Final Year · Major Project · Semester Two

Creating the Parallax: Research, Rationality and Implementation

For our introduction to the animated documentary, the team wanted to create an opening that would offer the audience a unique element of visual approach to that of the rest of the piece. In our animation, there were five fully rendered environments to plan and design, in addition to the partially constructed elements such as the doorway that is featured in figure 1.1.

Figure 1.1 – Shown above, Jess’s paintover of the doorway scene on the left, and her render on the right. The doorway was subject to experimentation in the earlier stage of the project and was one of the first pieces of three-dimensional development.

It was whilst in a team meeting that the concept of a parallax had first been addressed – the process would enable us to exploit the illustrative, expressive style of the animation visuals whilst complementing the format of storytelling, and as a tribute to the key source of inspiration, Walt Peregoy.

Pragmatically, the decision to create our opening scene using alpha planes and a multi-camera style approach would allow the team to be ambitious with the narrative. In the audio, there is a period of time from 0:19 to 0:29 in which we have no descriptive information, only background noise that was captured from the interview.  It is during that timeframe that the parallax would occur.

In addition to that ten seconds, there is an opportunity to extend it that bit further and acquire another five seconds so that there is sufficient time for the audience to become immersed within the story – if the pacing is too fast, the animation may lose its captivation. Timing was going to be one of the elements that would present itself as a challenge for this method of visual presentation.

parralex_test

Figure 1.2 – Shown above, the full rendition of the parallax landscape. It demonstrates the transition from Newcastle, to Kilkeel, before the pan finishes on the image of pupper mountain. 

Shown in figure 1.2 is the landscape that I quickly roughed up for the purposes of exploring multiplane camera techniques within the Maya interface. I used images that Ryan and Jess had gathered during their time in Newcastle to conduct the interview, and some images of Kilkeel captured by others. Whilst I was familiar with the location, I wanted to get a better impression of the environment and village characteristics that accentuate their charm.

Figure 1.3 – Shown above, some stills captured from the footage that Jess and Ryan took of the road into Newcastle, and the Mourne Mountains backdrop.

a91fb0402a8211a252d66a2bbdd1d2acc07b13b9

(A view of Kilkeel harbour and fishing fleet taken in 2002., 2002)

Figure 1.4 – Shown above, a selection of the images that I referred to during the design for the parallax environment. 

Key aspects of appeal were the seafront and Slieve Donard Hotel in Newcastle, and the prestigious fishing and boating community of Kilkeel. I felt that accentuating those elements within the scene would provide a subtle context to the audience, and create depth that alludes to the perception of gentleness and serenity that Attracta, Jess’s grandmother, communicates through her tone of speech and her exterior decorum.

I referred to these images when creating assets for the parallax, exploiting creative liberties to stylise them appropriately. The approach was first tested in early March, a few weeks previous to the visual style to be fully established – that is why the sketchy style feels disconnected from the work that we have been exhibiting recently. However, in spite of this, the ‘sketchy’ lined approach has a particular appeal, which is what ultimately persuaded the team to agree to it.

Figure 1.5 – Shown above, the stills are taken from the parallax render test which demonstrates the result of applying lined, transparent images onto planes and layering them over one another before introducing a panning camera into the scene

In total, the test mutli-plane amounted to roughly sixty different individual frames that required placement and adjustment in order to achieve the result, which can be viewed in figure 1.6. The individual frames were challenging to export as their own separate layer, despite the option to do so in Adobe Photoshop. Although I talk about this issue in greater detail in my notebook, it must be noted that the solution was to separately save each asset one-by-one, which was agonisingly long for an expectedly simple process.

Figure 1.6 – Shown above, the parallax video in practice. This playblast demonstrates the concept that I had regarding the environment, and how it would move amongst one another to give the impression of a ‘constructed’ memory. 

The purpose of the parallax had two objectives – one was a pragmatic choice. The use of alpha-planes would spare the team time and energy that would have otherwise been restricted towards creating these assets through the usual method of model – uv unwrap – texture. Taking into consideration how quickly each member of the team works, and their delegated tasks in other areas of the production pipeline, not doing so has saved us over a week of production time that can be spent on the more important tasks, such as rigging and character animation.

The second choice was driven by psychological reasoning – the construction of a memory. According to research conducted by Jesseril Surianwinata (n.d.), there is a discussion regarding the reliability of the human mind, and how that which is perceived may be influenced by aspects such as schemas.

A schema, as discussed in a recent conversation with a psychologist (Davies, 2016), is an external belief that influences the way that we absorb and process information. Whilst everyone in the population implements this form of cognitive behaviour in some method or another, it must be noted that schemas may change vastly from one end of the spectrum to the other. As Surianwinata describes, memories have the potential to be reliable if they are in some form connected to an active schema. In contrast, a negative schema has the power to influence the encoding and retrieval of that memory.

In Attracta’s account of the family acquiring Zara, she hesitates in her action of remembrance. Reviewing her interview, and the pacing and tone through which she speaks, the bittersweet ring in her voice is likely influenced by this notion. Whilst it is a positive memory and one that played a significant role in their lives for the foreseen future, the sadness is still evident. This is a behavioural trait that can be linked to a hypothesis of reconstruction that was first experimented by Loftus and Palmer (1974) regarding transport destruction.  Whilst the context of the study and the animated documentary are entirely different, the question is identical – does the language of the interview influence Attracta’s account of the experience of pet ownership?

It is to be believed that Attracta understood why she was being interviewed and that the inevitable topic of Zara’s demise would occur at some point during the proceedings. A person may suggest that interviews contain their own language which is designed in such a way to evoke an emotional response from their candidate. The concern is that Attracta’s hesitancy and difficulty in recalling events at the beginning of the interview may be caused by more than the perceived fear of unfamiliarity with the clause of being interviewed.

The predicament provided an interesting challenge – how do you make an environment that the candidate resides in, one that they are extremely familiar with,  allude to the impression of hesitancy? How do we present the concept of memory as an unreliable source, yet so clear and concise in our conscious?

Screen Shot 2017-05-18 at 02.22.01
Figure 1.7 – Shown Above, the original concept that I considered for the introduction to the animated documentary, and transition into the environment parallax.

The answer is trial and error – the initial structure of the parallax and the locations involved had luckily been addressed in the experiment. The next stage was polishing the work and reinforcing the concept of the objectives.

This presented itself through the evolution of decisions that have been made during the production stages of the parallax. The original concept, as shown in figure 1.7, was for the implementation of two-dimensional animation as unidentifiable feet cross into the frame, then walking the width from left to right before transitioning into the outline of a vehicle. This would be at which point a simplified, lined version of the country road scene would slip into the shot (as demonstrated in figure 1.8), before finally proceeding through the landscape of Newcastle and Kilkeel.

Parallax_Early_Version.png
Figure 1.8 – Shown above, the original concept of the parallax image which would have featured a two-dimensional animation of feet approaching a vehicle before the travel commenced. Due to time restraints, this sequence was removed.

The process was not without difficulty, nor faults. Once again, I followed the same procedure that I had implemented for the parallax experiment a few weeks prior – individual assets in their own layers, then applied to an alpha plane. The task proved more demanding now that we were rendering in Arnold; my previous method of applying lambert to plane and being able to view the effect was now not sustainable. In addition, the process resulted in duplicate planes being created for the lines, enabling the colour to remain in its isolated plane, as well as the alphas for each asset. The process was incredibly tedious, and consumed more time than I had originally scheduled. An example of a number of versions required is illustrated below, in figure 1.9.

Figure 1.9 – Shown above, the process that each asset within the parallax folder required in order to operate accordingly once plugged into an alpha plane. 

In addition to being a meticulous process, the alpha planes were not without faulty presentation. In figure 2.0,  the issues that occurred during the apply – render – adjust – render methodology are distinctly evident. In Arnold renderview port, the planes presented themselves as a very dulled version of the vibrancy and depth. Additionally, other frames were presenting themselves as solid black spaces in spite of the fact that the textures had been applied in the exact same process as those that are demonstrated to have worked. Furthermore, adjusting the render settings to try and discover the issue addressed showed that other settings were causing textures to appear white-washed and bleached, almost as if they had been exposed to sunlight. The skydome light that was within the scene had a set intensity of below 2.0, the standard to which the rest of the scenes within the documentary have been rendered. Frustratingly, the technical faults delayed the process of completing the parallax and rendering the entire sequence and due to the proximity to the deadline, this was an unexpected nightmare.

Figure 2.0 – Shown above, some of the technical malfunctions that occurred during the process of setting up the parallax in Autodesk Maya. 

Presenting the piece so far to the team and informing them of the difficulties that were occurring led Katie (thank you!) to remind me of a comment from a time when the parallax experimentation was presented for feedback – the fewer layers, the less mess of trying to implement the technique. It was then that I realised that my current process was unsuitable and that if I proceeded the way that I had spent the past few long days fighting over fidgety details, I would be putting myself and my team in a very difficult position.

I felt that it was important to revisit the information that was gained during the research stage for the first experimentation; investigating mutli-plane camera set-ups. The technical difficulties were distracting me from achieving the intended purpose of exploring the parallax technique, and as a result that I was not witnessing a benefit from adapting this process in comparison to placing a flat image within the scene and animating a panning camera.

In my research, I discovered a video-based short media study that discussed the origins of the multi-plane storytelling approach in ‘The Adventures of Prince Achmed’ (1926), in which the story processed through a series of silhouettes that created the illusion of a character. The film featured a number of influential Avant-Garde artists, due to the unusual approach to storytelling that the director explored for the premise. In a similar fashion to the German Expressionist art movement, it was to be expected that the West would adapt the technique once artists immigrated;  a result of the economic and social factors within the country at that time.

The most known use of the multi-plane technique was featured in productions from the Walt Disney ‘Mouse House’. The first appearance occurred during a sequence in Bambi (1942), implemented as an iconic woodland scene in which the character resides.  The technique continued to be used into the 1990s, most notably in the pride rock scene from ‘The Lion King’ (1994).

Figure 2.1 – Shown above – Disney In Honesty (2016) – the short video study on the multiplane effect. 

In the video featured in 2.1, the main prerogative of the parallax process is stated – immersing the audience within an environment. The primary objective is to create the impression of depth, as this enables the audience to understand the visuals as a mimic of natural form. In similar frustration to the fact that people look very different in photographs as opposed to real life, the environment can also fall victim to the focal length of a camera lens.

Depth invites curiosity, as the audience may find themselves captivated by the different elements that move on alternative timings and pace. The impact of a layered setting has the potential to make the audience feel like a voyeur within a private moment.

Whilst that may appear as a negativity, it is my intention to create this impression. The interview with Attracta is a very personal reflection; a result of an invasive process. The audience should not immediately sink within the piece – instead, they should be held at a personal distance. It is only once puppy Zara interacts with the camera (and by extension, the audience) that full immersion may occur – peak too early, and the integrity of the piece could be compromised.

Figure 1.9 contains what was present in the finalised version prior to being rendered.

Parallax_Complete_Image
Figure 2.2 – Shown above, the completed (flattened) illustration of the parallax.

Similarly, maintaining a balance between audience interaction and artistic exhibition, compromises had to be made during the production stages. In response to the feedback that I had received (and forgotten about!) regarding the reduction in the layers,  I revisited the approach entirely.

The photoshop document containing the individual assets was now the source file from which I would take the individual assets, and compile them together to create four different layers – a background, a midground, a centre, and a foreground. The centre layer – the road that cuts through the entire scene – was isolated with the intention of creating a depth of field, but after reviewing this with the rest of the team it was decided that doing such an action would negatively impact our production schedule and the resulting payoff would not be worth the pressure.

In figure 2.2 the entire landscape of the parallax can be reviewed. The scene was adapted in order to confine to the time restrictions and optimise the scene for rendering. The scene does feel less engaging than the one that was created for the experiment, but after witnessing how crammed the effect had at particular points in the timeline, I feel that this decision will lead to improvement in the natural flow and procession of the landscape.

The difficulty with the transition was the flow into ‘pupper mountain’, as it is the pivotal point where the medium of storytelling changes. To ensure smooth flow, it was a constant cycle of movement  – camera – render – adjustment. The hills contain a solid colour, whereas the pups have a mixture of cream toned fur with brown patches, which was an added element of difficulty within the piece. In response to this, I have placed the background layer (featuring the hills) behind the pile, and the foreground layer (with a blanket and hill) closest to the camera. This effectively encased the puppies within the environment, and from what could be witnessed by scrubbing through the Arnold renderview, appears to have been successful, reinforcing the concept of the mountain becoming integrated within the landscape.

Figure 2.2 – Shown above, some stills from the test render in order to check the effectiveness of the technique. 

Figure 2.2 illustrates the first test with the four layer process. The stills demonstrate that the colour and lineart are now working effectively – much to my relief, as previous attempts of layering the colour and line separately provided to produce less than satisfactory results (Figure 2.3). Unfortunately I was an idiot who rendered in .exe files, and as a result, cannot compile the images together for an accurate playback. However, scrolling down the individual images in quick succession gave the impression of how it looked, which gave me some ideas for adjustment. I could see that whilst everything was working as intended, the benefit of the multi-plane camera was not being utilised as effectively and the image seemed rather flat. It was lacking that unreliable and constructed, jigsaw-esque quality that had made the technique appealing.

Screen Shot 2017-05-16 at 23.34.39
Figure 2.3 – shown above, a rendered example of how the parallax looked at an earlier stage when the lineart and coloured underlay were on individual layers.

With this knowledge, I returned to the Maya scene in order to plan the adjustment. The stationary positions of the layers worked rather well with the camera, as everything appeared on screen as expected and there was no evidence of malfunction within the scene. In that respect, the layers were in an optimal state for some subtle panning animations to be implemented onto the scenery. The background panel received the least bit of animation; panning to the left as the camera arrives on the image of ‘pupper mountain.’ The middle ground and road received the most editing, with panning animation occurring in both the left and right direction. The foreground, with its close proximity to the camera lens, was more subtle to give the illusion that assets were moving at different speeds; an example that had been predicated in Bambi (1942) to achieve that depth within the landscape.

The unfortunate drawback with Arnold rendering and AIstandard materials is that within the user interface, the alpha-planes are a distorted mess. Whilst the Arnold renderview port permits some form of preview, the full formation of the piece is never cohesively clear until a render is completed.

Figure 2.4 – Shown above, the introduction to Pethood, featuring the Parallax and transition. 

Figure 2.4 is the completed render for the introduction sequence to Pethood. Whilst there are some adjustments to the animated pacing that I would love to have made, I feel that the piece is entirely sufficient for the purpose of the hand-in. My intention is to return to the piece post-deadline, and make changes to the timing in preparation for the final year show.

In spite of those small amendments, I am pretty delighted with the outcome. The result achieved my intentions, and in my personal opinion, evokes the psychological reconstruction that I have discussed in my research. The challenge for the piece was executing the concept that Attracta, despite living and being familiar with the area, would have difficulty with immediately recalling the memory. The use of colour within the sequence introduces vibrancy, which was intended as an expression of happiness and good connotations; it illustrates to familiarity within uncertainty. The movement of the panels was to signify the process of cognitive memory as the experience is brought back to the forefront of Attracta’s mind – this construct approach representing the secondary details that tend to resurface after the primary focus; Zara, in this case, changes those visuals significantly.

It is my personal belief that the representation of cognitive thought was greatly suited by the parallax because of that evolution. The animated documentary has three significant forms of media; live action, two-dimensional illustration, and the three-dimensional environment. The flow of the story is not only driven by the narrative given by Attracta, but also by the changes within the medium of choice. The further the audience is separated from the live action footage, the less the story focuses on the human perception of the piece; it becomes Zara’s story.


Bibliography:

A view of Kilkeel harbour and fishing fleet taken in 2002. (2002). [image] Available at: http://www.bbc.co.uk/northernireland/yourplaceandmine/down/A721757.shtml [Accessed 17 May 2017].

Bambi. (1942). [film] Directed by D. Hand. RKO Radio Pictures, United States: Walt Disney.

Davies, M. (2016).

Disney In Honesty (2016). Bambi: The Multiplane Effect. Available at: https://youtu.be/x66_BSGEeYY [Accessed 18 May 2017].

Loftus, E. and Palmer, J. (1974). Reconstruction of Automobile Destruction: An Example of Interaction Between Language and Memory. University of Washington, pp.585-589.

The Adventures of Prince Achmed. (1926). [film] Directed by L. Reiniger. Germany: Comenius Film.

The Lion King. (1994). [film] Directed by R. Allers and R. Minkoff. Buena Vista Pictures, United States: Walt Disney Pictures, Walt Disney Feature Animation.Surianwinata, J. (n.d.).

Surianwinata, J. (n.d.). To What Extent is the Memory A Reliable Process?. International Baccalaureate.

Advertisements

2 thoughts on “Creating the Parallax: Research, Rationality and Implementation

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s