Mark Eastland's DMUGA Blog: DMUGA Week 26

This week was a rather interesting one! We re-visited Bradgate Park, obtained feedback on our 3D work, did a very shaky (rather terrible) presentation and finally got round to doing some digital painting... Of which is of course very basic (I'm trying to learn!).

Monday - Visual Design

The whole point of today was to walk around and obtain textures (i.e. trees, rocks), pictures of the general landscape and draw whenever we took a break. In terms of the drawing aspect, one of the things I have struggled with from day 1 was drawing on the spot getting the perspective right, but throughout this term I have seen an improvement in this. Of course it's tricky from certain positions and I'll have to work on that, but for the first time I tried to just confidently mark the tree's form in a relaxed manor and it turned out quite accurate! I'm hoping this wasn't just luck and that it will continue, but this is the drawing I started when I was there:

Started the tree when I was there - Approx 30-35mins

Afterwards - focused on a different rendering technique - Approx 2hrs

For the texture aspect, I am disappointed to say I didn't get as many as I would have liked; this slipped my mind at times, but other times I was reluctant to take some because the lighting was so harsh it wasn't really worth it. In retrospect, I could have taken them anyway and used them as reference.
Using one of the pictures taken, I thought I'd at least try my luck at doing something digital, it's the only way I'll ever improve so why not? In my opinion, it turned out alright and I tried not to spend that long on it (1hr 30mins) because I have seen what level you should be getting to within this time. Obviously, it's not exactly a masterpiece but it's a start. There is a stage at which I get stuck at currently in all of the limited amount of painting I've done and that is detail. I'm not that bad at blocking in where things need to go and getting the composition ok, but for the life of me I cannot seem to add worthwhile details. I guess it's making me annoyed at myself which leads me to watch videos on how other people do it and then attempting it again.

Bradgate Park digital painting

Space centre digital painting (Background obviously made up)

The day itself was really enjoyable and I had a great time. Chris shared his current daily routines and it motivated me to try and do even more than I am right now. If someone else who works as hard as he does looked at the amount of time I technically have to do 'whatever', they'd probably say I could schedule in a lot more practice time. My quality of sleep has always been bad and it takes me a very long time to even drift off, so I was forced to change my 6hrs window for sleep to 8hrs as I was making myself ill by staying up until 2 and getting up for uni at 8.

My current routine every single day that I don't have timetabled lessons:
-Get up 9-10am-eat 1st meal, catch up on daily stuff, eat 2nd meal at 10:30-11am, go to gym until 12-12:30pm, eat 3rd meal from 1-2pm, (WHEN I SHOULD START WORK+Buy any food I need), start work around 3-4pm and carry on until 10:30-11:30pm with a lot of food breaks, try to sleep around 12:30-1am. With this I manage about 6 hours per day of actual work.

Having written this out, my problems are:
-Having too long breaks
-Not starting when I need to
-Gym in morning?

This is turning into a rant now, but I'll write about my ideas on what I can change at some point as I have an idea on what I could try. One of them is to just get on with it!
On a side note: I thoroughly enjoyed chatting to Chris about progress in weight training and dietary aspects, again, motivating me not just in work ethic but training too - Thank-you :)

Tuesday - Game Production Assessments - Addressing feedback

I know where things should have been better and I'm always open to any advice to better myself and/or my work. I am also 100% for the internal locus of control, but for some of the problems that arose, I really couldn't have done anything about it, or acted based on what I've been taught in class.

Generally, I thought it wouldn't be too bad. Nope. A whole host of things were problematic.
The main points were:

Character/Gladiator=

-Body texture looked "chaotic" and "patchy".
In the design document I wrote how 3DS Max wasn't liking the viewport canvas and so I had not choice but to submit it with all of the horrid seams. If I knew another way to get rid of them then I would have! Teach me please!!! I spent so long trying to get it to work, it left no time to make the diffuse better, but I'll try to correct the 'patchy-ness' next time.

-My scheduling was off.
I don't know what the optimal timing is for the stages in character creation, plus, I'd never rigged before and our tutor told us we must leave a lot of time for rigging. I gave it some leeway and my timings were generally pretty good. :)

-Naming needs to be more accurate + need to hide or delete unwanted items.
The only thing I can possibly think of for 'unwanted items' is the 3 lights, so that's a fair comment. Also, I name everything, so it can't be that and all my textures for this project were named fine in my opinion e.g. "Armour_Diffuse_001".

-Rigging needs immediate attention.
Honestly, the majority of it is ok, the real 'bad' part was the shoulders that are in need of major attention. The bottom leather parts had to be connected to his legs otherwise it would look odd with his hamstrings just clipping through everything.

-Over budget.
I know for a fact that in terms of the person, I was definitely under 2500 tris. The 300 tri budget for props was obviously over. I only did that because our tutor said it was ok - Why would I not want to create more assets if you have the 'go ahead', that's practice?

Transit Van=

The feedback for the van was a lot more reasonable and I know where the errors were.

-Should have used a composite material for the logo on the side of the van.
I know now, at the time it was too late to change it so quickly tried to hide the problem using the UV's.

-No specular map for outside + don't rely on crazy bump to create it.
There is a specular map, labelled "Van_Exterior_Texture_spec_001" and I make all of them using Photoshop from the diffuse as a base. In this case, the specular wasn't very good so I need to fix that.

As you can tell it wasn't the best day for me.
I later spent around 4 hours making the presentation all over again for tomorrow, and another 2 hours to script/practice.

Wednesday - Critical Studies

Wow, so after the big rush to re-do my whole presentation (first one had no theme to it) and practice it... I choked. I felt okay about the presentations and I was moderately prepared, but as soon as I started talking... Nope. I don't want to re-live it in my head again and describe the details. It's not like I've never done an important presentation before, although, I definitely haven't done a 'pecha kucha' styled one before. The 18 seconds for every slide may have been the reason why. 18 seconds was fine for what I needed to say but between the *thinking what I have to say* and saying it; caused me to go over to the next slide and speak even faster and just making it worse for myself.
That was the experience very briefly as I'm already feeling embarrassed again.

The rest of today was dedicated to doing some drawings of Bradgate and I finally got round to carrying on with the exporting of my assets into UDK with collision meshes.

Another Bradgate drawing:

UDK Exporting

Having successfully exported the bin project into UDK already, I assumed all the others would be pretty much the same process; they were generally (In the end). Although, after I'd created all of the meshes and begun exporting them, I hit a problem. My collision meshes weren't working. In my major panic to get something submitted on Friday, I just used ASE files instead of FBX and that fixed the problem. It wasn't until Thursday night that I figured out that the collision mesh in FBX format has to be named exactly (case sensative) the same as the mesh it resembles. *Facepalm*. I then went through, renamed and re-exported them. As usual, there's always one that doesn't want to work: