Wednesday, 5 November 2014

"All by myself, don't wanna' be all by myself..."

This is going to be a quick post mortem post on the film project, since it's been about a week now since we handed it in and I've had some time to reflect on it. There also might be the occasional sad comment about how I've now been set a solo project, and it feels a bit weird not working in a group for the first time this year.

I think the best thing to do would probably be show the original screenshot up against the finished product and then go from there really, so here it is.

So when I look back at this now so much jumps out at me as being really rather inaccurate. Once finished, I couldn't wait to push it aside and begin something else after spending 3 weeks straight on it, but now I look at it and wish I could spend so much longer on it to fix things. At the time, I have to say I was very proud of what we achieved. It may not be the most uncanny resemblance you've ever seen, certainly not as close as some of the other groups managed, but we did it in the time given and the models and textures were all pretty solid, even if the lighting wasn't so much.

One of the best parts about our project was how well we handled time-keeping for the most part. I worked out a system and certain deadlines for aspects of the project, and everyone pretty much stuck to them. It made us so much more organised, we hit the deadline with time to spare, and meant that we all were at the same stage throughout the project with each of our individual sets of models so no one got left behind.

Our teamwork ethic was pretty good as well. There were no major fallouts or arguments between group members at any time, and we all felt comfortable enough to critique each other's work when needed even though some of us had barely known each other before the project. I think the fact we weren't all necessarily 'chummy' with each other before we began the project helped us keep a sense of professionalism when discussing the project, but also meant I got to make some new friends (YAY FRIENDS).

And finally, another good part of the project for us was the models. It's hard to see in the final render, but we produced some models which were pretty spot on in terms of colour and patterns etc. The lighting has made the colours appear all wrong in UE4, but before we placed them all together in a lit scene the textured models looked good in my opinion. 

Right, and as for the things that didn't go so well, firstly there's the communication side of things...We weren't terrible, but we weren't great either. At the beginning of the project, barely any of us were in labs together, and the amount of talking we did in person was limited. We used Facebook for a lot of our chatting and feed backing to each other, but I found the most important information and critique we got was when we were talking face to face or working alongside each other in the labs. We had a couple of issues where people hadn't registered certain information that was given on Facebook and ended up doing things in a way which we hadn't intended or wanted, and I feel like this would have been avoided if this kind of information could have been given in person instead.

Another big thing, which I now realise makes or breaks a scene, is the lighting. I can't even believe how much of a pain it ended up being. I'm not particularly good at UE4, and I still have a lot to learn about it, but I had even assumed I'd be capable of managing some sort of accurate lighting within the scene. Turns out I didn't have a clue, and even the team members who were better at engine stuff struggled. We waded through iteration after iteration of the lighting, some making the scene turn a pinkish colour, others making the floor appear brown, and sometimes even without changing any lighting elements at all Unreal would decide to completely change the colour palette of the scene. It was just a case of getting it as close as we possibly could, because we realised quite quickly that we were never going to get it spot on. There's shadows cast in the wrong places, light sources look too strong and not strong enough in certain spots, and my main issue with it is the change in colour of practically all the textures. All the models we'd made look so accurate have now been changed completely for reasons unknown to us. The floor was probably the biggest let down, as not once did it appear the colour it was supposed to, and what was supposed to tie the scene altogether now screams "I'm wrong" every time I look at it. But we tried our absolute best, and if we could have fixed it I'm sure we would have done during the amount of time we spent fiddling with it, so I remain satisfied with our efforts.

I learnt a ton from this project. I picked up some experience of organising a small team, and attempting to communicate and synchronise workflow. The asset swap required you to work at a certain pace because we needed to pass assets on to the next person and so on, but this project didn't have prompts like that. It was up to us to assure we go everything done on time, which needed planning and making sure we all worked at a similar pace, something I haven't really had experience with before.

Another thing was just becoming technically stronger with certain aspects of the modelling process. I explored more to do with PBR, became much better at organising folders in Photoshop (I was dreadfully unorganised before) and got a little experience with Unreal 4, which I now need to explore much further to get a better understanding. All these little developments in my knowledge make me feel good knowing that I've already learnt quite a lot after only a handful of weeks back on the course. I've learnt so much, and feel like I've improved so much already that I can't quite believe we've been here for as little time as we actually have.

And so we come to the final part, my thoughts on what we'd do better next time if we could. Well, for obvious reasons, I'd make sure I made specific time slots that we all had to be in labs together. I can't even explain how much it helped at times to have one of my team members pretty much with me the whole time I worked in labs, I only wish the other two members had done the same and I think we'd have been so much stronger. I feel like it would also have been a good idea to get everyone's numbers as well, because when Facebook failed, and no one was in labs, a quick text might have got things done a little quicker.

AND LIGHTING. I mean, I'm saying that we could have done better lighting wise, but all I could have done really was dedicated more time to it. We spent a good 4 days or so doing the lighting, and I now realise that probably wasn't enough, but I'm not sure how we would have given it more time. We could have thrown in an extra day by cutting down modelling time or something, but I'm not entirely convinced an extra day would make all that much difference. We'd have needed a week more for all I know to sort the lighting out to a degree we were happy with, and if we redid the project I know we wouldn't be able to squeeze an extra week of it in. I think I need to do another project with a heavy focus on lighting to get a better jist of how long it takes to sort it out , and then I can base my future schedules and spreadsheet breakdowns on that instead.

But alas, I must let this go now and move on to the sentry project alone. *sniffles* farewell group work.

I've realised I said this was going to be short and it's far from it so I shall leave you now with a subtle hint of what is to come regarding my sentry concepts... I LURVVVVV OLD SEWING MACHINES (especially cute toy ones like this)


No comments:

Post a Comment