• Share it :

Off The Map: Post Mortem

3ds max alice in wonderland character artist engine game art game art design game dev post mortem level lighting off the map OTM UE4 zbrush It's safe to say that this post will probably be extremely long. Brace yourselves. The rant is coming.

I'm not actually sure where to start. I absolutely hated this project. I've found that over 3 months of doing exactly the same thing isn't for me - which I guess means I'll definitely be better off in a smaller studio! We had a strong idea that wasn't executed well, which made me miserable. In essence, the last term was spent being super miserable. But, that's life I guess. You'll get stuff thrown at you that you don't really want to do, but you've gotta man up and get on with it anyway.

Throughout the course of this project, working within a team of 6, we were required to make a functioning game level based upon Alice’s Adventures in Wonderland. The three topics we could choose from were: Oxford, gardens or underground. As a group, we decided to go with the underground theme as this appealed to all of us and seemed like some interesting work could be produced from it. 

A week or so before the project started, we all had to fill in a sheet and specify the roles we felt most comfortable in. This was to ensure there was an even split of people within teams. However, in our group, we ended up with 3 character artists and 3 environment artists but we worked around it and all chipped in on other areas - such an engine work. 
As we had 3 character artists within the group, we had to decide quite early on who was going to be modelling who. I was given the White Rabbit to model which I was both excited and nervous about as I’ve not modelled anything non-humanoid before.
However, as I managed to get him finished much sooner than what I'd initially planned, I took on a few extra assets and roles in order to lighten the load on my other team members. We needed a lot of assets to populate the level as it's so large, so I made a variety of picture frames, candles, lanterns, pocket watches etc. I was also in charge of modelling the skirting board, wooden panelling on the walls and the coving.

In addition to the character and extra assets, I also contributed to other areas of the project, for example:

• Concepts: Moodboards for the characters, style guide, book, warped assets & rabbit clothing. Concepts for the rabbit, final door and a storyboard.
• 3D Models: White Rabbit (Modelling, unwrapping, texturing, rigging, skinning and animation in Max & UE4.) 4 picture frame variations, candles, lantern, pocket watch, skirting board, wooden panelling (thick & thin variants), wrought iron door and the chest.
• Sculpts: 3 beam variations, candles, 3 frame variations.
• Unwraps: Steps, columns, lobby decor, stair supports, railings, ceiling supports, raised platforms, beams, chest and both sets of moveable stairs.
• Light Maps: Chest, 6 bookcase variations, camera, ceiling light, chandelier, desk, fireplace, cutlery, key, pen, plates, rose, wall light, furniture, desk lamp, rose vase, book, chest, final door and both sets of moveable stairs.
• Textures: Clean tile set, dirty tile set (for vertex painting in engine), 4 light masks, flip book texture for the candle flame, rabbits fur, rabbits body, pocket watch and normals for all of the frames.
• UE4 Materials: Rabbits body, rabbits fur, rabbits clothing, gold, glass, pocket watch, candle flame, subsurface for candle, frames, beams, desk, frosted glass, Alice's clothing, Alice's hair (subsurface), Alice's skin (subsurface) and the chest.

I also helped out with the population and lighting within the study area of the level and the population of puzzle area 3. The general level itself was populated by everyone in the group. 

Mystical floating candles..woops

I felt far too stressed out for the whole of this project, I'm not really sure why, but this was definitely the cause of my 'lack of enjoyment'. We just seemed to have one issue after another, which led to James and I staying up until 3am towards the end of the last week fixing things that should've been done weeks ago. For example, our playable character is Alice. Guess when she was put into engine. Yep, Thursday evening. A few working hours before the deadline. James had been asking for the assets for weeks, but it just didn't happen. However, I managed to get her working correctly with the default UE4 animations and replaced the blue dude in the third person blueprint. It was actually quite fun retargeting bones.

Whilst working on this project, we often had people telling us that it was looking really cool - but it was such a huge level. We found out that one of the judges is big on architectural accuracy and luckily for us, we had Jack helping us. He helped James out a lot with the overall structure of the lobby - especially the stairs and surrounding supports. The lobby is definitely the best looking area in the game, it still needs a lot of work, but we're developing an action plan of what needs to be done over the next 6 weeks to make it look fabulous. As well as receiving feedback from Jack, we also had weekly meetings with Steve and Chris to ensure we were on track.

Overall, I'd say our communication (or lack of it) let us down. Communication is such a vital thing in a project as big as this, and although we had a lot of it, it obviously wasn't enough. A lot of the time, it was unclear what people were working on throughout the day and the asset list I created on Google Drive wasn't being updated by certain members so we were unsure what stage people were at. This was a pain when people weren't in labs, or working on different floors etc. We worked around this though, and tried to play it to our advantages. Regular team meetings were - and still are - essential.


After receiving feedback around two weeks or so before the deadline, we realised that we really needed to cut down our level. What we had was underpopulated and it was suggested to us by our lecturers that we cut out the less important areas, and focus on making the remaining ones look really polished. This was something we'd been discussing for a while, but it always helps to get definitive feedback from others. After receiving this feedback, we decided to cut the following areas:

• Initial study area - being cut down into just the desk.
• Room in which you fall into from the rabbit hole.
• Corridor following said room.
• Puzzle area one is being cut down into one room instead of two. 

Looking back, I'm glad we had to cut these areas. At its current state, our level isn't completely polished. I dread to think what it'd look like if we hadn't cut it down...

If we were to start this project again, I'd definitely make a strict time plan and make sure we had a definite team leader in place. Our main issues seemed to be time management and communication which ultimately led to our project being a little more unfinished than what we wanted. I feel that if we followed our time plan more closely, and had daily team meetings, we could've reached all of the goals we set out to do. The three-ish random team meetings we seemed to have each week didn't seem to suffice.

Overall, I feel this project could have gone slightly better. We were way too ambitious with what we wanted to achieve and didn’t organise our time as well as we should have. I’m happy with the final outcome, but there’s so much more that could’ve been done. In a way, I feel disappointed. We all worked well as a team, but there was no real team leader to ensure that things were being done by certain dates – we were all relying on each other to have stuff done by agreed dates, which often worked but most of the time, it wasn’t as successful.

I’m happy with how I managed my own time though. I managed to get the rabbit finished (textured, rigged & animated) before the Easter holidays. I allowed myself time to improve on the rig and refine the animations after he’d been tested out in engine which meant I wasn’t panicking about not having him in the final level a few days before the hand-in. This also meant I had extra time to help people out with their work. For example, I spent the majority of the last two weeks doing the light maps and collisions as they hadn’t been done along the way.

I also spent the last part of the week fixing up silly mistakes. Files had been named wrong which led to confusion when importing, the scale of certain objects were all over the shop and pivots were in the totally wrong place. These were easy fixes though, and I'd rather they were fixed than left to cause more issues.

Ideally, I would’ve liked to have spent the last week polishing up my own assets and maybe making a few more to help populate the level but fixing up the assets we already have is obviously much more important.

However looking back, I’d say this is definitely the one project I’ve learnt the most on. Not only artistically and technically, but I’ve also learnt more about myself too. I’ve achieved what I wanted to and even did a bit on the side too, but even though I produced a main character for the level, I don’t feel as if I’ve done a horrific amount. The majority of my time has been spent fixing things for other people, which sadly, has meant some of my own work is slightly lacking. I've still learnt from these fixes though, which I'm extremely grateful for. It's nice knowing you've helped someone in your team out too. Yay.

In order to reach the deadline set for Off The Map, we'll need to put the stricter time plan into action. I went into the labs on Tuesday to figure out what extra assets need modelling and when it all needs to be done by. We've worked out that including this week, we've got 6 left. Which may sound like a lot of time, but if you include the fact that the labs are closed for the final week, it's not long at all. Welp. I've made myself a personal time plan so I can get all of these assets modelled within a week so I can get onto sculpting pretty things and Tom has also been super awesome and made one for the group. It's a week by week one and gives us a nice idea of what level we need to be at each day.

• Week 1 - Planning/organisation, bug fixes with models/textures and concepts/paintovers.
• Week 2 - Modelling of assets and concepts/paintovers.
• Week 3 - Sculpting/Texturing/Baking and fun engine things.
• Week 4 - Review of population and 'is it playable?'
• Week 5 - Finalising/polishing and OTM required documentation.
• Week 6 - Final polishing pass and packaging of the level.

The majority of the team has been in the labs this week anyway, and we've managed to get our level looking more finished! Well, here's to the end of second year. It's been a good one.