This animation production includes all the experiences and techniques I have learned in the past few years. On top of this, I realized many of my shortcomings and learned more knowledge, which benefited me a lot.
For me, animation is not just learning. It is my hobby. This mermaid animation has been planned in my mind for many years. For this reason, I also learned how to make this animation in other projects, including applying unreal. , camera learning, stone making, baking from high-poly to low-poly, etc.
Although the production process could have been smoother, I also made many mistakes. The most serious mistake made me waste three days. But I still made the animation what I wanted, which was inevitable without the help of others. I am very grateful to my classmates and teachers for providing beneficial suggestions regarding my technology, aesthetics, lenses, etc. These suggestions have really improved my graphics to a new level and made my animated characters more flexible.
Because I need to improve at setting up scenes, the convenient mechanism of Unreal software allows me to realize the pictures I imagined. In the coming days, I will work harder to learn this software and enhance my skills for new ideas. Make better animations.
Last week, I finished all the shots and animations and entirely completed the production of the animation pictures. I will take the working files from the dubbing studio to further modify them to ensure that every sound matches the dubbing. And can have appropriate background music.
First, I will complete the animation version, use After Effects to synthesize and export it, and then import the exported video into Adobe Audition again to adapt the sound effects.
I found that some sound effects could be more suitable because the animation has changed. The sound effects are too loud, so I corrected these minor problems before exporting new sound products and putting them back into After Effects for final synthesis. This can ensure the correspondence and accuracy of the sound effects.
Finally, I added the background music that I collaborated with Weiqi Feng to thoroughly complete all the animations, which were synthesized and exported in After Effects. My animation is complete after synthesizing all the data files and creating the showreel.
Last week, I finished making the final maps that needed to be changed and adjusted the depth of field of all the shots. This week, I’ll be doing a final check of all the shots before rendering them out. I did not have a lot of animation work this week; most of the time was spent writing my thesis.
I also selected a few nice-looking screenshots to show the adjusted depth of field:
Although some areas needed to be adjusted depending on the lens, the lighting and animations were adjusted in the end, and all the animations were rendered.
At the same time, I am following up on the progress of sound effects and BGM with my collaborators.
Since the color of the ground is really strange, the more I look at it, the uglier it gets, so I worked with Kaiyue on how to change it. She gave me a lot of advice, and eventually, I decided to change the color of the ground so that the ground blends in more with the environment, with a blue layer.
This week, I refined all the footage to make sure all the effects and animations were okay and bug-free. I also tweaked the footage somewhat and animated some of the depth of field to emphasize more of the visuals; this series of work made sure that I only had to render it once, which is a very time-consuming task.
After rendering, I organized all the resources created and imported them. Next week, I will consolidate all the help, finish my animation, and make a series of changes to the sound effects, as the video I gave to the dubbing studio and the final animation is somewhat different. I need to make a last change to the sound effects to ensure that each corresponds to the animation, which will also take a lot of time.
I had a very tiring week last week. Due to my lack of rigor, the model had a problem, which almost led to an irrefutable situation. I spent a whole week remaking the character model, troubleshooting the problem, and finally got a satisfactory result. The result of. This week, I will continue to focus on the settings of the scene file, including the modification of some textures in the scene and the location of the lens depth of field in the background.
Here’s some ball animation：
After spending some extra time modifying the textures throughout the model, I started animating the little octopus rolling into a glowing ball. I used a sphere in Unreal and animated the sphere in Unreal to correspond to the character’s dynamics. I used techniques I learned from my instructor in the first semester, using elements such as scaling and extrusion to make the entire sphere look very dynamic.
Moreover, while adjusting the mermaid model’s texture, I repainted the new textures of other models. Although there are not many, this can make this treasure chest more integrated into the picture:
The idea of this modification is also straightforward. My animation itself is stylized. If the things in the scene are too real, it will feel strange, so I repainted each texture and added the brush mechanism. , and finally outline it, it won’t look real.
Reflections and Challenges:
When I first started using a camera, I was new to operating a camera, and my understanding of depth of field was the ability to blur distant objects, adding realism to the concept of near reality and distance. I encountered many challenges during the production process. Sometimes, the depth of field would make the picture very unnatural, and sometimes, the focus would be inaccurate, causing the highlighted parts to be unrealistic. Each time, I needed to gain a deeper understanding of these. I passed countless tests. I finally mastered some functions more suitable for my scene and learned how to change the picture by changing the aperture size and exposure parameters to achieve the desired effect. This process was a valuable experience for me, and through continuous study and challenge again and again, I was able to improve my skills, make my images look better, and also give me a better understanding of this function…
This week has been a tough week. I constantly learn, absorb much knowledge, and grow quickly and hard. Most of my animation is done. Now, I need to make the final changes to the footage and have it ready to render immediately!
Next week, I will review all the shots again, fix minor errors, and finally render the finished film. Then, I will combine the finished movie with sound effects to complete my task. While I could make the sound effects myself, health and time issues prevented me from allocating more energy, so I chose to commission the sound effects.
Last week, I discovered an error in the model that was difficult to solve. This week, I spent two days modifying the UVs of the character model again and redrawing a new texture. At the same time, I imported the modified animation into Unreal. Then, I made new animations according to the needs of each shot. If I have free time, I will modify the camera’s depth of field.
This is my modified UV. I put all the seams on the back and hid them under the hair so that the camera couldn’t see the seams. The next step is to repeat the previous operation. I put the model into Substance Painter and re-baked it, then drew patterns and textures on the new UVs, and finally modified some seams and details through Photoshop. In this way, ugly problems such as seams can be avoided to the greatest extent.
After modification, you can see that the stitching is almost invisible, although there is still a seam in a dim environment. Because it is behind the back, the camera cannot capture it.
These are the finished modified textures. I changed the eyeball textures along with them, as the character’s previous eyeballs were overly complex and didn’t look quite suitable for my animation style.
Reflection：
In 3D character modeling, baking is essential for normal textures. I made a big mistake in painting with the SUBSTANCE painter, resulting in a lot of time needed to make up for the consequences. Re-doing the UVs is a redundant workload for me, and despite the small amount of time, it still requires a lot of effort. I should have checked at every step and made the necessary adjustments. Otherwise, I would be in this situation and have to take responsibility for my mistakes.
This failure was a painful lesson in the need to be very careful at every step of the production process. I will work harder to learn more about baking textures in the future and be consistent with the process to learn a more standardized and efficient way of production. I realized that even the most minor details should be noticed during this production, and I viewed this failure as a lesson learned.
I spent much time this week finally solving my model texturing issues. Next week, I will continue working on my scene placement and shot setup.
Last week, I finished adjusting most of the lenses. This week, I will improve the animation effects of each lens and modify the required particle effects resources according to the needs of the lens. Last week, I finished adjusting most of the lenses. This week, I will improve the animation effects of each lens and modify the required particle effects resources according to the needs of the lens.
This week, I placed each special effect in a different position to decorate each shot, which can make the seabed more colorful. Still, I don’t want to make too many moving animations in the scene because this will attract the audience’s attention, and every scene will be different. The focus of a scene should be the characters.
I’ve picked a few screenshots of shots that worked well and put them up:
Reflection:
Then, while observing the animation, I discovered a fatal problem. Because the ears of the character model are translucent, you can see the inside of the model. The inside of the model overlaps with the glowing hair, emitting white light, so I had to change the character model.
When observing the model, I discovered an even more troublesome problem – the regular seams of the character were pronounced. This was obviously a problem that occurred when I was making the texture! But because the preview effect in Substance Painter was almost invisible, I only discovered how noticeable the seams were after switching to the normal channel.
After that, I spent many days studying this problem and tried many ways to improve it. Because the animation had already been produced, I couldn’t modify the rigged model. I could only choose to alter the UVs and cut the tangent lines as much as possible. Hide it behind your back so you can’t see it in the camera.
But this problem caused me huge trouble. Re-cutting the UVs meant that I had to re-draw the texture. This situation is still caused by the fact that I am not proficient in model baking, and I once again realized the lack of my skills.
In addition, I am still modifying the animation. The current animation modifications are more detailed, including how to move the fingers, how to move the eyes, how to move the eyebrows, etc., so it only requires a little time.
Therefore, my work next week is to modify the textures, including the texture of some models in the character’s scene that do not conform to the art style, and adjust the lens depth of field. This week, the teacher still gave me some details that need to be changed. Although this animation is imperfect, it is the most flexible animation I can make. I can achieve this level in a limited time. I’m pretty satisfied.
Last week, I had the VFX assets all ready to go in UE, set up the shots in general, and posed the most essential scenes. This week, I will flesh out the details of each scene, import all the animation files into UE, and modify the details of the animations again.
These are some arranged shots and unadjusted animations. The process of importing Maya into Unreal is quite troublesome because the calculation method of each software is different, and Maya animation has another driving method in Unreal. I am also experimenting. I tried it many times and searched a lot of information and tutorials to ensure everything was fine. I summarized the principle of importing. Basically, we first need a “puppet” of a character. This puppet is used as a body, and then the “animation that makes the puppet move” is imported to drive the activity of this model, so the model file only needs 1, but the animation files require 33. I created 33 folders so that it is more apparent.
This is the effect after enriching the shot:
This is the animation sequence file I created:
This is the imported character model. I created multiple materials. To make the fins translucent, I made a separate material. So far, the effect looks good.
Challenge & Reflection:
The challenge this time is the replacement of software. The import of Maya and Unreal seems simple. Still, the steps are very complicated because Maya’s coordinate axis calculation differs from Unreal’s. After all models and animations are imported into Unreal, the world coordinates The axis will be cleared, so sometimes I have to manually move the animation to the position it should be. This is because I needed to fully understand this process when making the animation, and this may affect the new animation in the future. Animation replacement has an impact, but most of the animation production has been completed, and it is best not to move the world axis now.
Modification of animation:
Last week, the teacher made many specific suggestions for animation modifications, which is still the case this week. Most of the advice this week is straightforward and small, including hand movements, the speed of the character twisting the body, the rate of the hair, unnecessary key poses in some shots, etc. These small things can be discovered with the naked eye. Dynamics are more prosperous and more flexible. At the same time, through research papers on body language, I found some helpful animation methods. I tried to integrate these methods into my animations and saw significant results.
This week’s work looks relatively simple, but it took a long time, so I’m doing the equivalent of consolidating the conversion process. Next week, I’ll be further tweaking the animation details and UE FX shots, adding animation and FX resources to each shot.
Last week, I set up the introductory scene. I made the conceptual look with moss this week. I’m going to finish up the animation modifications and set up the camera in the unreal scene based on the animation footage, and then change different values for the effects through the footage settings to make the effects fit the scene better.
Challenge:
The lens setting is a new test for me. I must ensure that my picture is clear but looks good. This has something to do with the composition of the lens. Because all my underwater materials are luminous, This feature is eye-catching. I had to constantly adjust the values to ensure they weren’t too bright. Here are the shots after posing:
Special effects: The special effects in my scene are mainly particles. Some of the special effects resources were organized and reset by my classmate Kaiyue Wang and I. I arranged different vegetation in redesigned positions according to the lens frame. Because I need each shot to have a primary and secondary relationship, and I need to ensure I have the right amount of glowing plants so it doesn’t look too cluttered.
At the same time, we made a particle special effects material, which can be regarded as underwater plankton. I studied this material for a long time with limited knowledge and finally found a way to control the material’s density, area, and size. This made me very happy. A sense of accomplishment.
Setting up the camera is a process of repeating simple operations. Although it is not difficult, it is a lot of work, so it still took me a lot of time.
The next step is to improve the animation. I improved last week’s animation based on the teacher’s comments. This is the improved effect:
The feedback this time pointed out that there are still many areas for improvement in the animation, including the length of time, the movement of the hair, and some of the characters’ body language. Through the teacher’s suggestions, I changed some actions to make the plot more coherent, and no Unnecessary scenes will be deleted.
Reflection:
This animation modification made me realize a lot of shortcomings. I may still need to observe the character dynamics carefully enough, including some of the most essential things, and I need to understand them more. At the same time, because the topic of my graduation thesis is body language, I found a lot of helpful knowledge while searching for information, and I plan to use it in animation. In addition, this animation scene also gave me a lot of challenges because it is an underwater scene. The character’s dynamics are in a state that is greatly affected by buoyancy. Hence, the hair movements and other things that only occur in the water Inertia are completely different from ordinary land creatures. It is an entirely new production method for me.
Next week, I will import most of the animation files into UE, adjust the character’s position, re-adjust the lens according to the character, and then modify the animation.
Last week, I finished the models and painted the textures. This week, I will import all of the models into a built-up unreal scene and redesign the scene using the assets I have already finished making.
Although the process is straightforward, it is very troublesome, repeating simple actions repeatedly. I aim to create 33 shots in a scene, which sounds difficult. The scene is vast, so it takes me a lot of time. Moss is very suitable for outlining the shape of stones, so I adjusted the value of moss and planted moss on the ground according to the lens because there should not be too many models in the scene. The computer might not be able to support it. The production of these assets was discussed and completed by Kaiyue Wang and me. I am very grateful to her for her support. Because of her support, my work speed is much faster. She taught me a lot of unreal usage.
To save time, I searched for free treasure chest resource models online. Still, because this model was too actual, I would change the texture to a model that suited my animation style when I adjusted the assets again. I made the models of these stones myself last semester. I changed the arrangement and turned the scene into a completely different look. Sunlight and many kinds of light sources have been added to ensure that the seabed will not be cloudy and things can be seen clearly.
I think the effect of the glowing moss is perfect, and you can clearly see that the scene is more decadent. I learned how to use the plant module for scene editing. This practice made me more proficient in plant plants in unreal scenes.
This week, my scene has a more detailed structure and a more straightforward layout. I am delighted with it and have carefully thought about where each camera should be placed. Next week, I will develop my camera further. Placement and placement of various aquatic creatures
Of course, to make animations, I exported the arranged scene model. I put it into Maya to realize the interaction between Maya animation and unreal scenes.
Challenge & Reflection：
There are many challenges this week. My classmate kaiyue wang gave me a lot of help. She taught me to use more and more complex unreal editing. In the production of plant modules, I often needed help. Because I was not proficient in using the software, I could not control the number of plants well, which led to unreal crashes. Ultimately, I tried about five or six times before finding the right one. Plant in quantity. Moreover, I can use various functions more skillfully through constant and repeated practice while using Unreal. A very troublesome thing happens when the model is imported from Unreal into Maya. Because the unit calculations of Unreal and Maya are different, when I import the scene model into Maya, the model is 100 times larger than before. Of course, this is also what I have studied. The result after about an hour… This made me realize that I still know too little about unreal. There may be an easier way to solve this problem.
Animation：
At the same time, I am constantly improving my animation production. I record the animation through the playback function of Maya and then use UE to synthesize it into a video for the teacher to check:
before：
after：
Before the summer vacation, I made a version of the animation through the most basic storyboard, but that animation had many problems. I submitted the animation to the teacher. I had extensively revised it this time.
From our teacher’s feedback, most of the problems are that the characters’ body movements need to move in an arc, so they look very stiff. There is also a very troublesome place – the hair. I needed to make the look of my hair floating in the water entirely manually. The teacher gave me a lot of specific suggestions for this, including how the hair should flow at this time, at what speed, in which direction, etc. These are very useful.
Next week, I will complete the placement of the camera in unreal and the scenery required by the camera and use Photoshop to change some textures into another art style.
