full force on it. The original idea was a man who got captured by a bunch of Thespian outlaws. He must go through 4 Plays (Romeo and Juliet, Les Miserable, A Western, and Noise off) to win back his son. As I was developing this story, script and all, I found a lot of pleasure building this world and creating corky characters. The antagonist called Mr/Dr Hyde is a great but vain theater actor. The kind that memorizes and quotes all the positive reviews but unable to accept the reality of his world. Don't get me wrong Hyde is an excellent actor, considered the best in the world. He has been propped up his whole life that he probably considers himself a god. Think of The Menu (2022) level of snootiness with the psyche of Birdman (2014). The protagonist, is simple called Main Character, and he is looking for his son. As you might guess Twisted Theater is Mr Hyde's story and the protagonist is the players point of view as this story unfolds. Yeah, I was juiced to created this world. 

STORY

There are a bunch of missing elements that really would have put this whole thing together. Due to our 9 day time constraint that wasn't possible.  After the first cutscene, the player is handed a gun and asked to attack the enemy. Instead of shooting the gun, the first attack the player uses it like a hammer. Why wouldn't somebody use a toy gun like a hammer, it certainly is more effective. This does nothing to the enemy. You hear Mr Hyde say something that these are the highest calibor of actors. They will go t any length, there is no limit. And Hyde simply demands that you shoot. When the player does so the enemy kills over. The idea is a toy gun or sword might as be a real one to these actors. Once that is over Mr Hyde scaffolding comes loose slightly and he has to adjust himself to hold on. He then commands the player to go through the doors. The next thing we see Mr Hyde put on a wig and transform into a young women( Juliet)

Man: What are you?

Dr Hyde:  Don't you mean who am I?

Dr Hyde: " I was blind, but now I see. A chameleon of our generation."  -  Le Monde

Dr Hyde: "Incendiary" -  New York Times

 Dr Hyde: "A Thespian, A genius in which all are molded"  -  Mother

Hyde opens some very tall curtains bleeding white light and disappears in the light. 

This is where the player starts walking through the hallways of credits to the white light. 

In the next scene the player finds themselves on a stage.  Where a cutscene of Mr Hyde speaking, 

" O Romeo, Romeo, wherefore art thou Romeo?

 Deny thy father and refuse thy name? Never!

 O cursed name, Montague! What fool would wear

the guise of mine own foe,

And think love’s whispers make a traitor true?

Thy presence here doth mark thee false of heart—

A thief that steals into his rival’s keep!

Ho, guards! Awake! "

When the player has dispensed with all the guards another cutscene of Juliet jumping down to fight "Romeo" In this cut scene you see an arrow just barely missed Mr Hyde. After Romeo and Juliet the gameplay gets pretty standard. During Les Miserable you shoot a bunch of French soldiers to proceed. The same goes for the western. For the last play Noise Off a big puppet of Mr Hyde  would be fun to fight, I'm thinking Serious Sam level big. At the end of the puppet defeat some loose scaffolding falls and lands on Mr Hyde. Here you get a similar ending to what was in the game. 

If you couldn't tell this was already a setup for failure. But I genuinely  thought we could pull it off. The gun system is a repeat for all three levels and only the sword system for the first stage would need to be built. The game was too big and too complex for 9 days. Even with a team of 5. Most games Jams you don't get to utilize those 9 days. The first 2 is idea and team gathering, including mental and job preparation. Stuff only starts rolling after day 3. 

*Starting to get tired, I'm only mentioning this because I noticed I'm making a lot more grammar mistakes then normal.*

In a lot of my GameJams I'm the only 3D artist. Often one of the Devs dabbles but that is about it. This time around I though to bring in some extra help. We brough on AJBartArt for assisted 3D and Craig Henderson for music. With the gang ready we were off. This was AJ's first gameJam, I believe so I made no assumptions on his knowledge level. We jumped in VC and I showed him my Asset process. (See Workflow). Around the same time I touched base with Craig. I having so little knowledge on music dev I just let him do whatever makes him feel good. Craig mentioned that a general vision of the cutscene would be helpful. This turned out to be surprising important request. It really gave me a deadline and laid the foundation for the whole game story. In 4 days I rushed through getting the Cutscene in shape.  Some of the characters like Mr Hyde wasn't completed so the Dad stood in for both characters. I also did a vis-render of what the Credit room would feel like. Once again, documentation is clear communication. By this time AJ has proven he is an excellent 3D arts and had two characters completed. By this time we had 5 days left, and we only had Stage 1 half completed out of 4, and  4 characters out of 7 protentional. Around this time AJ mentioned a Trap mechanics for Level 2,  that decision was backed by Joppe and myself. I'd like to mention when I make games I typically have a general framework of common games. I do this to make it easy on the devs as I don't know what they can accomplish. So coming up with a more exciting idea wasn't off the books. Needless to say it wasn't completed. We dropped the last two worlds entirely and I pushed to complete the ending cutscene. I put more effort in cleaning up Level one and two. Of those two only Stage 2 got added to the Jam Build before the deadline. I enjoyed making this game. I find the art style having a high tolerance level for mistakes. 

BENEFITS OF GAME JAMS

When I first started gameJamming back in 2020. My skill just skyrocketed and I saw similar results over the 2nd and 3rd game.  But after about the 5th game my skills started to plateau. After wards I did roughly 1 or 2 a year at the most. Its nice to get something out of these Jams. And I got a lot out of this one. The most important one is game documentation. I was getting frustrated with people not being able to see the vision that is so clear in my head. And as far as I know so clearly and eloquently described. Joppe convinced me that I am not as "clear" as I think I am and encouraged me to write my idea out. Simply just describing what the player does on each level worked wonders for communicating. Clearly communicating was important to me, as I considered this to be my game. And one I really wanted to see come to life. Now, I probably wouldn't put communication so high on the list but for some reason this really clicked in my head. I started applying game documentation to another game I very much wanted to make. And because of its some what complex nature documentation would be ultra important.  So Ultra clear communication is right on the list. Another important on that list is the LOD (Level of Detail) for game devs. As an artist I wouldn't want to give the devs anything that wasn't pretty much finished. But after speaking with David I understand how important it is for the devs to have something to work with. As David explained, give them the stuff as its being made.  This is why I call it the LOD of dev. Because the first thing you should give them, once done, is the base geometry character model and animation if it has it. This is a different workflow with multiple benefits. First it helps establish a workflow, from Blender to Mixamo, to Unity. Second it helps the devs out, as constraints, targeting and triggers, require some time and precision. This is a similar idea of box modeling a level for them to start the process. 

*Side Note* When you export the FBX from Blender don't Join the whole model. All them to be segmented so the devs man move stuff around.  

WORKFLOW

BUILDINGS

Looking back at it now the things that really help get the ball rolling was finding a good work flow. It typically went like this. Search the internet for inspiration. I like to use Google Maps  and search for places like Paris old town district. Take a screen shot, go into Affinity Photo, clean up the images, downsize to 256 or 128 and export PNG with Near directional and Web quality. I'd then go into Blender, import Image to Plain. Place a cube behind it and turn on x-ray mode. This allows me to see the points behind the image which allows me to box model it. Once all the points are where I'd like to be I'd Shift+ A in Faces mode to grab all the faces. Press U > Project from view. Apparently Blender things Project from view is an important feature as it is now the first option in that category. 

*Side Note: textures work best in a square so 128x128. When I used a different ration I noticed the textures would stretch. But I haven't tested this outside of one instance so I might be wrong. *

Once the model is done i'd export as FBX. A few good lessons. 

1. Under File > External Data> Pack Resources doesn't seem to do anything. It will say that (how many textures you have) was packed. But shows no difference in process when importing into Unity. I attempted this to try and resolve missing textures issues. This issue happened when I imported an FBX and all the textures were missing. So I imported the textures, which worked, until you separated the FBX and materials. This issue was fixed when I when I went to Materials inside the FBX. Changed "Use Embedded Materials" to "Use Local Materials (Legacy)" Hit apply. Then switched it back to "Use Embedded Materials" and hit apply. This became my workflow process as it seemed to get the best results. 

2. For some reason I'd get these py. errors when I attempted to export FBX with Light sources, like area lights and spot lights. I just removed them and everything worked fine. 

3. Exporting to FBX worked best when I left all the export options to its default. Options like, select only visible never seemed to do anything and when they were selected never exported a successful FBX. 

PEOPLE

For People it was just about the same process, except instead of Google Maps i'd use https://thispersondoesnotexist.com/ website. It is a great way to using a face that is essentially not real, or in this case doesn't exist. I would find clothing. I'd typically use different sources for the top and bottom and head. Often times the head had multiple sources because thispersondoesn'texist website will often cut off the top of the image, leaving me to find a haircut as well. In my example case, no hair. Bring them into Affinity photo edit them. Place textures at 256 x256 and export, Blender, then FBX. I'd take the FBX into Mixamo, download the animation with FBX to Unity, 24 frames and  with Skin on. I'd take these into Unity change Rig settings from Generic to Humanoid and call it a day. 

For my head example. You'll noticed the skin color extension. I originally thought I could just duplicates the image, scale and add a gaussian blur. But this never gave me the results I needed. I would pick a general color likeness and paint the outline. Id then eye drop the rest of the colors and better blend.  With the female face,  the face, top of the head and chin/neck are all from different people. *Side Note* The first person I made was the Main Character and set him to 6feet so that everybody will have something to scale. 

CUTSCENES

Once you have a workflow down the art seems to be no problem. But, like any good project there was new stuff I wasn't so familiar with, that being creating Cutscenes. I created stuff when I used Cinema 4d, But that was much easier to add motion compared to the jank that is Blender motion. The biggest difference in Blender is putting the proper time frame, making sure I used MP4 and  super duper make sure to change the file name between each take because Blender write over it. I'd then Import the video file into Devinci Resolve, which has a powerful video editor that is similar to Adobe Premier. 

BLENDER COMPLAINTS

The cutscenes are vital to the story and game. I feel like I wouldn't have too much of a problem but Blender is God awful when it comes to motion. The Noise modifier in the graph editor is wild and feels unpolished. Which blows my mind because random noise is extremely important when created quick fun animations. Another complaint would placing objects on a spline with equal distribution then have object animation on a loop. This should be a basic feature for any spline it should be built directly into the spline controls. But Blender this the wrong attitude of not making any assumptions on what people are doing with there project. So instead I needed to watch a video on Geometry nodes, lame. I guess a lot of my frustration comes from how easy it is in Cinema 4d to add basic animation and motion to things. Another big complaint is Blenders terrible Camera system. You maneuver your camera into position go to View> Camera > Active Camera to see through it. No problem. But what if you have two cameras? If you are wanting to use the second camera View> Camera >Active Camera doesn't work, and in most cases breaks the camera Location. Instead you have to right click and say Make Camera Active. Why? I feel 98 of the the users using a second camera is because they want to look through it. Another big complaint is if I press on create a new camera it creates it at 3D curser. Instead of creating it from my viewport. The Viewport is the easier thing to navigate, why wouldn't create camera to viewport be a benefit. I understand there is a shortcut for all my complaints but that just highlights my over all complaint of that Blender doesn't want to make assumptions on the user, when they really should. Blender does this all the time, node wrangler is a good example of a third party being incorporated. Before people start breathing down my neck, I donate $120 ish dollars a year. So I feel like i'm allowed to complain about it. 

DAVINCI RESOLVE  COMPLAINTS

When you want to delete a shot it will delete everything blow it. I think that is a big weird as it must be common for editors to mix and move shots around. Constantly deleting shots blow it, including the music must be a pain. I got around this buy right clicking > Cut option. I also noticed there connect or magnet between shots isn't very strong. Id place a shot down just to notice I left like 1 frame of space between. 

GITHUB DESKTOP

I always said not one artist ever touched GitHub during its development. Before I forget here is the GitHub process. Make changes in Unity, commit those changes, Go to Branch> Update from Master (make sure your on your branch). If there is no conflict Push those changes to the Origin. Change from your branch to Master branch. Go to Branch> Merge into Current branch. Drop down to Merge your branch, make sure your on the Master, Merge. Afterwards switch back to your branch.