Week 30; From Humble Beginings

  Well, I managed to reach the end of my first year at university (relatively) unscathed! It's been such a blast I can't believe it's over, I've thoroughly enjoyed every part of this year and have made sure I won't regret anything that I haven't done this year. I feel like it's the first time where I've really wanted to work for something, which has been completely alien to me in my past experiences. Enough of the sentimentality though, I may tear up. However, I will leave you with a comparison of what I started out the year with, and what I ended up with.

Week 4 Sci-Fi Crate

Week_30

  So, as you may well know this is the summation of all the work we've done for the past year - the zombie pick-up truck. It's been great fun to model, however sadly my free trials to nDo and substance designer had run out, so I couldn't complete the whole PBR look, however I tried hand-painting the different maps (even though I fully well know it's nowhere near the same). What I did find out though is that UE4 have updated their launcher in which their new version supports Metalic and Roughness maps! I wish I had known this before instead of fiddling around with them in the previous project! At least I know now and I can fully utilize this function, which is great! 

  Right, lets move onto my thoughts on the outcome of this project. I'm indifferent at the moment. Having just finished it and handed in, I'm not really quite sure what to think. I think under the circumstances of having no substance designer and trying to make the textures PBR I did quite well. However I just don't like the design. It's just too bland for my liking. I should have gone back into the diffuse of the truck and fiddled around with the map there, however it wouldn't look quite right as the metalic and roughness map wouldn't register the extra information and I didn't have access to substance designer or nDo, so fresh blood I add to the body work may seem really dull and rust, for example. I've also learned a good lesson in not adding extreme detail to areas which wont ever be seen, for example, for the gear stick in the interior, I made on the normal map the inscription on top of the gear stick, however no-one is ever going to see or notice that unless I tell them it's there, which rather defeats the object. All this being said, I'm happy with the aesthetics of the asset, it's just the design I really have a problem with. I'm impressed that I was able to push myself to try and learn this technique in time to submit a project with it, and it'll only help me in the future. 

  Now, what I'd do differently. Firstly I'd spend a larger time concepting until I was completely content with the final outcome (although this may be unrealistic as I will need to make the asset some time, we'll see). I would also pay closer attention to the allocation of space I set out in my UVW Templates, as sometimes I give too muck texture space to a small detail. For example, the barrel on the side of the model looks hideous and I'd make the texture larger and just change the amount of noise in it, it really doesn't fit the top of the barrel and sticks out like a sore thumb. Obviously, with the great news that UE4 now supports metalic and roughness texture maps, there had to be a cog in the works. For some reason when I export my truck as an OBJ and import it into engine, the spikes don't show up like they do in 3DS Max (see flatshader). So if I were to change anything, that'd probably be first on the list. In all honesty I think that's about it, other than what I've mentioned before about the bodywork, I will conclude I'm actually quite happy with the outcome, as opposed to what I wrote before. 

  That concludes my first year. I have a plan for the summer which I'll divulge. I plan on working with my Father, who happens to be a programmer, in creating small app games and keep working on my 3D. I'm also going to be setting myself certain drawing tasks from life since it's summer, and I should be going to some nice places around England in which I can draw. I also plan to save up for a wacom tablet so I can texture from home as well as model, which would help immensely over the summer as I wouldn't want to fall behind on that! Other than that I'm going to be looking at game engines and what they can do more closely. I really want to be able to fully utilize UE4's potential as I currently feel like it's being a bit wasted on my understanding of it. I also plan to make my own website for my blog and move all the posts onto there, as I feel that it'll enable me to design it exactly how I want it and accommodate my requirements. I will also keep blogging what I'm working on over the summer as well as I wan't to stay in a good habit of blogging.

Thanks for reading!

Week 29; The Penultimate Week

Hello! I'm very happy to join you again after a stress free Easter holidays! In all honesty the last project I did had me completely burned out as I was effectively trying to do double the work so the break was much appreciated and I can come back into the final two weeks fully rested.

So to start off the week we had our presentations on our plan for the Zombie apocalypse. I had prepared this over the Easter and had been practicing the recital, however that still didn't prepare me for the uncharacteristic nerves that set in. I'm usually completely fine with things like this, I guess it was the pressure of it being a formal assessed piece to count towards my final grade. With that being said I think it went quite well, I had some key points from my last presentation I wanted to work on, one being conveying evidence to back up a point. I hope I did this effectively as I reasoned my concepts satisfactorily in my opinion, even though they were actually quite silly. My whole plan was based on riding a motorcycle which I have no clue how to do then riding to a high security prison half way across the country, so I tried to give as much evidence for why a person who is not completely insane would do that. I also tried to project my voice well and capture the audience as much as I could, trying to throw a bit of humor in with the whole silly idea of my plan. Despite the nerves that had set in a few days before the presentation, like I expected, it was painless and over in a flash and I actually quite enjoyed it! I actually wished I had a bit more than three minutes up there! One of the things I could have definitely improved on was the aesthetic quality of the presentation; it wasn't the prettiest. However I also believe that to keep an audiences attention they shouldn't be sidetracked with the imagery, there does need to be a balance, however I do admit I didn't hit the balance.

So that was Tuesday over with. The rest of the week was dedicated to concepting my Zombie modifications! The project I've been waiting for! So I started out with (after the moodboard) sketching extremely roughly over my base truck I'd made previously and annotating my ideas, just to get them on paper. After this I could move onto the next stage which was bringing a few decent concepts forward and photobashing them up. In all honesty I wasn't too satisfied with my concepting, I think I could have done way more with it that what I chose (such as my hoover-leaf blower idea), however I chose the most practical out of some quite wacky designs.

The rest of the week was taken up mostly by writing my Extended Reflection essay which I hadn't got round to in the Easter. It wasn't that I was dreading it, I don't actually mind writing, it was just I did need a break or I'd completely burn out on the cusp of the year.

So, essay done, I'm now moving onto modelling my assets for the 'Zombified' pick-up truck. I'll join you next week with the finished asset and the end of the year! (Wow it's gone quick...)

Until then!

Week 25; It's the Final Countdown

Well, I didn't promise anything about song lyrics!

At least this weeks weak reference actually has some meaning behind it. Writing this blog now we've just broken up for the Easter holidays, meaning we only have 2 weeks left after this before our major hand-in. This includes the zombified pick-up truck, along with our extended writing piece, and a presentation due sometime when we get back from the Easter within the first week. The year is coming close do an end and it's getting really quite scary to think I've already completed my first year of teaching!

Anyhow, lets get onto what I did this week. For the most part; texturing. Lots and lots of it. So as I've covered before I have been using substance designer to create my PBR textures. I've been able to work out how to use the mesh to create maps like world space normals which essentially tell where the verts are pointing (from what I can understand from it), so it can apply wear and tear to certain parts of the mesh which would have more tear than others, or functions like bottom to top dirt.

Albedo

Normal

Metalic

I think this has gone quite well, I've really enjoyed the process of procedural generated textures and really believe this has got to be the way forward, since you can create a material and re-use it on a different mesh to have vastly different results. The interface did take a bit of getting used to but I did get there in the end and created my final asset!

The only problem I had was putting it into UE4 for the final beauty shots... UE4 as an engine doesn't support metallic or roughness maps, the format that Substance Designer uses. So I had to essentially faff around with trying to convert them into specular and gloss maps, which are still completely different to metallic and roughness maps. I'm going to be honest and I don't quite understand how I managed to make it work in UE4 but it does to a degree (apart from the chrome on the hub caps not showing up correctly).

I'll leave you with the beauty shots and I'll see you after Easter!

Week 24; Toyotally Awesome Progress!

OK, I promise this will be the last pun. Seriously.

I confess much of this week I've been a hermit. I've only really seen the outside when walking to lectures on Monday and back again. There is also an excellent reason for this!

I downloaded Substance Designer and a set of programs from the "Quixel Suite" (including nDo - normal maps, dDo - similar to Substance Designer from what I've deduced, finally 3Do - an extremely useful tool allowing to see your asset in PBR lighting scenarios). The whole of this week was really just testing the water with these programs and getting used to them. I've made a few test materials in Substance Designer but I'm saving that more for next week.

This week I've mostly been focusing on nDo and creating the normal maps for my truck. The reason I actually chose nDo as a program to work from is that I see a benefit from the way it creates normal maps. To create normals in nDo you draw and edit vector shapes straight into photoshop without having to model a high-poly version and bake the information onto the model. This is an incredible advantage to hard surface texturing, as i don't have to make a completely new model to bake my normals! But in all honesty that's pretty much all I've been able to do project-wise with this week.

However, I did also start my own project. I've decided to aid my knowledge of PBR texturing by making a petrol station. Hopefully this will familiarize myself with the programs better, alongside doing my Toyota Hilux project. I doubt I'll be able to finish off this petrol station but I managed to do most of the modelling rather swiftly (I'm really getting the hang of this modelling thing now!). Anyway, here are the updates from the past week:

Screenshot from 3Do

normal painted in nDo

Week 23; Keep Truckin' Along

Firstly, I'd like to address the well thought out title of this blog.

Okay, now to business.

This week we go straight into another project, and I do have to admit I'm still feeling the effects of the crunch time we had in the character project. Luckily though, we have a relatively straight forward project! This will be the first instalment of our Zombie Pick-up Truck final project.

We've been given 6k tris and three weeks to model and texture the exterior (and interior as a stretch goal) of a Toyota Hilux truck.

In all honesty, I think this actually too much time spent on this part of the project as we already have been provided with the blueprints of the truck and we can follow them quite strictly then when moving onto the interior (which I fully intend, given the time we have). Therefore I've made the decision, following an announcement of a talk coming up on normal maps by one of our third years, to fully PBR texture my truck.

So, what is PBR texturing? Well, PBR stands for Physically Based Rendering, in which the light directly and dynamically effects the textures of the model. This gives a somewhat realistic effect when done properly. So, how will I make these PBR textures? In preparation of this I've been following a series on youtube where (insert name) creates an asset for a level using a Photoshop plug-in – nDO – to create the nomal maps straight onto a photoshop document, wherein normally you'd have to create in 3DS Max or zBrush, a high-poly model and “bake” the information onto a low-poly model. I will be using these normals in conjunction with the normals that I'll get from a program called “Substance Designer” that I'll be using.

My ambition of using and learning a completely new program in 2 Weeks will be quite difficult to achieve but I'm up for the challenge and hopefully it'll make my model look quite pretty!

However, at the end of this week I've created fully the interior and exterior of the model and I'm ready to continue onto the texturing and unwrapping for next week.

Interior

Exterior

Until then!  

Post mortem

  So, as I've covered in previous blog posts, we had five weeks to complete the character project. Two weeks for concepting in which after the first we had feedback to improve on for our final concept to be approved. This was an individual project and we had two 1024x1024 texture sheets and a 6k tri limit. In hindsight, I'm quite happy with the whole process, the reasons I'll go over in the following paragraphs.

  What I think worked well:

  In all honesty I'm actually quite happy with the project overall. This is the first project I actually spent a considerable amount of time (2 weeks) concepting and managed to paint up a final concept of the model. I liked the overall aesthetic of the character, and how it turned out in the end. I'm particularly happy with the textures on the model, I think they have sufficient detail and readability and fit the project brief quite well.

  What didn't work well:

  My topology. I briefly covered this in my previous blog post but I got in a muddle with the topology of my model. I went against my own intuition and decided to model the folds into the model, which I am aware messed up the topology to make it hard to rig. I'm also not too keen on the pose I decided to model him in. I tried to get an elderly man ambiance for my character so I modelled him in a stoop, however he seems to come across more stunted than stooping. I also should have made it more obvious he was older with the use of greying hairs. I also should have allocated more time for texturing as towards the end of the project as I found that I had to rush the last refinements such as the shadows on the face.

  What I would do differently?:

  As I've said before, I would completely redo the topology of the model to fix the rig of the model and in turn the deformation of the model. I'd also fix the colour scheme of the model. In my opinion it has too much green for a fisherman which should have more of a blue tinge to the design in correlation to the theme of the sea. I would have spent more tris on the accessories and more time in texturing them as well as they come across rushed and also don't deform properly with the rest of the model.

  So, knowing all this now, I'll redo the character project but limit myself to around 2k tris. I may change the subject matter to what I want, reconcept over the summer or in downtime between projects and reflect on what I've learned and how long the new project has taken me.

Week 22; Coming to a close.

The Final week of Character project and along with getting feedback from my third year peers, brushing up textures and rigging, this week has been quite a busy one, unlike what I'd had expected.

So I spent the week brushing up on my textures after receiving some feedback from my third year mentor in which I was advised, for the face of the model, overlay a picture of my actual face then paint over onto the texture sheet. I also didn't have the correct shadows under the nose and the neck and forehead area so I had to change those as well. I was also informed that I should make the back of the poncho less detailed in order to mask better the fact that the texture is mirrored. Also I cleaned up the hair and hair line of the model.

Now the rigging part didn't go so well.

Due to the way that I had modelled my character (modelling in the folds) I now had to deal with best the bad topology I had made. This in turn meant that the character did not deform correctly when put through animations and took a lot of fiddling around with the weight of bones to fix. For the most part I went through Mike Pickton's tutorial on rigging a character (https://www.youtube.com/watch?v=fpU1yqsT8_A) and following his instruction step by step. Even though I did this I still had my fair share of problems that ensued.

However at the end of this week I'm happier at the knowledge I've gained and I now have a proper appreciation of why topology is so important!

The End Result:

We were also asked within our review to do a post-mortem of our character project. In the next blog post I'll discuss my end feeling with my project, how it went and what I've learned more in depth, until then!

Week 21 & action plan

Back in Week 20 we were asked to prepare the current stage of the character project we had reached at that current point in time. After submitting my screenshots of the model I had my review with Mike and Emma.

After about ten minutes with them I was tasked to come up with an action plan on the way I'm going to tackle the next part of the year in respect to what we'd discussed in the review.

One of the comments that was made was my seemingly lack of response to feedback which I admit is something not documented in my blog, so therefore I'm making a conscious effort to document it not only in my head but on my blog as well.

I also take too long on projects as I find myself not balancing my time with my personal work and development as well, so I'll need to compensate with the quality of my projects in order to develop further as an artist in my own respect as well.

Now I had misinterpreted the hand-in for what they wanted to look at for this review, I thought they just wanted to have a look at the point we had reached in the project so I had just saved out some jpeg screenshots of where I was at in 3DS Max. However I think from the feedback they wanted something more presentable than just a screenshot, therefore I need to be mindful of how my presentation will effect my grade.

I was also told to re-do the character project in my own time at around 2k tri limit, theoretically taking half the time of this one in which I'll understand better the processes I've been going through.

So to recap:

Commit positively and document feedback.

Be mindful of completing projects in sufficient time.

Pay attention to aesthetic quality of work and presentation.

Redo project with a 2k tri limit.

Complete a post-mortem of project afterwards.

In other news I've been modelling this past week and I've reached the end of my modelling. At the time of the review I had just completed the body, however now I'm about to complete the texturing phase of the project too.

The main problem I found was that in the 3 Weeks leading up and preparing us for the project, we had a week where we modeled folds, which I'm now aware that this is not a good idea. This reason being that it completely destroys the topology of the model and therefore will not deform correctly when rigging. However, ironically, I don't have enough time to redo my topology completely as I will then be behind for the rigging and texturing of the model.

For now I'll leave you with the screenshots of the model so far and next week I wont have too much to show as I'll mostly be rigging and tackling that!

Week 20; The Modelling of Me!

  The modelling of the head didn't go as smoothly as I'd of liked it. In fact, I spent the majority of the week modelling my head, which was a huge time sink. I think I personally invested too many tris and too much time in modelling my head and overlooked how many tris I'd need to model the rest of my body.

Breaking my face into Face Loops

Start of modelling the head using box modelling - Front View

Left View

  Never the less, after the final touch ups on my concept and researching facial topology I cracked on with modelling the head for my model. It was actually rather a simple process which I just found very time consuming as I wanted to get it near enough perfect since when you look at a character the first thing you notice is the face. After many cuts and swift loops later, I ended up with this at the end of the week:

  At the end of this week, after modelling most of the head I decided to do a quick paintover of the features to make myself look considerably older. I debated whether to model the creases in my forehead in for next week, but after consultation, decided it'd be better to conserve my tri count and add them in on my texture. 

  It's not perfect yet, but I'll keep on cracking on with it over the course of next week, and hopefully tomorrow I'll start modelling the body and the rest of the accessories!  

Paintover of head model

Week 19; Feedback and Continuing Concepting

  This week we continued with our conceptual character development based on the feedback we got last week on Friday. My feedback entailed annotating my mood boards and further idea generation of my photobashes. We were also set a task to complete 20 iterations of value design (separation) for our characters so we can apply colour easily as an overlay.

Clothing Board

Annotated Mood Board

   So I started Monday with further pulling ideas from mood boards and clothing boards I had compiled, after annotating them and figuring out exactly what I wanted from them, then arranging them on my mannequin. Like last week I warped and used the perspective tool to clothe my mannequin until I had around 32 iterations of idea generation. I then put these images into greyscale mode so I could separate the values easier for the next stage.

Photobash Continued

Face Paintovers

  I then went on to some further idea refinement by doing paintovers from the four successful photobashes. I added accessories and such quickly by painting over the concept, which enriched the character and provided a purpose for the character. However before this I did research into what accessories a medieval fisherman would use and came up with a simple rod, a pronged harpoon and a basic net form, whilst adding detail to the belt area with various utilities that a fisherman may find useful. This added a layer of realism into my character which is quite valuable.

Accessory paintovers

  After this step I went onto separating the values, which actually was rather harder than it seemed. My initial plan was to just use an overlay layer with different value combinations until I found one that looked okay, but the layer blend mode didn't give the drastic change I was looking for, and ended up painting over some values by hand to get the correct definition. This in turn actually watered down the details I was wishing to keep so by the time it came to collecting different colour combinations the detail had almost gone and the concepts seemed rather watered down.

Value Separation

  However I proceeded in producing the colour iterations after picking a value separation that I liked. For this I collected a board of moods and colours from screenshots of Disney films and general happy mood images, which would be in keep with the project brief. I then colour picked the compositions and applied them over my concepts.

Colour Mood board

  They do seem rather radical but I figured that I'd make them radical to see what worked and then edit and mute the colours on a later pass of the concept stage.

Colour Design

  I then landed at my final concept, seen below:

  However, I was unhappy with the presentation level and detail of the concept so I went back in and added texture definition and surface definition, re-adding the folds I'd lost in the value separation stage. We arrive here, my final concept (middle), with hue/saturation tests either side of the final just to see if I wanted to tweak any final colours.

Final Concept - Refined

  That's pretty much where I wound up this week, I'll join you next week where I can finally start modelling!