Thursday, February 28, 2013

Oscars controversy: Rhythm & Hues

The Academy Awards ceremony took place on Sunday 24th February - whilst remembered by many as a glamorous celebration of modern Hollywood and famous actors, it's also spurred a lot of controversy, particularly from the visual effects industry.

Rhythm & Hues, the VFX company that worked on Life of Pi (and has previously worked on The Hunger Games, The Lord of the Rings and Harry Potter), won the award for Best Visual Effects. But when supervisor Bill Westenhofer accepted the award, his speech was cut short following his announcement that the company was experiencing "severe financial difficulties". 

“What I was trying to say up there is that it’s at a time when visual effects movies are dominating the box office, but that visual effects companies are struggling. And I wanted to point out that we aren’t technicians,” said Westenhofer, who shares the award with fellow winners Guillaume Rocheron, Erik-Jan De Boer and Donald R. Elliott. ”Visual effects is not just a commodity that’s being done by people pushing buttons. We’re artists, and if we don’t find a way to fix the business model, we start to lose the artistry. If anything, Life of Pi shows that we’re artists and not just technicians.”

"Several hundred people reportedly congregated outside the Dolby theatre in Los Angeles as the stars walked the red carpet, demanding better treatment for the artists who make the spectacular visuals for blockbuster movies possible. The protest was planned after the well-known Rhythm & Hues effects house filed for Chapter 11 bankruptcy last week, shortly after winning a Bafta for its work on Life of Pi."

"One protester, Nancy Evans, who left the visual effects business a few years ago, told Variety, “If these companies are getting subsidized by taxpayer money, where is the money going? If it’s not going to artist salaries and it’s not going to the effects companies, where is it going? It seems to me that’s a continuation of expressing money from working people to wealthy investors.”
According to Variety, among the industry practices that protesters declared detrimental to visual effects artists were the current fixed-bid business model and the need for artists to move from place to place around the world to find work."
It seems astonishing to me that in an age where many films are wholly dependant on visual effects, the artists responsible still go under-appreciated and underpaid, even in the highest profile productions. The work is extremely time-consuming and intense, and requires years of training to reach a professional standard. The fact that the industry is suffering financial problems, on top of the issues of students being under-trained and generally unsure of how to get started in the industry, doesn't give good impressions. I can only hope that Hollywood and also educational constitutions begin to change their approach to this crucial side of filmmaking.

Wednesday, February 27, 2013

Spaceship - modelling, texturing and animating

I struggled with coming up with an interesting spaceship concept - all I could think of were the typical flying saucers and the sleek sci-fi stuff that's popular in video games nowadays. 

My alien is a sort of stone-like structure, so I decided to look into old ruins, particularly Aztec ruins, for inspiration. I put together a moodboard of some aspects I liked: 

Here's some concept sketches, and the final concept art I used as modelling reference. I adapted the general pyramid-like shape of the ruins, but experimented with different proportions and shapes. 

There's also a rough floor plan of the interior showing the general shape I'm going for, although it's quite clear that the interior and the alien are far too large for the ship - I'm thinking it's more of a transportation vessel, and the protagonist awakes on a different planet rather than inside the ship itself. 

Due to my lack of inspiration when it came to the spaceship, and resulting delay in finalising an idea, I didn't really leave myself much time to model. I hand-painted a texture using many stone-effect-like custom brushes in Photoshop, and added a bump map, but the geometry is, as you can see, not very intricate at all and doesn't really resemble the concept art. 

I'm not at all experienced in modelling architectural structures such as arches and columns, so I had to leave out a lot of those details. If I'd had more time to spend on the ship, I'd have happily looked into tutorials, but sadly I didn't.

(the UV map messed up a little when I decided to add in a load of bevels after applying the texture... I hope to hide this using clouds/lighting)

However, in the actual film, the bottom of the ship will be all that will be immediately visible to the viewer, so I decided not to worry about it too much (again, comping is the most important part of this module). 

To animate the ship accurately, I created a new camera and imported a screenshot of the relevant film footage as an image plane. Rather than moving the ship itself to fit the perspective of the buildings, I moved and oriented the camera until the image plane was roughly in line. This meant that I could move the ship up, down, left and right without worrying about it's rotational values. 

The animation will be exported as a .TIF to retain transparency values, and then imported into After Effects ready for comping. This will either be done as a single pass or through different layer passes (diffuse, shadow, occlusion, etc.)

Although certainly it would be more beneficial to render my animated footage in different passes so I can manipulate them in After Effects, it really depends on how much time I have left.

The ship's interior environment: displacement mapping

I wasn't sure how to approach the interior of the ship, as I think largely it will go unnoticed due to the overall darkness of the scene. My original thought was for it to be piles of rubble and rock, but as I have lost quite a lot of time and am approaching the deadline, I've just ended up using default Maya textures.

I was intrigued when I came across a tutorial explaining how to create a terrain using displacement maps.

The results of this technique can look very realistic and intricate. Here's my results:

My main issue with this is that it looks very large-scale. I would only need to use a small section of this terrain to animate my scene. I tried to repeat this technique using my own hand-painted displacement map, but it came out looking very strange and 'crystallised' as opposed to organic like above. 

Unfortunately due to time restraints I've decided to abandon trying this method. I went back to the original scene I created in Maya, and added a 3d rock texture with a noise bump map. I experimented with different effects until I came to one I like. (You can see some of the results in the previous blog). I lowered the 'eccentricity' of the specularity so it didn't look so smooth, and increased the bump map to create a 'glittering' rock look. 

The geometry of the scene itself isn't really what I wanted to use in the end - it doesn't look particularly good.   However as there's only  few days now until the final crit, I really needed to start working on the animations and get started on the comping, as that's what this module is revolving around. In After Effects I am going to maybe add a vignette effect and generally play a lot with the lighting to hide the surroundings as much as I can.

Experimenting with volumetric lighting

Since the designing stage of this module I've known I want to put a lot of time into creating an eerie, mysterious atmosphere. I'd like the lighting, shadows, colour scheme, mise en scene etc. to all be relevant and express a specific 'feeling'.

I want the lighting of my scene to be dark and tinted blue, with blue-turqoise 'edge' lighting to add more definition. 

The results of some experimentation:

Although I like this outcome, it seems a little too harsh. (On a side-note, you can see here how much the bump map is brought out in these lighting situations.) 
My tutor suggested I look into volumetric shadows, so I followed a couple of tutorials.

This type of lighting makes it appear as if the atmosphere is foggy, and creates interesting varied shadows which effect the surrounding fog depending on the shape of the object. Essentially you just add a spotlight to the scene, and enable 'light fog' under the 'light effects' tab in the attributes, and also enable depth map shadows. You can then vary the intensity of the light, the fog and the fog shadows to get the result you want. For some reason I haven't looked into yet, it only works using the Maya software renderer, not mental ray.

When I replaced the sphere with my model and tweaked all the attributes a bit, it definitely began to create the softly lit, enigmatic feel I'm going for. 

I couldn't quite recreate this exact lighting within my landscape - at the moment the volumetric shadows look a little grainy and difficult to notice. I will be tweaking these further before rendering. I also added in the harsher edge light to emphasise the form and significance of the alien and prevent him from just fading away into the background.

Before I added the edge light:

And just to show how it looks with intense fog and fog shadows (in this case the spotlight is angled slightly lower behind the alien):

This is something approaching the final lighting result (before I tweak the grainy fog shadows)

Here you can notice the flash light which comes from the actor's perspective. The alien will otherwise be a dark silhouette. This is all to make him appear more foreboding, though not necessarily intimidating - I'd like it to be fairly ambiguous whether or not the alien is malevolent. The main theme here is his mystery.

Something I've noticed throughout this course is that I actually really enjoy creating the lighting in a scene, and generally adding those final touches to the ambience. It's a surprisingly time-consuming part of a project, but drastically changes the outcome. I'd perhaps like to look into this a little more as part of PPP. 

Thursday, February 21, 2013


I recently signed up to CGHub, a community of professional and amateur digital artists, both 2D and 3D. Not only is it a useful place to showcase work, what seems particularly useful is how active the forums are. Artists often have "sketchbook" threads where they upload their sketches, ideas, WIPs and finished pieces, and others are able to look and offer their opinions and constructive criticism. It's a great, laid-back way to get valuable feedback on your work and discover ways to improve. It's also easy to get into contact with other artists, offer your own feedback and advice, and develop a network.

Another useful area of the site is the 'challenges' section, where bi-weekly competitions are held in a variety of categories, including character design and creature design. The bi-weekly 'drawing jam' covers a huge range of themes, ranging from the typical such as designs for space stations and super villians, to slightly more unusual opportunities to test your skills and design something different, such as cherubs and redesigning Disney characters . I'd definitely like to become more active within this community so I can get feedback from more professional artists, and take part in the challenges and develop my skills.

A while back when we determined our 'areas of improvement', I wrote about how I lack a lot of confidence in my work, especially when it comes to sharing it online, and I often fail to actually finish my art to a polished, fully rendered standard. I also wrote how I lack an understanding of the industry and the sort of skills and knowledge I need to become more professional. I think CGHub is a valuable place to develop upon all of these weaknesses. I can share work, gain experience in talking with other artists, become more comfortable with sharing my work and understanding my weaknesses, and hopefully improve.

Wednesday, February 20, 2013

The VFX pipeline

Here is a really useful and informative page detailing the typical VFX work pipeline. The amount of work is quite overwhelming. All the departments are interdependent, and particularly in post-production each is involved in extremely intricate and detailed work. Of course, the pipeline varies on the scale of a project; on a small scale project one person would probably be responsible for many of the different jobs.

The page goes into detail about each of these steps:

It's quite worrying that I'm still in the modelling/rigging stage... This pipeline also doesn't take sound and music into account, which is an important part of creating a good piece of film.

Something I found interesting was how software that you generally associate with VFX and CG work, such as After Effects, Maya and 3DS Max, are often not powerful enough to achieve the results the artists want; therefore they create their own plugins, tools and even entire programs that are catered to their needs. This takes place in the 'Research and Development' stage, and involves the input of technicians, mathematicians, programmers, etc. As creative projects become more technically advanced, people who are more logic and technically minded rather than creative are still crucial assets in the filmmaking process.

There are a few steps I could have worked on to make my film more accurate. For example, in a big budget film a lot of effort is put into tracking and match-moving so that CG elements and filmed elements match up with each other seamlessly. I'm not sure what the focal/aperture length of the camera was at the time of shooting, so I won't 100% accurately be able to place the model in the scene, although in my case it won't make a difference as the actor and model aren't actually in a scene together (the shots with the model will be 100% CG). My main worry will be the shot where the spaceship arrives; I should ideally make a rough 3d sculpt that is to scale of the area where the footage was shot, so that the ship will match it's surroundings, but I'm not sure how to go about that as I won't have any way of knowing the exact height/shape of the buildings. I'll have to rely on judging by eye.

Alien: Texturing, rigging and blend shapes.

Finally made progress on my "colossus" alien. 

I finished modelling, UV mapping and applying the texture. This time round, when UV-mapping, I made sure all the pieces (the head, body, armour etc. are all separate pieces of geometry) fit together on the same map, which is generally much more economical file-size-wise and also makes it easier to paint the textures consistently. It also saves a lot of unnecessary work when it comes to creating and assigning the shaders.
I wanted to retain the jagged edge appearance of the model so I haven't smoothed it, only softened the edges. So the final model is actually quite low-poly, and looks, I think, much more suited for a game than a film, but I'm hoping dramatic lighting and composition will detract from that. (Low-poly also means shorter render times, which is always a plus).

I experimented with different brushes within Photoshop to achieve the "stony" look, and also applied a bump map to make the textures look a little more refined and realistic, emphasising in particular the texture of his armour. It's subtle, but noticeable.

I can now finally begin thinking about the character's movement and animation.
Here is the rig:

I realised fairly quickly I made a mistake with the geometry (incidentally the same mistake I made with my last module's model); I didn't model his arms and legs straight (for example his arms should ideally be at 90 degree angles to his body). When testing the final rigged version, it didn't seem to cause any major troubles. If it causes the legs to bend at the wrong angle, I can use the knee control to manipulate the pole vector, so I'm hoping it won't be too troublesome to animate.

I also made sure this time to lock off and delete any attributes that the control won't be needing. In a professional environment this would make it much clearer for the animator how the body is able to move, and won't cause any needless and annoying movement of controls themselves.

I also added some blend shapes. Although my model hasn't got an expressive face, the body beneath these stone structures is organic; I can't achieve a realistic level of muscle movement, but I created a blend shape where the arms and neck bulk up a bit to mimic the muscles tensing. When he moves his arms around, this should hopefully add that little more realism. (Ideally I would have spent more time investigating how to model arms so that they looked anatomically correct, this way I could distinguish what muscles are where, and use blend shapes to create more accurate muscle movement).

Already, in hindsight, I would have liked to have broken up the face "plate" into different sections so that I could animate some sort of facial movement. I've definitely not gone into as much detail as I would have liked.

Monday, February 11, 2013

Alien model progress

I found a sheet I made earlier in the module of rough silhouette ideas for my alien concept. As I've mentioned before, I struggled to break away from the typical and create something more unique - they seemed to be either Giger's Alien or a Mass Effect rip-off. I decided to abandon this approach, try to clear my head of preconceived "alien" aesthetics and attempt to think of something different.

I reached a point where I started panicking over the fact that I couldn't settle on a final idea, so I rushed into an unfinished concept and tried to start modelling, with the intention of creating the details of the alien in Mudbox. This had quite disastrous results...:

Not only did it look awful, the geometry was very poorly modelled. I didn't give myself a chance to figure out the correct topology and edge flow. I quickly left this model and started from scratch.

From the few rough concepts I'd made when I decided to take a Colossus-inspired route, I picked this one (I accidentally saved over the others when making the turnaround). Parts of the body are covered in armour-like structures, meaning I don't have to worry too much about modelling muscles. I also generally liked the look of this one in comparison to the others. 

Here's the final alien concept:

Some shots of the modelling in progress:

Each part of the body was modelled on separate layers so that I could hide the armour and ensure the torso underneath had reasonable edge flow. It's far from perfect, but I very quickly rigged the model and tested out a couple of poses to see if it generally will work okay (as I didn't do any weight painting it's obviously far from what the final model will move like). I think it will considering there's no dynamic movement, it's just a very slow walk.

I'm not sure what I'm doing in terms of smoothing yet. I'm also not sure how to connect his legs to his torso, so I think they might end up "floating" like seen above. I might add some strange lights in areas like this to hint at it being some sort of "alien magic" keeping him held together.

I tried to send the model to Mudbox, but my lack of understanding with modelling meant I made quite a few errors and inconsistencies with the geometry, so it wouldn't work. I'm going to have to spend a lot of time creating a detailed texture to make up for the fact I can't add the level of sculptural detail I wanted. 

Trollhunter and Cloverfield VFX breakdowns

I found a few videos giving an insight into the development of the models and VFX techniques used in the film Trollhunter. The film is a "mockumentary", shot in a similar style to The Blair Witch Project; although it is arguably an overused cinematic technique, I thought it created an intriguing atmosphere in the film. Seeing these massive, grotesque creatures through the lens of a camera being held by a character makes them seem much more intimidating and scary.

The below video in particular gives details about the structure of the models and the different effects applied to the video. Muscle simulation was applied to the rig to create more organic movements of the mesh. Additional 3D details like growths and plants are then applied on top. It's clear to see that a great deal of work is put into post-production to ensure the monster fits the natural environment; the footage of the man on top of the jeep is in front of a green screen, the sky, troll and other environmental elements are dropped in, and many techniques and effects are added to emulate the "handheld footage" feel: matte separation, lens distortion, camera shake, grain match and colour correction.

The video also shows a few of the steps taken to create the scene of the monster being viewed through the jeep window; the troll's wiremesh is placed in the correct area, occlusion effects are applied to approximate realistic lighting behaviour and then VFX elements, like snow, are added last.

There are many different aspects to the rendering of the troll. As listed in the video, these include the beauty pass, vegetation pass, volume light pass, blast light pass, fur ambient pass, fur key light pass, ambient pass, soft reflection pass (and that's not all of them).

I don't know much at all about this side of visual effects and model rendering; it all seems way over my head at the minute. I did consider experimenting with muscle simulation in my model, but seeing as the focus of this module is on compositing and post-production work, I don't want to create much more work for myself than I already have.

I also found a breakdown of the effects used in Cloverfield, which similarly is through the point of view of a character and their camera. Here we can see techniques such as virtual set extension, (which I was surprised to discover a while ago is used widely in TV series so that they don't have to shoot on location). 3D elements include the head of the statue of liberty, bridges, aliens, etc. which are all composited with background and sometimes foreground plates to create the final clip. It seems the aliens are animated within a 3D structure reflecting the environment of the film to dictate where parts of it might be hidden from view, or where it interacts and collides with the surrounding buildings.

It's very clear to me that much of what makes a model convincing is how well it fits into it's environment, in terms of aspects like lighting and colour correction. The model I'm making for this module might not be anywhere near Hollywood standards, but I'm hoping if I spend a lot of time tweaking lighting and grading it won't look too separate from it's surroundings or from the actor.