Going Live update 15/03/23

Going Live update                                                                                                                                           15/03/23

Figure 1: Animating Nikko in Maya

I've currently moved onto animating Nikko in Maya whilst Tinny is currently still being rendered, and despite some periods where the process has been particularly stressful, the animation is becoming smoother and more refined.  Starting off, the feet proved to be particularly troublesome, mainly when it came to getting the positions just right in the midframe, but I've managed to find a workaround for this.  Despite the obviously much smaller proportions than the Kayla rig I've been using previously, I've managed to transfer the movements from the basic walk tutorial fairly well.  I'm moving onto the fine tuning section of the animation, and I do know that I will need to update the reference model in order to save time on animating again.  Thankfully this is a model that, for the most part, looks decent on almost every angle.

The adjustment of the animation in the graph editor will prove to be particularly tricky given how the movement has reacted to some tweaking in the past.  The first run through of the animation as a basic walk cycle has turned out smooth enough.  It may well need some more work when showed to my peers but hopefully it has still made the grade.

Figure 2: Nikko graph editor

Figure 3: Nikko walkcycle test

Figure 4: Nikko double-bounce graph

The double bounce walk for Nikko, adapted from the standard masculine walkcycle, usually requires slowing the animation cycle down from 24 frames to 33 frames so that the bounce is less sudden and jerky.  It also allows the viewer more time to observe the movement.  This is the method employed by the tutorial I've been following, and it's translated very successfully.

Figure 5: Double bounce walk cycle:https://www.youtube.com/watch?v=uTk3PYTTrJw

Figure 6: Double bounce walk cycle

Advanced Production Update 13/03/23

 Advanced Production Update                                                                                                                 13/03/23

Figure 1: Pencil sketch filter: https://www.google.com/url?sa=i&url=https%3A%2F%2Fwww.photofunny.net%2Fcat-image-processing%2Fconvert-picture-drawing-a-pencil&psig=AOvVaw3YUnGBlVYIHP9_itKk_hPO&ust=1678809286950000&source=images&cd=vfe&ved=0CA8QjRxqFwoTCPiBsoui2f0CFQAAAAAdAAAAABAE

I've begun experimenting with some more pencil sketch filters and some new footage to create some more innovative shots.  The same criticism could easily apply to them in that the static footage is looking very unprofessional but that's been well established and with some well-staged reshoots this can be resolved.  I am aware that this technique might not be entirely relevant but I believe it can be classified as VFX.  The technique is accomplished by clearing up the footage in After Effects and then applying different filters and textures to them in Photoshop before taking them back to After Effects for final compositing and colour correcting.  The key to making this effect work is to make sure the animated footage is concentrated, so that the objects can be easily identified by the viewer.  

So far, only three or four shots have yet to make it to the first draft phase.  Once these shots have reached this stage, I can focus on narrowing down the final edit of my film.  I recognise that I need some extra screen recording footage for my assessment video as well as my showreel, so for my next compositing I'll have to make sure I've been recording video of it.  I will also need screen recording footage of me editing the rough cut of the final project to give evidence of my video editing knowledge.

Figure 2: Typewriter composite test

Shots that have reached the first draft phase: 

• Typewriter test
• Pictish beast
• Pictish penguin
• Pictish figure
• Rotoscoped handjoin
• Glasses filter
• Pictish snake
• Pictish boar
• Pictish deer
• Jute growing
• Harp test
• Lava

Shots that need to be composited/started:

• Rotoscoped distortion
• Flying whale
• Rotoscoped vision
• Alien landing

Things to make screen recording of:

• Animating in Maya
• Compositing 
• Animating in TVPaint
• Video editing

Going Live Update 10/03/23

Going Live Update                                                                                                                                      10/03/23

Figure 1: Rendering tests

Prior to the next meeting with Wild Child, I've been experimenting with the Kayla rig in an attempt to make a more refined strut cycle.  In order to make the arm movements neater and more refined, I've decided that for the strut, it's better that the fists are closed because it shows that the character displays more confidence, as well as the arms being straight against the body to avoid bad arm manoeuvres when viewed from the front.  The lack of central poses in the strut cycle make the animation process harder than the vanilla walk cycle which had poses that were easier to map out.  The strut cycle has four poses between the start and middle and four poses between the middle and end meaning that whilst the waist position can be copied and pasted, there is no definitive pass pose yet, which makes it hard to make these sections fluent.  I must also confess that I've been doing the poses in the masculine style with a female character, mainly because the characters in our story are both male.

Figure 2: Homemade strut walk cycle

The double bounce walk cycle was made by simply adapting the standard walk cycle and replacing the up poses with down poses to make a bouncing effect, as well as by spreading out the cycle from 24 frames to 33 frames to make the animation slower and the bounce movement less sudden.  I've also found methods of animating the camera to perform an orbiting shot and other camera tricks.   This method will be perfect for Nikko, since this movement will easily match his persona, whilst the strut cycle we've agreed on for Tinny will be harder to pull off.   

Figure 3: Double bounce walkcycle

Figure 4: Maya Graph Editor: https://www.google.com/url?sa=i&url=https%3A%2F%2Fforums.autodesk.com%2Ft5%2Fmaya-animation-and-rigging%2Fonly-purple-curves-in-graph-editor%2Ftd-p%2F9559946&psig=AOvVaw1XNRtOeMbrkeB2ajCtpX9c&ust=1678534932018000&source=images&cd=vfe&ved=0CA8QjRxqFwoTCLCI7ISk0f0CFQAAAAAdAAAAABAE

I've also discovered how tricky the Maya graph editor is.  Whilst it is essential to master, the curves on it do require a lot of patience and care other wise the movements can become erratic.  I'm going to have to learn Maya for many years to fully avoid these sorts of shenanigans. The main trick is to make the tangents flow into each other organically whilst keeping an eye on the figure movement. The main problem child I've found with the Kayla rig at the moment is the right arm which is getting obnoxious and hard to fully control.  I really need a tutor overview to get this sort of thing ironed out.

Figure 5: Strut compilation

 

Advanced Production Update 07/03/23

 Advanced Production Update                                                                                                                      07/03/23

Figure 1: Salmon composite test

This is the latest test of VFX composition with my salmon animation, using a low angle shot of a bridge beam as the basis.  The animation was adapted from a royalty free model I found online which I adapted into a Pictish salmon.  I think this one turned out better than the first attempt because it had a longer body which was closer to the Pictish symbol, and a flow of plausible movement that could be turned into a GIF-like cycle that made animating smoother and more efficient.  I also found it easier to incorporate the animation into the composition with flowing, graffiti-like staging and a relatively simple way of getting the angle right to make it look like it's been painted on the beam.  I've also added some simple little bubbles on either side of the composition to add some underwater theming to the shot, as well as the necessary glowing on the fish to make them distinct from reality.  Whilst the original shot needs to be reshot to add the necessary movement on the camerawork, the VFX should require minimal alteration at best.

I've also been experimenting with some pencil sketch filters on my rotoscope videos, since whilst they look less rustic and personal than my usual method, they can produce more refined results and can be infinitely more efficient.  The main thing I've found is that they generally work better when they are filmed on a green screen, because you still need to rotobrush the footage in after effects to keep the video focused.   I've tried applying video filters in photoshop to the footage to make it resemble an outline done on a sketch, and whilst the results can be very precise, it also picks up everything in the background.  It should also allow more time to work on the compositing and reshooting the original shot if needed.  This is an occasion where static shots to be rotoscoped work better because the constant movement in frame can make it harder to rotobrush.

Figure 2: Pencil sketch filter test

This is the latest pencil sketch attempt I've tried on Photoshop where I took the footage and ran it through a tutorial designed for an image.  Whilst there were inevitably points where the functions were clearly designed for a still image, and when I compressed layers they were just turned into one still, most of the functions and filters seemed to work essentially as they did in the tutorial.  The main challenge at this point is to find a video that has minimal background interference.  I will continue to experiment with different filters and methods since this is proven to be an efficient method.  

Advanced production update 06/03/23

Advanced production update                                                                                                                      06/03/23

Figure 1: Caird Hall: https://en.wikipedia.org/wiki/Caird_Hall

I've just finished the first composition test of the Caird Hall shot with a roto-brush effect so that I can have a figure of myself walking on top of animation.  The composition needs some improving for the second draft, including the incorporation of the double discs and a glow effect to amplify the animation's distinction.  I've found that the rotobrush makes it a bit harder for the computer to render footage whenever you add new effects, but this is something that isn't a major gripe.  The bezier warp effect makes the effect on the animation blend into the background fairly effectively.

Figure 2: Harp composition test

I'm also currently developing my Pictish beast composition with shot from around West Ward Works.  I've applied some royalty free footage of some mist to try and add some more atmosphere to the shot, and used 3D to make the Pictish beast appear as though it's flowing in and out of the buildings.  I initially tried using a divide filter on the animation, but that made the glow effects behave weirdly, so I had to apply the invert filter to make the glowing filter work properly.  The next draft of the video will definitely require some better timing on the animation, with maybe some more frames in the TVPaint file so that it can appear to flow better.  This is one of many shots that will need to be redone with a moving shot so I can move on.  Overall, I am happy with how this shot has turned out, but I recognise that it might have some issues to iron out.  This shot also needs some glowing lines to indicate the impact point where the beast hits the wall and passes through.

 

Figure 3: Pictish beast composition test

Shots that have reached the first draft phase:

• Pictish figure
• Pictish penguin
• Pictish beast
• Hand joining
• Deer animation
• Lava
• Pictish snake
• Boar running
• Glasses filter
• Jute growing
• Harp playing

Shots that need to be started/finished:

• Pictish salmon
• Rotoscope distortion
• Rotoscoped vision
• Jute machine rotoscope
• Whale flying overhead
• Typewriter rotoscope
• 
  

Figure 4: After Effects Pictish beast composition

Advanced Production Update 05/03/23

 Advanced Production Update                                                                                                                    05/03/23

Figure 1: Panning photography:https://digital-photography-school.com/6-tips-master-panning-photography/

We're only about four weeks away from the end of the semester, and progress on the visual effects for my final project is going more or less OK.  I've just managed to rotoscope enough footage for a harp shot which will accompany a shot of the Caird Hall to illustrate the location's history and current function as a musical venue.  After my most recent group tutorial, it's become clear to me that my current approach to the project is very much flawed.  I've rescinded a point about possibly doing a showreel instead of Seoras, which was just foolhardy in retrospect, and I've made it through about half of my shot list to complete with about six or seven shots left to reach the first draft phase.  I've also found the methods necessary to improve the shots I've done so far.  However the faults with the film so far are ones that have been recurring throughout the series in that the film is too long and will result in too much work for the timeframe we have available, and that static shots simply will not do, no matter how many effects or plugins I dress them up with.

Admittedly I initially went with static shots because they would make the composition a bit easier, but I probably should've tried to find a find around this.  Upon the tutor's insistence, several shots will have to be redone, including the shots of the walls, the shot at Cox's stack (though I have a panning shot on my hard drive which could do the trick), the shot of the lava flow, and basically any shots intended for VFX where the camera isn't moving.  It will be a bit obnoxious but it can be done.  These shots will definitely be included in the final draft of the film, as hopefully a more rough draft will suffice for the video material for advanced production submission.  If I wanted to do a drone shot for Cox's stack, I'd have to ask permission from video owners to use some of their footage, since the airport has made it illegal to fly drones just about everywhere in Dundee.

My rotoscoping and animation practice also needs improving, as my drawings are too detailed and too frequent to be productive enough.  I'm aware of a plugin for After Effects called Mocha Pro which easily speed up the rotoscoping but it's expensive and obviously, only the administrator can install it.  The main option I have is to use the AI the tutor recommended, flaws and all.  If this doesn't work, I need to develop an animation technique where I draw on fewer frames per second and with less detail, given that my current approach is to animate on twos and to draw the frames with as much detail as possible.  If I try a looser approach, such as animating on fours and using less details and more abstract designs, maybe this could yield a faster animation process.  I have to find a technique that doesn't require subscriptions, such as the AI known as Anaconda, which uses python coding to install and utilize it's features.  

Figure 3: Anaconda figure rotoscope test: https://www.provideocoalition.com/automatic-rotoscopingfor-free/

My tutor has been given a copy of the folder I've been using for my project to try and salvage something from what I've filmed so far.  With the animations of figures for my project, it should be a bit more forgiving because I can draw enough keyframes to the point where I make the animation gif-friendly and produce loops which make compositing slightly easier.  I will aim to do this technique with the Pictish fish and other animations that focus on one figure.  In terms of the new VFX shots for my film, my main objective will be to add tasteful movement to the shots to make them more interesting and show them to the tutor before I proceed to compositing.  With making the final compositions, I need to remember to put in the extra effort to make them distinct enough from reality with effects such as glow, displacement map and turbulent noise with mist.  

Figure 4: https://www.youtube.com/watch?v=9nLALXAM5aU

Second Assessment Video draft script 01/03/23

Second Assessment Video draft script                                                                                                     01/03/23

Figure 1: Assessment video: https://www.google.co.uk/url?sa=i&url=https%3A%2F%2Flearn.solent.ac.uk%2Fmod%2Fbook%2Fview.php%3Fid%3D1270730&psig=AOvVaw0bNtRnqDoaPttrhWKn0L2P&ust=1677749911388000&source=images&cd=vfe&ved=0CA8QjRxqFwoTCJDx_823uv0CFQAAAAAdAAAAABAE

My name is George Salisbury and this is my assessment video for the modules Going Live and Advanced Production.

For our client project in Going Live, we were assigned to create a pitch for a children's television series for the Stirling-based animation studio Wild Child, a project that they could easily put into production and produce a successful series. To save time on development, it had to be based on a pre-existing property in the public domain.  After much discussion and debate, we settled on developing a spin-off story based on L Frank Baum's The Wizard of Oz, taking place after the departure of Dorothy, the main protagonist. The two main characters we agreed to move forward with were a revised version of the Tin Man, whom we called Tinny, who in this version was younger but still well-mannered, compassionate, moral and helpful.  This also contributed to the title we agreed on for the show, Tinny Adventures in Oz.  A glowing heart feature was added to indicate the lessons taught and the journey that an episode displayed. The other character we agreed to focus on was the winged monkey Nikko, who originally was a servant of the Wicked Witch of the West but is now a more independent figure.  Whilst he is the usual troublemaker in the pitch's narrative, he is not irredeemable and can learn from his mistakes and improve himself, making him an essential reflective figure for the show's audience Philip even created a brief animatic of an example episode that demonstrated the character's purpose in the series, with Tinny  The age demographic we decided to aim for was between four and seven years ago, and so the characters had to be aged and proportioned to resemble said age group, with short statures and rounded body elements.  It was a team effort, and we all contributed to different elements of the process.  I volunteered myself to be involved in the concept art alongside Sachien, as well as the animation stage. My concept art was more suited to reference material rather a direct basis, but it was valid input nonetheless.  The main contributions I made in this field were a map of Oz and some maps of the Munchkinland set piece that could be used to plan how to block out and place assets in a scene.  I had some experience with animation previously in 2D with TVPaint and After Effects, but this is the very first time I've tried animating in 3D, and so a lot of practice became necessary.  I tried animating with different models, but I always acknowledged that I would be working with short characters.  The main rig that I could find that was best for the task was a character called Kayla, a young child close to the age demographic with a simple stature and movement capability that could be recreated with Tinny and Nikko, especially with the latter who stature was to be more humanoid than monkey-like.  Even though the rig's flexibility was more exaggerated than our main protagonists, it did put me in good stead for animating them.  The movement in the animation had to be engaging but nothing melodramatic, so I couldn't just do a simple walk cycle. The best walk cycles we agreed on were a strut for Tinny to indicate his confidence and mannerism, and a double bounce walk cycle for Nikko to indicate his relatability and more childish nature.  It is also said that the double bounce walk is said to be essential for a person to walk comfortably and how humans have evolved over time.

So what we found is that it is really crucial in order to walk economically, to load your ankle joint, to then release this torque really explosively like a catapult to shoot your leg forward. - (Renjewski, 2023)

Here is an example of our production pipeline on Tinny (insert pipeline motion graphic here) 

Figure 2: Animation pipeline: https://www.google.com/search?q=animation+pipeline&rlz=1C1GCEB_enGB1047&sxsrf=AJOqlzVGIRZlqBZpEJ6F7qnnKsaOeEAnEQ:1677666331252&source=lnms&tbm=isch&sa=X&ved=2ahUKEwjThOeNwrr9AhVQXsAKHbCkDasQ_AUoAXoECAEQAw&biw=1401&bih=932&dpr=1#imgrc=pmd72jXCmehwlM

The Advanced Production module was mainly centered on developing our final project in stages such as storyboarding, scriptwriting, crafting an animatic and other pre-visualisations for the final module in the last semester, as well as refining practice with 3D.  This involved sessions developing skills in Maya and Zbrush with techniques such as rigging and sculpting.  These sessions were slightly more troublesome at times than I would prefer, but I still kept persisting with them as much as I could, and led me to conclude that I definitely prefer Maya to Zbrush at this point.  I could adapt to Zbrush over time and I know how to import models from Maya into it, but it will take a lot of practice and persistence.  It was during this time that I filmed the live-action footage (over several days as opposed to just one as I had originally planned), so I had a basis for my animated footage.  I also discovered during the crafting of the animatic that several shots really did need the effects incorporated to serve a purpose.  I was aware of several faults going into this stage of the project, such as an overlong treatment and runtime, and some deeply flawed shot structure, and a very rough live-action cut that lasted over ten minutes in length and a narrative flow that just didn't quite work.  The main takeaway from the tutor is that static shots simply will not do in a professional shoot, so I had to go back and redo the footage where the camera wasn't moving but still focused on a still object for the 3D motion tracker to work on.  The first showing of the rough cut was, in hindsight, a rather poor performance.  At this point, I began to consider making a showreel of animation for my final output, but because we still have some salvageable assets, I will continue forward with the production.  I was also in a place where I had too much footage, which all things considered is like a safety net.  The animation stage of my project was certainly distinct from my colleagues, with a greater emphasis on 2D as opposed to 3D.  The animation was consistent of rotoscoping, and 2D animation characters, with a distinct focus on rustic styling and linework to resemble the Pictish carvings that were the basis for my film.  Admittedly, you can tell that this is my time doing animation to this extent, with some minor issues with timing and movement here or there, but nothing that cannot be refined over time.  In terms of compositing the animation, I had originally starting out with importing the animation from TVPaint into After Effects, adding an invert filter and practising with different blending modes.  Upon feedback from the tutor, I knew that these shots needed more visual flair to make them more distinctive from reality, such as a glow effect, an expression on the transform filter to make the camera movement more engaging, or an additional, surrealist video filter with mist or grain. It was from this experience that I learned the main difference between compositing animations in After Effects and using After Effects to make motion graphics and graphic designs. The final cut of my project will hopefully have a greater emphasis on the animation and VFX elements from my work, as the degree was originally focused on.

  

References

Rascoe, A. (2023) New research finds why humans walk with a 'double bounce', NPR. NPR. Available at: https://www.npr.org/2023/01/08/1147737212/new-research-finds-why-humans-walk-with-a-double-bounce (Accessed: March 1, 2023).