Category: ADVANCED AND EXPERIMENTAL 3D COMPUTER ANIMATION TECHNIQUES PROJECT1

The Deep Mist (A Cthulhu Style CG Trailer)

Project1 Showreel

My role in the team:

Story

I provided the very first idea, wanting to do a Cthulhu-style film, in which there could be a monster of huge size, human size is very small compared to it, with the feeling of gigantophobia. Then Guanze Wu enriched the whole story line and provided more additions to the story. Finally Guanze, Yaozhang and Ziying and I all worked together to finalize the entire story and details.

Storyboard

We made our own version of the storyboard based on the work we had agreed to do and then put together a collage of everyone’s storyboards to form a complete version.

 Layout

I created a dynamic storyboard for the project. The purpose of doing this was to select the BGM in advance, and in addition I could let the collaborators know exactly how much time each shot would take.

Character

I made two more characters as I had agreed at the beginning, which included the initial sculpting to the topology, then to the texturing of the two characters, and finally I made the materials for the two characters.

Comkpositting

I did all the compositing parts that I needed footage for in AE, which included sub-channel adjustments and adding effects, as well as color grading.

AE SUBTITLE EFFECT VIDEO

I created all the titles for this project, including the project title at the beginning, several sub-titles in the middle transitions and the closing titles at the end of the film. I did all of these title effects in AE.

Editing

I ended up making my editing section, doing all the transitions effects, and adding more ambient sound effects, which made the whole project more complete.

Hyberlinks:

Week01: https://zhenzhongliang.myblog.arts.ac.uk/2022/06/06/week1-storyboard-reference/

Week02: https://zhenzhongliang.myblog.arts.ac.uk/2022/06/06/week2-layout-confirm-work-flow/

Week03: https://zhenzhongliang.myblog.arts.ac.uk/2022/06/07/week3-investigators-character-modeling/

Week04: https://zhenzhongliang.myblog.arts.ac.uk/2022/06/07/week4-topology-uv-splitting-layout-uv/

Week05: https://zhenzhongliang.myblog.arts.ac.uk/2022/06/08/week5-male-and-female-head-modeling-mask-modeling/

Week06: https://zhenzhongliang.myblog.arts.ac.uk/2022/06/08/week6texturing-in-susstance-painter-baking-maps/

Week07: https://zhenzhongliang.myblog.arts.ac.uk/2022/06/12/week7-character-shading-rendering-ae-subtitle-effect-video/

Week08: https://zhenzhongliang.myblog.arts.ac.uk/2022/06/19/week8-compositting-footage/

Week09: https://zhenzhongliang.myblog.arts.ac.uk/2022/06/22/wee8-editing/

This week I already collect all the footages and compositted them, so I just to need to put them together and edit them. Then I will try to add more sound effects.

Thanks to my previous project files in layout production, I only had to replace these shots separately in the first step. In addition, I specifically mentioned to my collaborators that it would be best to add an extra second to each shot before and after the time indicated on my layouts, so that we all had more time to edit. The advantage of this is that when I need to do a lot of transition effects for example, sometimes the time before and after is not long enough to cause the transition effects are not fully displayed, if there is an extra second this problem will be largely avoided.

Next I did more transition effects on it, for example, between the very first title and the first scene shot, I wanted to make the transition between them more natural, so I noticed that the title shot had some smoke effects. I then found a footage of the smoke rising and dispersing and placed them in the middle of the two shots, then adjusted the overlay to multiply and then adjusted the transparency. This gave me a nice transition effect at the beginning, which made it look more natural.

Besides, most of the transitions were made up of simple zoom in and zoom out and other basic camera movements with some camera shake or shutter angle and blur effects, because I didn’t want to make the whole film look too complicated.

Finally, I added some ambient sounds, such as the ticking sound of the hands turning faster and faster, and in the last shot when the character’s flashlight light shines on the monster, the faint sound of breathing on the monster. And with the monster’s eyes open, this breathing sound becomes more and more obvious, these small details will make the whole film more rich.

I’ve already done the character’s modeling work, and also make the texture and shader for them in Maya Arnold. So this week my collaborators will give me the footage that I could continuely do the compositting work.

Here I would use one of the video compositions to illustrate my method, as the rest of the footage uses this same workflow.

So I drap all the footage into After Effects, and create one composition. Cause I’ve already told Guanze, in compositting work which kind of path I need for example ID, emissive if have, Z depth etc. So I put one original footage at the bottom, then duplicate it and give it a extractor.

I could extract different paths I need with this step. For example I extract the emissive path, in this scene the light from the flashlight is the only source of emissive. So I wanna its color become more cold this can have a strong contrast with the eyes of the monster behind the crimson color. I edit the curves and increased Blue content in RGB. The same method I uesd to the human. Cause the light is pretty dark in a cave, so audience can’t see where the human is, I especialy increase the edge light of the figure by the method I told before to allow the audience to enhance the perception of this scene between the characters standing.

For the monster’s eye, in the very first time I extract the ID path, and use color key to extract eye’s position. After that I added another original layer under it and set it as ‘Alpha Inverted Matte’. So that I can adjust only for this area. I use a new plug-in which is ‘real gow’ it could bring a perfect glow effect. So I added the glow effect when the eyes are open, and key some animation to make its glow effect look flickering, which makes the whole picture become more vivid.

Finally I add 2 more adjust layer, one for edit the whole gamma some of our footage’s gamma isn’t well, another one for edit color and filters. Editing gamma is simple, the gamma value that best suits your picture is the best. There is no trick to it, it is entirely based on your picture. For the editting color I use a plug-in ‘Looks’ it would be very fast to adjust a lot of desired effects, even the focal length and so on can be adjusted very convenient. In this project, since we are a Cthulhu project, I want the frames to be in a blue-purple tone, so I spend a lot of time to adjust it.

Finally output the whole footage the compositting work is done. I basically used this method to composite all the shots I needed, and it wasn’t as much trouble as I thought it would be, because as soon as the first one was done, the later ones just needed a few simple adjustments.

This week I would make the shader for the 2 characters and set rendering, after that I would also use AE make title video and subtitle videos for the trailer.

As I seted before texturing the material, I just need to put the texture from Substance Painter into Hypershade window. Then link different textures to the right position.

It is worth to mention that some nodes in the Arnold material require manual adjustment of parameters. Rounghness, metalic, opcity(if have) and displaement textrures require to change the color space into Raw and turn on the Alpha is luminance this could make the rendering corectly. Besides if you use normal map make sure add one bump2D node and link normal map’s alpha to bump value in the node then out normal to the normal camara. Also change the color space into Raw.

After linking creat a skydome light and put a hdr texture in it. This step is just make sure everything is right in the final rendering under a normal light set. Besides I also made some changes somewhere I feel unsatisfied. For example I thought the glass masks on the helmet is too clean so I add one extra noise map to get more dirty detail and mixed the base color and the noise map together and link in base color. Therefore I got the final character rendering version.

I really like the title design in DeathStranding, especial the hair strand effect. Basicly this is mainly achieved through effect expressions. Just need 2 expressions of vector mash to control the random seed and size finally also need to add another expressions to control the size of the effect.

Next week I probable foucused on other subtitle video and wait for collaborator’s rendering video to edit, this project is coming to the ending~

After modeling and unfolding UV this week I could do the shading and texuring work.

First I need to figure out which part’s texture and material is same so that I could give them one shader and name it. This is important in Substance Painter and Marmost Toolbag, this allows me to make and bake the mapping separately depending on the material.

The next I put the sculpting model which has nearly 1 million faces together and rename for each part for example XXX_low and XXX_high. The reason I did this I need to put them together into Marmoset Toolbag to bake Normal map and Curve map and AO map. The software could devide them by the name XXX and put them in low model and high model by name.

After that I got basic Normal and Curve and AO maps. Then swich to Substance Painter, upload the low model into it by the way make sure your model is in soft edges otherwise the height(displacement) map and Normal map will have some wireframe on them. Choose the map I baked before and the software will automatic assignment to the right place.

My personl habit for texture in Substance painter is that making a fill layer in the first time to control the base color and roughness and metalic if necessary. Then I probably add several dirt layers to imitate the dust or some edge marks of wear. And gave all of them with a folder and name it and creat a black mask to choose which part I would wanna have this shader.

Finally export all textures and chose Arnold default maps (in this point you could also add other texture as you want and need) and set a name space and file path.

In our script there are two roles present a male investigator and a female investigator. Now that I have a basic investigator’s clothing, I still just need to get the heads of the two characters done to match the investigator clothing.

In order to save efficiency I decided to use DAZ 3D to generate two characters out, which can greatly save production time.

So I could get two basic full body model, and then export them into FBX file with textures and rig information by the way but I won’t use the rig cause I only need the head part.

Then import them into MAYA and delete rig infromation. Finally delete other part of body just leave head and hair. I finally get the head model.

Next is to make the mask wear on these two character’s face. And I searched two images that close to my thought about the mask. It could have some sci features and there are two air outlets on the left and right side.

So I modeled them in Maya and move it to match the head on correct position.

After last week’s scuplting work I export the model to MAYA into FBX file but I don’t know why all the faces change into triangular face. And I tried OBJ file as well, the same happened. So I quited and decided to export the high model which had nearly a million faces and topo it in MAYA.

The method is simple, first just need to converts the selected surface to a live surface.

Then move to modeling toolkit and choose Quad Draw in the tools, and set points in the surface which you already select as live surface. Besides the point you will set in the face it’s better to use 4 sides face cuase it would have the best animation effect for your work.

After 3 days boring work of topology just to be honest I finally had a 90,644 faces model which keep the original detail I scuplt and could be rigged. The next I need to split and unfold UV to give Guanze Wu to rig.

There is no strict principle of splitting UV, just remember not to leave too many seams in the places where the camera is often easy to shoot, people usually cut in somewhere like underarm, crotch, chin even the back part of ear. This could avoid the audience to see these easy to have seams, otherwise there will be a seam processing work to do.

Finally layout the UV, my habit is to zoom in as much as possible on the larger areas of the UV shell, this could have more details on the possitive space.

As our script we need 2 SCP foundation investigators, so I searched several references which is closely to the investigators.

Firstly I just thought the area which our story happens would full of toxic gases even nuclear radiation phenomenon so I quickly reminded some medical staff in protective clothing, they would wear face masks in work daily and have enough protection suit. So I found some images relate to this in Art Station.

But our teammates Guanze Wu added that our characters should have more than just protective clothing that fits real life, it could have some sci equipment to against the huge Cthulu monster. With this point I linked to astronauts equipment. These equipments is strong enough to defence the body and it makes things more reliable.

I mainly use MAYA and Zbrush to model character, so in the very first time I quickly set several basic box to confirm the size of character.

Cause I wanna make a helmet to defence toxic gases so the head looks like a little big. So I move to Zbrush to sculpt the model.

I used ZSphere to get a basic body model and set differet sphere in right place and then I just need to work on sculpting and following the reference especialy sculpt some pleating effect of clothes.

Finally I got this basic model but the subdivision was too high, so next week I will focus on topo this model until 10,000 faces max.

After last week’s work it’s important to make a layout which could make our group members clearly understand the time of each shot and the overall rhythm of the film. So I need to use some images from our storyboard and references to imitate the final frames to edit.

First I found some free background music wich full of horror and gloomy as the basic music. And according to this I will follow the rhythm of music to edit different shots to make sure how many time we need for different shots.

I did it in Adobe Premiere and put all of our images and background music in it. And marked the climax of the music as the point shift different shots.

Finally others could just follow the layout’s rhythm to confirm how many time they need. So layout is essential for project.

Besides we also confirmed the work flow that for some far view scene and mid view scene we would use Unreal Engine to do it, cause the PBR effect could be rendered very efficiently with it. On the ohter hand for some close up shot we would still use Maya and Arnold to make. If we use Unreal Engine for them it’s too close the effect will be easily seen fake.

As last term’s presentation me and Guanze Wu and Ziyin Wang and Yaozhang Zhong decided to make a Cthulhu style project. So in the first week we had a online meeting to derive the whole script. In this meeting we basically reached a consensus that we could make a horror film trailer and it’s about 90 seconds.

I created a miro link to collaborators we could share our work and work progress on there. https://miro.com/app/board/uXjVO8cHs_Y=/?share_link_id=850618947548

After that I firstly did some researches on characters and the environment on artstation to gather some ideas. For the enviroment I want to make some Nordic topography maybe in a huge forest and the huge monster hide in the deep of some places like caves. Besides I also had some thoughts on the characters, as our script it would have some professional equipments like chemical protective clothing and helmet and mask etc.

Also I did a sotryboard to indicate which camera and shots we need. And I used a combination of drawing and photoshoping to create my storyboard, which could demonstrate the shot and the atmosphere that I want much clearer.