Friday, 29 June 2012

Material Experiments

Over the past few days I have been attempting to play with material settings in Maya. Please note the lighting has not been properly set up and I have been using bright lights, the theory being that if I can make it look interesting in bright daylight when I add more shadow and murky atmosphere it will be more interesting. 

The first task was to get a close up stone texture which would also work from afar. This has been much trickier than anticipated but I feel that I am starting to get the right kind of results. It is just a phong shader to which the diffuse, specular and normal map has been applied. I needed a texture which would be rough, contain a high level of detail; yet contained a degree of "shininess" to reflect rain when wet. I had hoped that once sorted I could save the presets and reapply to various textures across the model. Sadly this is not the case but by focusing on the more prominent textures at the front of the building I shall have something working successfully to start the rendering process on Monday. 

 Close up detail development of the entrance elevation:

Although the textures are working well on certain close up areas the bigger ones with less detail are proving much less successful and it may be a case of having to go back and re-map the larger areas such as the train tracks.Unsuccessful texture of a larger train track area featuring stretching at the edges. Think more detail is required and perhaps a larger map should be used:

The second material to work on is the metal. This was trickier than the stone as the starting point and textures were not as finalised or as interesting as the stone. Having experimented with both phong and the mia_x materials I'm still finding it hard to find a realistic, weathered texture that applies well and looks good. At present I'm experimenting on the windows as this is where the texture is most noticeable. I have tried some of the car paint and metal presets in the mia_x material however find that the result against the texture maps is rather jarring on the eye. I will be re-working the metal textures in the hope of improving the look. Render test on side entrance elevation: 

The final big material to worry about is the glass. For this I have used two different types. A murky frosted glass for the stained glass texture and the doors, and a more reflective solid glass for the main towers. Initially I was worried that I wouldn't get a good result with the glass but in some respects I am happier with that than the other materials. I feel that I am getting a good level of reflection on the main towers and by applying a rough bump map have got a good level of graininess which works well. 

The only issue with the glass which I am finding is that when Global illumination is turned on I am getting some extreme lighting conditions which could be just to do with the lack of a proper lighting set up. Its also tricky for me to know what or how things will change when they are divided up into render layers: 


Wednesday, 27 June 2012

Super Quick Thank You To Team Ridiculous..

... for saving me from Maya and treating me to cocktails and my weight in sushi.


Render Layers

As soon as I am finished tweaking the materials (in the next two days with luck) I will be going onto  the rendering and hope to complete that by the end of next week while the rotoscoping onslaught begins. Unsure about what render passes I would need I got advice from Dylan who suggested that I would need:

direct lighting (diffuse) 
indirect lighting (global illumination)
shadows (though may have to use maya hardware over mental ray render as it can act erratically)
ambient occlusion
alpha/ matte (for compositing)
3D motion vector blur (not required in all shots but may prove useful)

Taking Dylan's advice I'm going to start on the external shot looking in as I fear that some extra detail in the textures may be required and also some of the full external shots which use still frames. From experience with the going live module I certainly anticipate that this will be a far tricker process than I am expecting but with luck I'll at least be able to rotoscope and set up a composition in Nuke while the rendering occurs. 

Final Internal/Close Up Maps

Below is a video containing all the maps I created for use in the close up of the external looking into building. As with the external,  although I've created occlusion and displacement maps I've found that I have been getting a better result with most textures when using only the normals and the specular maps.  Although trickier to complete to a higher standard the textures in this shot are simpler in many ways and I may have to look into adding extra details on planes to make the appearance more varied and interesting.  

Final External Maps

Taking Pat's advice I changed the png to targa files and have had much better success regarding the quality and application of the textures in Maya. Below is a video containing all the maps I created for use in the external building. It was a lot of work and required a lot of tweaking. Although I have created occlusion and displacement maps I have found that I have been getting a better result with most textures when using only the normals and the specular maps however it is handy to have them just in case. 

Friday, 22 June 2012

External Textures Update

Finally the diffuse maps have been applied and the correct materials assigned to each map. This took far longer than I thought as I experienced technical issues and many of the textures has to be re-worked further. Originally I was using uncompressed tifs but was having severe problems loading them in Maya, after discussion with class members turns out that Maya 2011 has glitches with tifs so have used png files instead. This was taken further when I was having difficulty files if they had a resolution over 72 dpi, thankfully with a bit of tinkering and offsetting values and using bigger map sizes instead I have managed to get them all there. The quick render below was taken without the correct lighting and don't have any of the diffuse, specular, bump, normals or occlusion passes imported yet but gives a rough idea of colours and detail to be worked on at least.  Next week tinkering will occur when the internal ones are done too. Currently there is only three materials applied: a phong for the stone and basic textures, an mia_material_x_passes (though am also experimenting with dilectic material) and a mia_metallic paint for the bigger areas of steel which looked awful when a texture was applied to them. It will look much prettier soon but here is work in progress....

Stage 01 : Internal Complete

I used the same technique of re-naming and re-uvmapping for the shot looking into the building as I did on the external. The process was far quicker this time and has allowed me to simplify what was a rather complicated model. Unlike the external this one required a lot of re-modeling but thankfully practice means I'm getting much quicker at this.  The new maps have been exported and will be completed this weekend ready for material experimentation and rendering next week. I feel like I am finally making progress but acutely aware of what is left to do.

Tuesday, 19 June 2012

Program of Study 4.0

In order to pacify some of my tutors I decided to re-write my programe of study so that it reflects more clearly what it is Im hoping to achieve in the remaining 5 weeks. It's getting very scary how close we are getting to the end but have to say I'm looking very forward to the finish. Been a tough few weeks!

Stage 04: Render Tweaking

Finally I am at the point where I can begin to experiment with textures. had a good meeting with Dylan this morning about how to set up effective shaders and materials so that I can begin to start tweaking the textures I already have. Going to try a modified phong shading group for the main stone and metal and a glass mia_x_material for the glass. 

Below is the first very quick test render of the first application without any tweaking in the wrong lighting but already there is a marked improvement with the texture, need to realign the UVs but glad the detail is starting to come out. Now the fun can begin! 

Monday, 18 June 2012

Mission New UV: Stage 02 and 03 Complete

For the external model in any case. Internal will be finished tomorrow once I get some feedback from Dylan regarding the way I have done it so far. So the external model has been re-UV'd and 4k textures have been created for the new maps. Although this was a frustrating process to go through again it should allow me to have the textures on the model at a higher resolution giving greater emphasis of detail. I am anticipating that more detail and variation will have to be included later however taking Dylan's advice I kept the textures relatively simple this time as it is more effective when applying to a larger (or in this case ridiculously huge) model. Updates on how the textures will actually work regarding the renders will be uploaded tomorrow following the meeting. I've included some of the maps below as an example of how I have split the model up further and the level of detail now included.

Sunday, 17 June 2012

New Animatic

More cuts made. Hopefully more achievable.

Jeanette Tutorial 07

On Thursday I had my tutorial with Jeanette. I think both Jeanette and John are concerned over my progress however hopefully by the time I have the new maps done and have a meeting with Dylan on the 19th where I stand will be much clearer to both myself and the tutors. Overall though despite knowing that i will not get the project completed I feel that the peer review helped me clarify how I would like to present my work and what I am aiming to achieve by the start of August. To aid this and to help explain to my tutors I plan to update my Program of Study while my aims are still clear in my head to help me explain it better and give me more of a focus to work towards. 

Dylan Stage 01 Complete

I have taken the time to rename and re-UV the Maya models. This will hopefully ensure that by exporting as a series of 4K maps that the overall graininess of the close up textures will be lost and I will be closer to achieving the textures I am aiming for. The reason that this will be more successful is that I have analysed the bigger areas giving them a higher proportion of the available texture map space and have divided each building section into more maps, allowing for more space and better division. In addition I have decided to arrange the maps into co-ordinating materials for when the maps are applied to the mia_x materials. eg. all the glass for the main doors are in a different map from the stone facade. Although this requires a lot more work it should make the later texturing stages more efficient and result in a more finished, professional textured model. 

The Help

After Wednesday’s peer review Felix Zhu agreed to assist me with the rotoscoping element of my final live-action footage. I decided to give him the shot from under the bridge. This is a tricky shot as it contains cars and people who pass in front of where the building will sit so I am very grateful for his co-operation and hope that there is a way I can help him in return with his project.

Shot 008 and 002 Development

After further work on Shot 08 and 02 I have decided to drop them from the film and display them as still images instead. There were several reasons for this but the main one being that not having a sufficient amount of moving footage and having to rely on static images would jar with the flow of the film. I do think that they are both interesting shots and will continue to develop them after the deadline but in the meantime will strengthen the narrative of the project by continuing to develop the narrative away from the screen.

Peer Review Feedback

On Wednesday I had the peer review for semester 3. Overall I think it was rather successful and helped me clarify both my progress and what it is I am hoping to achieve in the time left regarding the final masters submission. In short I am trying to create a realistic sense of what would of happened had the futurists built Waverley in 1914 working in an interesting narrative between the juxtaposition of the futurist/ neo-classical architectural styles and the aging process of how their vision would look in contemporary Edinburgh today. From the group discussion that followed the class agreed that I need to work on the timing of my animatic and that cutting the number of shots down would be beneficial if I could keep a slower pace with the shots themselves. Despite being a predominantly 2D shot it was thought by the group that shot 03 would work as it stands and that it would fit well amongst the 3D shots. This was a weight off my mind and has been moved lower down the priority list. For the shots that I will not be able to achieve or complete I discussed my idea of displaying a series of images alongside the footage which would add to the narrative of my project and this was deemed a successful way of showing how I would progress the project on its completion. I was happy with the way the review went and from it was able to enlist the help of Felix who is happy to assist with the rotoscoping of the final footage while he awaits renders. 

Friday, 8 June 2012

Good and Bad

So had a good chat with Dylan this morning who helped me to clarify some issues I was having regarding my workflow and progress with my Masters project. He has helped me to break it down into some stages to work on over the remaining 7/8 weeks which should prevent the onslaught of panic I am feeling. I do have to revise all the models and am looking at having to re-do the textures and UV maps (once again). It is deeply frustrating as I am back to where I was in February but overall its probably for the best. 

By the next meeting on Tuesday 19th:

- Rename model and elements
- Redo, Break down, finish and finalise all models
- Set up a final Maya project with the finalised work

- Redo and break down the maps further into more appropriate elements/ useful maps
- Revise layout and overlay where applicable, use extra layers to modify details
- Keep the current 2K maps for the distant shots
- Use a series of 4K, 2K and 1K maps and identify which are to be used in which shots
- Create a series of default maps against which things can be tested
- Create bump maps based on the revised UV maps

- Create and apply new textures for the shots
- Update Glass and Metal elements to mia_materials

Basic timeline:

3D finished (modeling, texturing, rendering, match-moving, tracking lighting) : 24th June.
2D finished (rotoscoping, compositing, rendering) : 5th August.


Thursday, 7 June 2012

Updated Leica Reel

This is just to help me keep on track with what is happening and how it is all fitting together.

Shot_08 Development

"That decoration as an element superimposed on architecture is absurd, and that the decorative value of Futurist architecture depends solely on the use and original arrangement of raw or bare or violently colored materials"
Antonio Sant'Elia 
Manifesto of Futurist Architecture 1914

Its a bit spot the difference between some of these videos but here is a record of the development of Shot_08 as it stands. The final one is probably my favourite but I still feel it needs tweaking. I prefer the thinner spotlights and much happier with the colour too, both of the more orange tone to the Waverley lights and the darkening of the fore-ground. I will most definitely come back to edit the haze around the station and will try to edit my layers more efficiently to create a better sense of depth in the shot. 

Friday, 1 June 2012

Evolution of Shot 08 Background

One of the issues I face with some of the shots is how to turn an obvious day shot into night. Over the coming few weeks through illness and general life-madness I have come to the conclusion that photo-realistic is not perhaps achievable so have decided to make it more surreal or more magical realist. I have also been limited in what I can do away from the studio so have been working on 2D roto and what Photoshop work I can do. Ultimately it's more me but much sillier. Below is the background for shot 8. It is a work in progress so comments are welcome. I have shown the evolution of it so far in case you know of a better approach...