Masters Project - Week 06 / 1st - 7th August.

Week Hour Breakdown:

Wednesday - August 1st - Five Hours & Forty Minutes - Light pylon asset modelled, smoothed and partially unwrapped.

Thursday - August 2nd - Six Hours & Forty Minutes - Unwrap finished. Optimised and unwrapped floor panel asset.

Friday - August 3rd - Seven Hours & Seven Minutes - Unwrapped and textured floor panel asset.

Saturday - August 4th - Five Hours & Forty One Minutes -Created ceiling panel asset metal beams, unwrapped. Created floor edges, unwrapped and packed. Tested floater baked detail on stacked UV's. Watched stylised painting tutorials.

Sunday - August 5th - NA - NA

Monday - August 6th - Six Hours & Forty Nine Minutes - High poly planks for ceiling asset. Checked baking. Modelled wall panel asset, made edits, added detail. Resized objects to apply more negative space.

Tuesday - August 7th - Seven Hours & Seventeen Minutes - Edited mesh wall form errors. Modelled Floor container asset. Created lock floaters, unwrapped asset / baked. Altered wall mesh with chamfers / smoothing groups, unwrapped 3/4.

Total time spent this week - 39 Hours and 14 Minutes.

Overview:

This week of work consists of modelling, baking and unwrapping any assets which are integral the scene no matter what state it gets to by the end. This includes the walls, floor, ceiling, the light pylons that make up part of the floor etc. These assets need to be set up in Substance painter ready to be textured. So worst comes to worst, I'll have the assets in the Marmoset Scene with the baked detail and AO applied, which will allow me to apply just a simple base colour and material properties if time becomes a bit thin before the submission date.

Wednesday / August 1st:

I decided to start with something small to get me back into the flow of the approach I wanted to take creating these items. So I knew that with the floor panel of this environment I was going to need to create the light pylon and storage pylon separately from the floor. I started off with the Light Pylon asset as it was quite simple in form. I spent most of the day adding the details I wanted such as the small control panels and maintenance panels. Then the sliders to block the emissive light portions took a little time to model correctly. These didn't cause much issue though, and allowed me to smooth and partially unwrap the object before the end of the day.

Thursday / August 2nd:

During the morning, I finished off the unwrap of the Light Pylon asset, and made sure to bake the AO and other maps before creating a new scene in Marmoset to bring the pieces I create into. I decided the floor panel was the next step so that I could actually place the Light Pylons in the right spots in scene. I optimised the current floor panel I had available by removing most of the faces that were hidden. The way I had created the floor before hand was by duplicating the hexagon with the faces on the side still all present to I could play around with changing the height of them to create staircases and such. At this point, there were a lot of faces not being used that were bringing the polycount up high. After deleting all unseen faces, I unwrapped the floor panel with stacking UV's in mind. I decided to allow four different types of texture on the faces of the hexagons which I would stack upon. This meant that repetition could be noticed if viewed, but saved a tremendous amount of UV space for the floor. I also thought forward to when I would have to texture them all, and even though there were methods to quickly generate some of things I needed, I wanted to play it safe as I expected me to try and hand-paint some details on.

Friday / August 3rd:

The whole day was spent in Substance Painter studying references to see what kind of approaches I should apply when texturing the floor panel. ZugZug studios that was mentioned by a tutor in the Viva was already an Art station account that I followed. Most of their work had a very noticeable consistency with the way they textured their stylised assets and I tried to replicate some of those traits. I spent a long time trying to get the kind of edge wear that was applied to some of their gun assets, generators that I used didn't give the clean effect that I was attempting to get, and even then, something with my model meant the curvature settings on some of the generators were not picking up all the edges that I wanted. There were also hand painted periodic scratches that were applied to the edges, which I ended up hand painting in along the edges. The biggest thing that I overlooked up until I finished the edge wear detail was the use of gradients. I painted in gradients on the top faces with a hexagon brush with the opacity altered, and used a generator for the gradients on the sides of the hexagon pillars that were visible. To finish off I added a few circles and lines that had a minor indent height setting on them to fake some screws and separating panels. I decided not to work on this any more after it was late, and only decided to come back to alter colours or metallic and roughness values if needed as I had a lot of other things to accomplish.

Saturday / August 4th:

I was quite hesitant to work on the wall of the scene as I felt it had been neglected quite a bit and would need some extra attention. So I opted to create the ceiling panel that would allow the light to come through in a spotlight manner, and the ceiling also helped to round off the scene a bit better rather than just have the walls fade up into nothing. I felt as though the ceiling asset wouldn't be as complicated as some other pieces, so I broke down the asset into creating a single metallic panel which I could duplicate. After that I could just focus on creating the rims that would connect the beams and the asset to the wall of the scene. I thought again about stacking UV's and saving some space, but this time so that I would be able to run a grate like texture along the top and bottom of the metallic boards using floaters to bake the detail down. I knew that if I packed all the board separately together, the resulting quality of the bake would probably be too low to bother with. So in anticipation, I decided to create three boards to be used at the three UV's. Then, I would stack any other boards duplicated from these three, so that I only had three board UV's that were separate. I took the asset into substance with just these three boards and the border, to see if I could bake the floater onto these boards, and then duplicate the boards in 3DS max without changing the UV's, to finally replace update the asset in substance while it hopefully retains the bake details and spreads it across the duplicated boards without any AO issues. This ended up working as expected when testing, and allowed me to comfortably duplicate the other boards and slightly reform them at a later date. I spent the evening researching some stylised painting tutorials to pick up some extra tips and look out for information consistency.

Monday / August 6th:

The next morning I decided to finalise the ceiling panel. I did the time consuming task of duplicating the boards, and placing them in the places I wanted to spread out the various boards so they could be textured slightly differently. I attempted to snip the board by not dragging verts around and causing the UV unwrap to warp, but instead add a new line between edges, and just cut pieces off that way, so that I wouldn't stretch the UV. This ended up working, for most of the boards, but along the way some boards would stick past certain areas where they would be visible and needed to be stretched and edited. This caused more texture stretching than I expected, and decided to re-approach the ceiling panel on another day as to not get stressed and caught up on it. I moved onto the wall panel, which caused me some bother. This asset had been left pretty plain and untouched beneath all the assets I had been building on top of it. The wall was very plain and flat apart from all the indented portions which housed assets such as the Crystal tubes. I had a period of mental block, unsure of how to approach this asset. I ended up looking at reference of space station concept art to look at how sci-fi walls that tend to be one material / colour would be formed and made interesting. I came across some very informative concept art by the art station user Adrien Girod, who currently works at Ubisoft. The concept art he had designed for an Exoplanet Station, had some fundamental aspects which I realised were what I missing. He used separated panels in the textures, and angled edges along the walls. The angled edges were what helped with my design. this allowed me to separate some of the details which were lined up with a slight angled portion to change the depth of the next layer of detail. This helped to break up the wall enough to help the scene. The other major thing I took note of when doing these changes is that I had blocked out all view of the wall by making all the assets too large or placed too many, making the wall overwhelmingly cluttered. Admittedly, that was another piece of feedback during the Viva presentation, that the scene was very organised and busy, and that a little bit of negative space and discord in the symmetry of things could help guide the eye around the scene more efficiently. So, I decided to shrink down most of the panels and pieces along the wall to let the wall panel asset have more of the space in the scene. This gave me more negative space to work with so I can have areas of rest in the scene where I can add a simplistic decal. This greatly improved how dynamic the scene was. At the end of the day, I then went ahead and added a few vent insets into the wall, along with a few small panels to help break up the wall in a non uniform manner.

Tuesday / August 7th:

I decided that I was happy with the changes in form I had made to the wall asset and also needed to move forward with other aspects of the scene. So I started working my way along the wall clearing up any NGons or issues with the mesh. I found that the protruding portions of the wall that held the Crystal tube assets up had a lot of issues with NGons in the corners where verts didn't match up with one another. This took longer than I wanted to clean up, but after multiple attempts at doing so managed to find a way to correctly link up all the necessary faces. I moved onto the other asset that needed to be accomplished to take a break from spending time on the wall panel. There was a large gap in the floor still where a floor container needed to be placed. This asset didn't need many edits from what the blockout of the mesh already was. The only thing needing changed was a way to display how the top of the container was locked. Since it needed to comfortably slide into the floor, I couldn't have any chunky hinges on the sides, and this meant I had trouble finding a way to portray this. Eventually I decided to edit the shape of the top panel slightly to allow space for a floater which would act as a lock. These flat circles would spin to unlock, and allowed me to find fix without any complicated modelling. I then proceeded to unwrap and bake this asset, to then be placed in the Marmoset scene to make sure the bake results had transferred correctly. With the remainder of the evening I tackled the wall panel again, to go through the motions of chamfering the edges, unwrapping and smoothing the asset. By the end of the evening I had accomplished most of the unwrapping for the asset.

Week Conclusion:

The week in conclusion was very mixed for me personally. The progress I was making was fairly good, I felt myself accomplishing the process at a quicker pace with each asset. But considering the plan I had made the week before for the Vivas presentation, I was going much slower than I needed to be to even get close to finishing the scene to a slightly cut version. This disheartens me obviously, as I'm trying to work my way through what I need to do. I feel uncomfortable with the prospect of cutting it very close during the deadline dates, but I understand that all i can do is keep working and see how far I progress. Hopefully my next week of work will be even swifter, and I can add most of the assets that will fill out the environment to a point that I'm pleased with it. At least the bare bones of the scene are in place, and it's just adding most of details that will create a narrative.

-RyanHarleeJones