Some feedback points & actions taken:

 

F1 - During the idea process, I was thinking about what kind of base to make. It was suggested that I make it look like a pop - up book.

 

F2 - The inital idea for the building needs more character - try adding slight distortion using the FFD 3x3x3 modifier to make it look like it's popping out. Also, the sides are too blank, fill it in with some architecture.

 

F3 - Colours need some development (this was half way through when I had imported the textures into UE4), make it more vibrant and inviting as a sweet shop should be - Use pastel colours to play on the sweet shop theme.

 

F4 - The large flags at the top can be the pointed castle flags instead of the square banner type - helps with the silhouette.

 

F5 - "The steps leading to the front door could be chocolate bars". It was at this point that I realised I need more things to suggest it is a sweet shop.

 

-------------------------------------------------------------------

 

A1 - I didn't go with the full on pop-up book where all the folds & cuts are made perfectly to make it work functionally, instead I went with the idea that it is literally popping out of a magic book. I felt this went quite well as I could give the book a lot of motion by using wavy lines to suggest it has just flung open.

 

A2 - I took both these points onboard and they worked out great. Without the modifier, the building looked very rigid, once applied it felt as though it was a bit livelier. I knew the sides were a weak point because I hadn't concepted anything for them, adding the balcony was a good choice as it allowed me to add more points of interest with the lighting.

 

A3 - My initial choice of colours really weren't that great and as I decided to do the colour changes in UE4, it meant that my texture sheets look very different to the actual outcome. I'm glad I did do it this way because I struggle with colours so it allowed me to play quite a bit, but it also has taught me that it may just be best to draw over a 3d rendering of the model to test the colours beforehand (should have stuck to what I know in terms of the workflow).

 

A4 - I had a bit of trouble deciding what these flags should look like, so when I changed them to the castle flags it looks a lot better, and increasing their length was a result of me thinking about the flow of the model.

 

A5 - Instead of making the steps chocolate bars (I liked the purple colour gradient currently applied), I had some ideas for other things I could add e.g. the heart lollipops, jelly beans, smarties, floating sweet wrappers, closed sign.

 

Conclusion:

 

It was my first attempt at a stylised environment project and overall it wasn't bad, however, knowing what I know now, there are many things I would do differently to achieve a higher quality result - some of these are exaplined below in the 'improvements' section. I enjoyed the experience and now I know what I will be able to do in the future when it comes to stylised. The fact that our FMP is coming up soon and we need a solid idea also effected the amount of work produced for this project as there was a significant amount of time dedicated to group meetings & general discussion about the FMP leading to a weaker overall result in this project. Nevertheless, I'm glad I finally have done something stylised and my mind is at rest knowing what it is I could potentially produce when it comes to stylised (I was skeptical about the quality of my hand painted textures/ability to give objects character).

 

Improvements for the future:

- Overestimating what is required to be on unique texture sheets: This is something I rarely do, but as I haven't done a project in this kind of style before, I went with what I would do instinctly. What I should have done was a quick test to see the quality of the textures before continuing in my pipeline. It wasn't until the end where I realised the tileables more the most part can be just 256x256 in this style. This would have given me much more texture space to use for unique assets. In addition, because I didn't test properly what should be tileable or unique, I regretfully ended up putting UV's in the unique texture sheets which could easily have been a tileable, or even overlapped & mirrored - saving a lot of space. To sum this point up, a lack of planning when it came to what I wanted meant I didn't know what I wanted so it all got a bit muddled, resulting in wasted texture space and inconsistent texel density.

 

- No matter what, everything needs beleivable structure: As this was a more cartoony style, I assumed I could make things certain shapes just because it is aesthetically pleasing. However, this was not true when it came to the structure of the shop. The roof at the top was made that shape because it worked well in the silhouette, so I therefore kept it. This left me with the weakest area of my shop by far because you can tell it hasn't been as thought out as much as the other areas and doesn't structurally make sense to the viewer. I will never make this mistake again.

 

- Modelling speed of low poly assets: I gave myself a list of things to model and thought it would be a good amount for the time frame, but part-way through modelling, I realised I could make pretty much everything I wanted to make in half the time and I still had plenty of time left to make a lot more things. Unfortunately, this combined with the 1st improvmenet point meant that I couldn't really model an interior because there had been wasted time and wasted texture space - The decision to only do the exterior was made about 1 week into the project causing an unplanned for situation (Went a bit muddled).

 

- Plan what colours would be good together beforehand: I stated my method of establishing colours were to edit them in real-time in UE4. This was good for this case because it allowed me to constantly but rapidly change the colour scheme. However, now that I am getting to be more comfortable with how I should use colours, in the future I will use a combination of material instances & actually thinking more in-depth about the colours prior to colouring my textures. For example, by taking the untextured model into photoshop and colouring over the top to get a better idea. This is a general practice done by pretty much everyone, but I regretfully left it out during this project.