Friday, 17 April 2015

What went badly? Container City


Unwrapping and Geometry


Kieran right from the start was adamant that he was going to be on prop duty, we were fine with this and let him get to work. Kieran then proceeded to make around 70 individual assets and he unwrapped them. This was all extremely impressive at first. At this point we noticed huge issues with his unwrap and gave him allot of feedback at this level of the modelling process. A handful of re-unwraps later we had to move on and texture. At this point we noticed that the problem had not been solved. We also now had an opportunity to sea the geometry we would be working with. The geometry was far too low poly relative to our agreed quality of output. But it was too late.

This issue could have been stopped had we kept a closer eye on each other’s work. The problem here is that we were all very confident with each other’s abilities so we were all comfortable letting team members simply crack on undisturbed.

Difference of opinions 


We had slight disagreement with the design of certain areas of the world and how we would place certain assets. The series of bridges that connect the residential containers, the Buddhist shrine and the market in Joshes original set up were extremely stylised. Josh wanted them to have an exaggerated flimsiness to them; he wanted them to be perilous. This idea stemmed from the idea that the people that lived in this community could not, and their fore did not suffer from vertigo.


But the rest of the group found them unbelievable, so we pushed heavily for them to be altered and made much safer and sturdier. Josh resisted hard in this area, the bridges in their final state are a result of compromises from both groups. Sadly though these bridges are still one of our greatest visual stumbling blocks, people when they play through the game still question.

We had other similar problems with placing candles directly in the player’s path and whether or not this would lead the player, or if it was an obstacle a player would assume you wouldn’t walk straight over, thus blocking off a path.

Shadow Puppet Theatre

 Josh in this project was given the vast majority of the engine work, because of this we did set aside a single container for him to have his own little zone.

For me this was the shrine, 

Mike the Noodle Bar

and Joel the Lift.


Due to the pressures of being the sole engine guy in the team Josh didn’t get to creating assets for this area until pretty late. At this point we already had many assets ready, and some parts of the game were populated. I personally believed that we didn’t need another themed area and we already had enough assets to make a perfectly fine game ready space.


But instead valuable time was used making additional assets and more unique texture space. We already had issues with pixel density we didn’t need new additional props, especially props that had a single use. Again here we all had to compromise this room was going to go ahead with being a shadow puppet theatre, but a few assets that were modelled were not actually going to be used due to not having the budget or time to texture them.

Also right at the end the core feature of this room didn’t actually work. Josh wanted to have light and shadows projected on to a sheet in engine thus creating the shadow puppet theatre.  Sadly we were unable to achieve this finish; this has left this room barely readable. Again the problem here could have been limited by more and better communication through the group. We should have used existing assets, this would have given us more time and texture space for the assets we already had.


Texture Imperfections

There are three individual assets I can think of as having a low texture quality.

The yellow railings have little to no texture variation, they appear to be simply yellow, there is no rust, peeling paint, warn edges etc.

The Rubbish piles have not been alphered round there edges so there is an area of flat texture just visible at points in the level. I wasn’t responsible for rubbish piles, I noticed when looking at their finished versions that they could have done with a projection mapping pass, it was at this point that I noticed that they didn’t have a roughness, so I did this and then the metalness. Sadly I failed to notice that the alphers hadn’t been made either.

The lift interior was textured well if it were meant to be a clean brand new lift. I showed Joel how to produce a seemingly complex geometric pattern which would then be projected. As a result the lift has a really nice floor pattern. But the lift it’s self has not had a wear and dirt pass. As a result the lift floor does seem out of place if you happen to notice. 

Personal

I personally could have done a lot more, right at the start of the project I chose the task of designing, modelling and texturing the Containers. As these are going to be one of the core assets for the level I really wanted to do a good job here.


Also as this was the only core task I was given I ended up spending allot of time on it, as it was all I was given to do. I spent allot of time over designing the container this was time that could have been spent elsewhere.



(I forgot to put this in the things we did well Blog. Only a quarter of the container is unwrapped, as the container is symmetrical on two axis’s. This means that 75% of it can be over lapped, and only 25% of it has to be textured. This has resulted in a container with an extremely high pixel density)

This image shows the texture quality of the floor. The geometry is actually flat here, but I have used projection mapping to make the boards appear warped. 

given the importance of this asset and size this was extremely important. Or alternatively, in order to achieve this pixel density the container would have needed at least two 1024 if only 50% of it was over lapped, using the method I used the container and all its variant parts, hinges, wires, Wooden window shelf etc, are on one 1024)

Basically I spent too much time on every asset I did, but as a result they would look a lot better. If I hadn’t done a projection mapping pass on Joel’s metal assets they would be a mear shadow of what they are now.



I also spent allot of time using the cloth modifier on all my draping assets.



Again this only served to improve them. Also I completed all the tasks I was given and then had to get tasks off other members of my team, all of which I completed to a high standard.

One of 3 posters. This is a photo bash of an amazing but mostly unknown Russian Statue and Glastonbury Festival. I've also tiled military Jet planes across the sky and included the Chinese Symbols for Infantry. I was really happy with this, I wasn't comfortable to begin with as Graphics isn't something I have done in a long time. I really like the ambiguous history I have created here. 

Engine

The greatest mistake that was made during this project was set in stone on the very first day. We decided as a group to give Josh the sole task of doing all the engine work for this project. As a result he would spend the vast majority of the 4 week project working inside unreal. As we now know, and all have personally experienced, prolonged exposure to Engine work kills you. Engine work can often make you feel like you are achieving nothing as the majority of your time is spent doing failed experiments, or making something and then inexplicably and seemingly permanently breaking it. Leaving you feeling like you’re a dead wait on the team, although this is not true. Engine work has a bumpy learning curve; all the mistakes and frustrations have to occur.

Because of the seemingly endless queue of tasks and requests from the rest of the group, when doing engine work all you want to do is go back to some more traditional modelling and or texturing. This is probably what led to the Shadow puppet theatre.

The engine work for this and all upcoming projects has to be shared out; one person cannot spend an entire project in engine. The decision we all made on the first day was the cause of the majority of the grievances during this project.

Although

All the problems that I just listed are all relatively small issues. Although giving Josh all the engine was the greatest mistake, Josh did pull it off.


The texture inconsistencies are mostly hidden in the level. Also these assets at least do add to the bricker bracker. They also are textured with all the maps. There are no flat textures on any of our assets, and their definitely are no untextured assets. The vast majority of any scene you see in our level is textured and modelled well.

The shadow puppet theatre is still a shadow puppet theatre and if you were to take time in this area you may be able to discover this, it insinuates there were kids here, and this adds even more depth and life to the world. 


The bridges although sometimes questioned, they are mostly liked. They have great silhouettes, and they add a level of danger to the world and they tell you something about the lives of the inhabitants as Josh intended.



















No comments:

Post a Comment