With drawing and animating water, simplicity is paramount - much more so than other elements due to the sheer amount of detail involved in even the smallest action. Gilland (2009, p.95) recommends learning to identify the main formations seen in water references, and then break those down into simpler, graphical shapes. He says “if we attempt to draw all of it, we will quickly be overwhelmed by the details”, and in addition to this, if after struggling through the details we do complete the animation, it could become too visually messy for the audience and thus lose its appeal. Simplification and stylisation are therefore incredibly important for this element.
Before getting onto the theory of a water splash, both Gilland (2009) and Kathleen Quaife (2013) talk about the importance of good perspective with a water splash effects animation. It is common to think that perspective isn’t that important due to the abstract nature of water, however if the perspective is wrong, the “viewer will feel that something is amiss, and their attention will be drawn away from the story” Gilland (2009, p. 96). Where this is prevalent is within the water ripples. Often times, the ripples are animated outwards from the base ellipse with even spacing (which is fairly intuitive as they will in fact be moving at roughly the same speed), however as the water splash is three dimensional, the ripples furthest from the viewer should be closer together than the ones nearest to the camera. Drawing a ground plane and a base ellipse over that grid to represent the water splash will help with the animation’s perspective.
Theory:
To help understand the theory behind a water splash, I used information from Gilland (2009), Garrett Wren (no date), Kathleen Quaife (2013) and Howard Wimshurst (2018):
The primary force in a water splash is caused by an object entering the water. The size, speed and shape of the object will affect the type of splash generated. When the object enters the water, it will displace the water and cause it to explode outwards from the impact in a sheet. This sheet will be carried on a trajectory dependent on its initial velocity outwards, slowing down as it is affected by gravity. As the sheet gets higher, it also thins, and the surface tension causes holes to break the surface. The holes expand, eventually causing the sheet to break into droplets of water. The period where the sheet reaches its apex is called the hang time. This is when gravity has overcome the upwards velocity, causing the droplets to start arcing downwards. Whilst this initial splash is happening, the impact caused by the object also causes ripples along the water’s surface. When the droplets crash down from the primary splash, they create their own ripples that expand outwards. They may also create their own little splashes as well depending on their size.
Once the object enters the water, an air pocket is formed due to the displacement. After a fraction of a second, the water then rushes into this air pocket and gets launched upwards creating a secondary splash, again creating more ripples along the water’s surface upon landing.
As you can see by the theory, there will be a lot of overlapping elements within a water splash, which is what makes it so dynamic and appealing to look at. Playing around with each part’s timing, as well as the length of the hang time will give different results and effects. Another point to be wary of is that the primary splash creates a lot of droplets, however when animating them hitting the water’s surface it is important to not add too many secondary splashes and ripples to avoid “making your animation overly busy”, Gilland (2009, p. 121).
The size of the object entering the water affects the size of the water splash, however different sized water splashes will have different timings. Gilland (2009, p. 102) says “the key to the size and duration of a splash, is how much water is being displaced”. A raindrop splash is the simplest water splash to animate due to its small size and thus length. They can last as short as 2 to 5 frames and don’t have a secondary splash (especially if the rain falls on a solid surface). For a slightly larger object - such as a pebble - it can last about a half second, with the ripples being very short and the secondary splash being either non-existent or a small, 4 to 6 framed droplet splash. A medium sized splash usually lasts a second to a second and a half, and a larger (big rock) one would last about 2 to 3 seconds, with the secondary splash being almost as big as the primary. These are all timings to play around with, but generally the larger the object is the more water it will displace, and thus the slower, higher and more detailed the animation will need to be.
You can see this theory in effect in the following video:
Reference Video by Videezy
This reference doesn’t have the holes forming in the primary splash, and after studying other references I believe this is due to the smaller size of the splash - as the primary splash doesn’t go up that high the water is not stretched enough to break the surface tension. I also noted with these studies that the overlapping timing of the secondary splash provides a lot of the appeal.
MaginPanic (2017)
Matt Timms (2011)
Chris Graf (2020)
Chris Graf’s (2020) water splash was in the same style as MaginPanic’s, however on a larger scale. This meant that the hang time was even longer and the animation was a bit more detailed overall, as well as having more evident ripples from the crashing down of the primary splash’s edges. The secondary splash of this animation was also much larger.
Studying all of these I noticed that deciding when to show the secondary splash is important, as it gives a lot of information about the size of the splash. Generally, the longer time between primary and secondary splashes, the larger the splash. I quite liked the longer durations as it also made the animation more punchy.
Before animating a proper water splash, I decided to warm up by animating a raindrop, as it’s much shorter duration would give me a feel for animating water:
Now for the actual water splash:
I found this very challenging, mainly due to all the elements I had to keep track of as well as because it’s animated straight ahead, which was necessary to avoid it looking too mechanical. I studied the animated references for the right timings to use and then just started, focusing on creating interesting shapes and following the perspective grid I had set up.
I animated up to the peak of the primary splash, where I then went back and started adding the holes in as I wanted my splash to break apart and come crashing down in two sections - the bigger drops and the main body. From there, I animated the holes and the main body, before going back to animate the bigger drops. I really wanted to focus on getting strong, punchy timing, therefore I decided to have the hang time be quite long and the crash be quite short, which I think I overdid slightly. I feel I could have increased the duration of the primary splash’s crash.
The drops were probably the most challenging part as following the arcs and having overlapping timing was quite difficult, and I feel like they are the weakest part of the animation. I should have spent more time studying how other animator’s approached them.
For the secondary splash, I wanted to have quite a big duration between it and the primary one to create a swing-like timing, which I think worked pretty well.
To add polish and to make the animation more believable, I added smaller splashes for the drops and for the secondary splash, as well as adding the ripples. I found the main splash’s ripple challenging due to the same problem as with my observational drawing - circles in perspective.
Overall, I was really paying attention to the theory learnt from my research to make this animation as appealing as possible (namely the punchy timing, overlapping actions and strong shape design) and I think despite it not being professional quality, it is significantly more appealing and believable than my initial water splash animation. The timing is much stronger, and the animation has much more energy and character to it. The shapes are also much more realistic and appealing to look at due to my observational drawing practice. Finally, one of the biggest issues with my initial animation was the start of the splash, as it moved too slow and uniformly, resulting in a pie-like effect. This animation definitely fixed that issue, with the splash feeling like it’s truly reacting to something rather than moving of its own accord.
I animated up to the peak of the primary splash, where I then went back and started adding the holes in as I wanted my splash to break apart and come crashing down in two sections - the bigger drops and the main body. From there, I animated the holes and the main body, before going back to animate the bigger drops. I really wanted to focus on getting strong, punchy timing, therefore I decided to have the hang time be quite long and the crash be quite short, which I think I overdid slightly. I feel I could have increased the duration of the primary splash’s crash.
The drops were probably the most challenging part as following the arcs and having overlapping timing was quite difficult, and I feel like they are the weakest part of the animation. I should have spent more time studying how other animator’s approached them.
For the secondary splash, I wanted to have quite a big duration between it and the primary one to create a swing-like timing, which I think worked pretty well.
To add polish and to make the animation more believable, I added smaller splashes for the drops and for the secondary splash, as well as adding the ripples. I found the main splash’s ripple challenging due to the same problem as with my observational drawing - circles in perspective.
Overall, I was really paying attention to the theory learnt from my research to make this animation as appealing as possible (namely the punchy timing, overlapping actions and strong shape design) and I think despite it not being professional quality, it is significantly more appealing and believable than my initial water splash animation. The timing is much stronger, and the animation has much more energy and character to it. The shapes are also much more realistic and appealing to look at due to my observational drawing practice. Finally, one of the biggest issues with my initial animation was the start of the splash, as it moved too slow and uniformly, resulting in a pie-like effect. This animation definitely fixed that issue, with the splash feeling like it’s truly reacting to something rather than moving of its own accord.
No comments:
Post a Comment