Tuesday, 31 January 2017

Four: Building Biolab Elements

  Work continues on the biolab, and the template system I suggested in my last post is working well to an extent. It takes some reshaping in some areas but so far everything works. I plan ot leave UV mapping the rooms themselves to be the last stage, when all the other construction is complete.

  To control light exposure and mitigate objects phasing though solid surfaces, I created grooves that doors can slide into but should the camera catch sight of the inside these areas. This design will be especially important, come to think of it, for the isolation unit doors, which have large glass panels that will likely end up phasing though the current models.

The door designed for the offices works just as well for the decontamination unit used to enter the laboratory. This room can easily be furnished with a bench and lockers (the lockers could be kept opaque to avoid designing hanging coats or hazmat suits) to fill the space.

Monday, 30 January 2017

Four: Biolab's Doors and Windows

  Back to work, and while I have a few models to go though I decided it best to work on the biolab, as it is a key area for the animation to take place in. With some initial assets constructed, I've spent the past day finalising some of the models.

  This door I suppose is something of a hero prop, it's going to be the camera's window into Four's cell so it felt like the first thing to make. Due to its complexity I have a plan of a lower-resolution stand-in with a normal map for more distant shots. 

  I was able to quickly rig up a system that controlled the door (although maybe I could add some stagger) and allowed for easy switching between the high and low-resolution models. This system could be less complex with the other two more static doors, as it is only Four's door that will open.,

  The door and one of the windows are also UV mapped. One of the possible advantages to the highly artificial environment is that assets can be easily cloned and used elsewhere in a modular fashion, The character of the environment could then be delivered in placement and any fatigue so it all doesn't look completely identical.

Wednesday, 25 January 2017

Four: Progress on the Corridors

  Some work continues on the biolab. While the door and most of the corridor could be signed off on, they need their UVs fixed. The more distant sections of the corridor give me a good idea of which models to create normals for the more distant sections of the scene.

After changing the bracing I believe the door looks a lot more plausible now. A console to open or lock the door is planned ot go inside the indentation in the centre, which can help give an indication of the size of the room.

   To keep up the pace, the biolab was next on the list. The door to the isolation unit feels complex due to the unconventional shape, and I wonder if the indentation can be done with a normal map for more distant instances. I need to enlarge the room a little and decorate the walls.

  I will need to do research into locaations for potential lighting fixtures. Progress has been a little slowed as the deadline for my dissertation draws near.

Tuesday, 24 January 2017

Four: Ceti Base Rework Begins

  I wanted to make some progress on environments so I made a start on assets for the corridor. The primary elements being segment doors and the tunnels.

  As I get further on the plan is to make normal maps out of sections of the hi-res models that will replace the details in more distant segments of corridor. The pipes and the paneling in the middle are specific locations where this technique will likely be used.

  Researching for the segment door I didn't like the first outcome. Looking back on my influences I took notice that while there were indentations, they appeared to represent points where the door's structure was weaker. More solid areas I believe are more reinforce points to reinfoce the door's strength. These spans are noticeable on places like the doors to the bunker in Star Wars: Return of the Jedi.

Because of the shape of the hole the door fits in, I might be able to shore up some of the geometry neat the corners. I worry that in these areas there could be a good deal of stretching.

Monday, 23 January 2017

Four: Inserting Younger Teeth

  I took a break this evening from dissertation writing to attend to the juvenile. It was a rather simple task: Install the teeth into his mouth. In order to not disturb the UVs I had worked on but still have room to work, Four's body was copied, lifted up and cut in half and was to act as a template for the full mouth.

The arrangement of the teeth and gums was done by adjusting the gums using a lattice deformer and then inserting the teeth back in one-by-one in order to minimise shape deformation. The roof of the mouth may be a little high, I had to buff it as the roof was very flat in the main model.

I attempted to neotinise the teeth in an attempt to make them fatter and stubbier to hint they are from a younger mouth, though perhaps some of them could do with a little more de-aging. And while I haven't got around to them yet, the claws will receive similar treatment in terms of reshaping.

Friday, 20 January 2017

Four: Juvenile Ready for Mudbox

My dissertation deadline is coming in so some priority has been switched to that. Earlier in the day however I dedicated some time to modifying the juvenile model. The aim was to make it ready for a session in mudbox to add the finer wrinkles and create an ambient occlusion map when mapping the skin. 

I left space so that once I transfer over the tongue, gums and teeth I can insert them into the map, hopefully in a way that I can save time by transferring some of the details of the adult's UV maps. The tongue and teeth, hopefully I can transfer over without a session in Mudbox remaking the layers.

Wednesday, 18 January 2017

Four: Juvenile Geometry Nearly Complete

  I neglected to upload yesterday but here is what I was working on Tuesday. I was working to refine the model, finishing some of the chubbiness of the juvenile's body and fixing the UV maps.

  Getting the 'stubbiness' of the juvenile's front fingers was a little more difficult as they are a little longer due to the extra bone in each finger compared to the toes.

  Yesterday was also used as a time to fix the mouth. This turned out to be a little more difficult that I first assumed as her more juvenile form mouth opens with a vertical mouthline. With the current head this was difficult but not as difficult as it could be as it was meant to be in a transition phase anyway. As Four gets older her mouth gets more normal-shaped.

  A modified mouth cavity however will mean the gums will need to be modified to fit the mouth. To fit the idea that these teeth are for a more childlike creature I may have to stub them or sink them into the gumline. I was partially inspired by canid mouths so the easiest solution will be a matter of researching how the tooth sets of puppies and kittens look. Or I go back to otters as they manage a certain cuteness with their teeth despite having enlarged canines like dogs do.

  As well as adjusting the gumline, some work still needs to be done on the UV maps. I've made a start with the larger areas and most are ready for receiving their mirrored versions already.

Monday, 16 January 2017

Four: Starting the Juvenile Form

  I made a start on the juvenile form today, as a start to the next tasks in preparing my animation.

  To save a lot of time and to keep it looking like it could mature into its adult state, Four's juvenile form used the adult as a template, preserving (mostly) UVs, foundation geometry and the shaders used to make her believable. SOme UVs will need fixing and remapping, as I had to break a couple of them to save on gemonetry.

  The posiive is that I feel like she's coming out better in this form now that I know better what I'm doing. There is still some work to do before moving on to installing the skeleton. I need to make sure her eyelids conform to her eyeballs, correct the base of her tail, neotinise her front fingers to they look more childlike and maybe give her a bit of pudge in her belly 

  The body of her juvenile form is very loose. Some gemoetry was adjusted to look like fat that would disappear as she got older and more mature The eyes and claws I can rather straightforwardly import from the completed model and resize where appropriate.

Sunday, 15 January 2017

Four: Updated Turnaround

With her shaders improved I decided to roll out a new turnaround. As well as showcasing her new skin this turnaround also attempts to better portray her personality though posing.

Friday, 13 January 2017

Four: Retouched Skin

  After being briefed for the second term, I figured I would try and add some finishing touches to Four to close off the minor project. The shader above uses a mila material to provide sub-surface scattering, but it appears to have an added benefit that the light interacts with her body much better than in the last design.

I also retouched her skin to give it more life, as before she was rather pale and dead when a direct transfer of the textures from a blinn to a mila shader was involved. It also gave me an opportunity to retouch her complexion to bring it more in line with her final concept.

Thursday, 12 January 2017

Wednesday, 11 January 2017

Four: A Near-Final Model

  Today has been spent applying the finishing touches to Four's body. Using a texture sphere I transferred the texture I was working on to the main model and apart from what looks like an ambient occlusion anomaly under her belly, she's come out looking really nice. 

  I might modify the tinge of her tongue to add a bit of blue or orange. A way of giving it that alien vibe. Another factor is the tongue needs to be wet, It needs to glisten which could be done with a specular map as it shares its shader with the body..

  There does appear to be a shadow under her eye when the cornea is included in the render. It could be it's reflecting the world, which is pitch-black as there is a 40% reflectivity value. If so, it might be gone when Four is put into an environment with more than directional lights.

Tuesday, 10 January 2017

Four: Body Progress and Rendering Wizardry

  I have come to conclude that Four's adult form is almost complete. I might have sacrificed the time on the environment to make her work but I feel the sacrifice has some merit to it.

  I continued working on the textures and decided that because Four is fairly small (ish, perhaps the size of a large dog), her skin folds could be easily seen. I have had some setback as on the university PC I was using, the while PC would drop to five frames a second while extracting the occlusion map. What complicated doing it at home was the file couldn't be read, perhaps due to the way I had moved it while I packed up to leave. So I will try and get that first thing in the morning, as the ambient occlusion is crucial to giving something for the normal map to work on.

  The normal map itself came out a lot better after going over the Mudbox model using a stylus and a couple of sculpting layers to separate the detail. In particular what was improved were the folds in the neck and the hips, and the texture of the fins.

  Two of my favourite creatures from Potter spin-off film Fantastic Beasts and Where to Find Them were Frank the Thunderbird and the Occamy (the other features were each cool in their own ways). Watching the film I was particularly interested in their eyes. Scenes with Frank showed the way his eyes would twitch about while his pupils would adjust themselves to focus his vision and adapt to minute changes in the light, while the Occamy despite having enormous irises retained their curvature. Both creatures also have irises that occupy the near-entirety of the eyeball, something I tried with Colo Colo last year but because I didn't quite know how to pull off refraction the eye looked sliced off on the front.

Frank (left) and the Occamy (right) with the most prominent things in focus - their eyes. ((c) Warner Bros. Pictures).
  This got me inspired to add similar features to Four's eye. I discussed with Alan about a variable twitch to the pupil dilation that was ultimately discarded in favour of manual control as Four's eyes are small enough that extreme close-ups would be the only way to see micro-dilations. 

  The addition of curvature was quite straightforward and gives the eye more of a feel that it is a full orb, rather than being sliced off on the very front. Part of the eyeball cuts into the eyelid, which under normal circumstances would accentuate that the iris is flat. But using refraction the cornea appears to completely hide this little detail.

  Most importantly, the eye looks like a sphere again even if viewed in a way that would expose the flatness of the iris for all to see. In the days before I had been mentally questioning how to reflect the inside but refraction is so much less complex. They still need a texture to actually look like eyes and maybe a bit more gloss but that's a very quick job.