**Blending 3D and 2D Animation — Transcript & Summary | SozAI**
Source: https://sozai.app/transcript/blending-3d-2d-animation/

Olof Storm explains his hybrid workflow: blocking out animation with simple 3D characters, then hand-drawing 2D frames over them for a bouldering montage in his film.

## Key Takeaways

- Simple 3D blockouts can speed up 2D animation by providing timing, camera, and shape reference without final-quality modeling.
- Deliberately incomplete 3D (no fingers, no facial animation) helps preserve the artist's hand in the final drawings.
- Handheld camera motion combined with sparse character animation is an economical way to add life to a shot.
- Hand-drawing offers more control and a more human look than 3D shading tricks, at the cost of time.

## What the video covers

- Storm is animating a time-lapse sequence of two climbing characters working on a boulder project across multiple sessions and weather conditions.
- A follow-up montage samples various climbing moves in a rapid, non-chronological edit cut to music.
- He blocks out shots using simple 3D characters, keyframed per joint without fancy rigs, framed in a 3D camera environment that mimics live-action filmmaking.
- Handheld camera motion adds dynamism and lets him animate characters on fewer drawn frames, using camera movement to carry motion.
- He renders a wider static view with only the character animated, then imports it into 2D software as a starting point for hand-drawn final frames.
- The 3D model supplies general shapes and head angles, while mitten-like hands and unanimated faces force hand-drawing and manual cleanup.
- He economizes on drawings by holding frames where motion is minimal, mirroring traditional 2D animation practice.
- After Effects is used to preview final camera motion before all frames are drawn.
- His style uses no outlines, only interior lines for overlaps and facial details; the 3D lineart serves as reference.
- He prefers this drawn approach over cel-shading filters, which he feels still fall into the 3D uncanny valley.

## Chapters

1. 00:00 Introduction to the climbing time-lapse sequence
2. 00:40 Planning the bouldering montage
3. 01:17 Why use 3D characters for blockout
4. 02:32 Handheld camera and economical keyframing
5. 02:58 Building wide static views for 2D import
6. 03:59 Drawing over 3D: mittens, faces, and cleanup
7. 05:30 Holding frames and previewing motion in After Effects
8. 05:47 Line style and comparison with cel shading
9. 07:10 Recycling and adjusting backgrounds
10. 08:00 Closing thoughts and Patreon mention

Answers

## Questions about this video

Why does Storm use 3D characters instead of drawing a rough animatic in 2D?

He finds 3D blockouts efficient: the characters are posable, can be framed in a 3D camera environment that mimics filmmaking, and provide shapes and angles he can trace or refine. This lets him assess edits before committing to full hand-drawn animation.

How does he avoid the look of 3D in the final film?

He leaves hands as fingerless mittens and does not animate faces in 3D, forcing hand-drawing and manual detailing. He also uses no outlines in the final style, treating the 3D line work only as reference rather than a finished look.

Why not use cel-shading or filters to get a 2D look?

Storm feels those methods still lean toward the 3D uncanny valley. He prefers the slower drawing process because it gives him full control over where to rely on the 3D reference and where to override it with his own decisions.

## Full Transcript — Download SRT & Markdown

00:05

Speaker A

[Music] I'm currently working on an animated sequence that follows a shot I created quite a while ago.

00:24

Speaker A

This animated time lapse of my two characters working on their climbing project was a shot that I had in mind very early in the development of my film. It's there to demonstrate time passing and that they spend multiple

00:40

Speaker A

sessions on it coming and going in various weather conditions. I then want to follow that with more specific focus on the climbing moves. a sequence of shots, not chronologically edited, but just sampling the various moves in a rapid edit that gives us a more detailed

01:00

Speaker A

view of the process. Edited to music, it becomes a montage of bouldering. There are a few ways I could block out a sequence. I could 2D animate a rough animatic with loose gestures of my characters, which I do for some

01:17

Speaker A

shots. or what I've decided to do for a few of these, which has been really helpful and efficient, is to do the first pass of animation using 3D characters. I created these with the aim of having them look

01:32

Speaker A

as close to my drawn versions as possible. They are reasonably simple in terms of geometry and textures, but they allow me to pose them and then frame them in a 3D environment, which mimics real film making quite a lot. I have no

01:52

Speaker A

fancy rigs for these, but I just brute force animate them with key frames for each joint. I also only have to make it look okay from the view that my camera is looking at. So, whatever is out of

02:06

Speaker A

frame does not have to perfectly make sense. I want a handheld camera for most of the climbing in this film. This adds dynamics and a second source of motion to it all, which I think looks good in an edit. This can

02:25

Speaker A

also save me from having to animate my character on once in 24 fps, but use the camera motion for that and then have the character alternate between fluid animation as well as some pretty low amounts of frames for certain bits. The

02:42

Speaker A

way I achieve that is to first create a shot that I'm happy with in 3D that pretty accurately describes what I'm after. This goes for both the 3D character animation as well as for the camera. Then I create a second static

02:58

Speaker A

camera that has a wider view than that with animation applied to it. This view is sort of a average of the more zoomed in one so that it covers the full viewing frame from start to finish of that one that is moving. I can then

03:15

Speaker A

export that wide static view where only the character is animated and bring that into my 2D animation software. Here I now have a perfect starting point to do my final drawings on top of this. I can then later apply

03:32

Speaker A

the camera motion again to this wider view. What I like about this 3D character is that it's detailed enough that I get the general shapes for free when drawing.

03:46

Speaker A

The angle of the head is just there for me to grab, for example. But it's also simple enough that I have to do a bit of cleanup and detailing in the 2D animation process.

03:59

Speaker A

I like that the hands are basically mittens without any fingers because that forces me to draw them and not trace over 3D fingers which generally I think results in a better image. Also I do no animation to the face in the 3D software

04:17

Speaker A

so any facial expressions I will have to add later but the placement of the eyes and proportions of the head it just comes for free.

04:30

Speaker A

During this process, I can decide if I want to hold on some frames where the character maybe is not moving overly much. And I can approach it as I would if I would have 2D animated this from the

04:44

Speaker A

start, not animating every frame on the 24 frames pers timeline. I basically try to be a bit economical with my drawings here. At any time, I can open this project in After Effects and now mimic the camera motion I had on my first 3D

05:03

Speaker A

camera to this static view. That means we can preview the final motion of the shot without having all the frames drawn [Music] yet.

05:29

Speaker A

Converting the 3D character into 2D lines is just the first step I take in this process. In my film, I've gone for a style that does not use outlines on my characters. I only use lines within the colored shapes to describe overlap of

05:47

Speaker A

body parts or facial details and so on. This means this line work pause is just used as reference for my final look. But it really helps in the conversion process from 3D to 2D. There are other ways people take on

06:08

Speaker A

this challenge by fully relying on the 3D software using cell shading and filters that try to replicate the look of 2D.

06:18

Speaker A

But I generally finds that it still tends to lean more towards the 3D uncanny valley with those methods.

06:25

Speaker A

Whereas this slow but deliberate drawing approach gives me full control of where to rely on the 3D and use it as aid and where to overwrite it with my own drawing decisions. I think some of those shading methods have come really far in the last

06:42

Speaker A

few years. And maybe there are ways to really fool us into thinking it's fully drawn by now, but I just like the drawing aspect of it. So therefore, I go down this road. I guess I'm just getting old too.

07:01

Speaker A

[Music] Having created a bunch of shots all already set around this boulder, I've had to paint a lot of angles of the background. I could start fresh every time, but sometimes the angles match up relatively okay with the previous shot.

07:30

Speaker A

So then I can recycle a lot of that work and adjust an already painted background to fit this new one. To get through the endless list of shots on my own that this film requires, I will have to find

07:44

Speaker A

little shortcuts here and there to get anywhere. The 3D blockout method solves a lot of the film making aspect of this following sequence. I can pretty quickly create an edit of these that helps me see sections of the film without

08:00

Speaker A

committing to a lot of hours of drawing. But once I see it's working in the edit, I love the manual process of drawing it all. Hope you enjoy this video. I have a Patreon page with a lot more video

08:13

Speaker A

content and tutorials that you can check out. I'll see you in the next one. Bye.

08:29

Speaker A

[Music]

Topics: 2D animation 3D blockout hybrid animation animatic hand-drawn animation After Effects bouldering animation workflow Olof Storm independent film

Study this video

- [Flashcards from this video](https://sozai.app/tools/flashcard-maker/?from=transcript&id=91945)
- [Quiz on this video](https://sozai.app/tools/quiz-generator/?from=transcript&id=91945)
- [Crossword from this video](https://sozai.app/tools/crossword-maker/?from=transcript&id=91945)
- [Word search from this video](https://sozai.app/tools/word-search-maker/?from=transcript&id=91945)
- [Study guide from this video](https://sozai.app/tools/study-guide-maker/?from=transcript&id=91945)

Free, made by AI from this transcript. No signup.


---
This is the markdown twin of https://sozai.app/transcript/blending-3d-2d-animation/ — the same content, without the markup.
Published by SozAI (https://sozai.app). Reuse and quotation are allowed with attribution and a link back.
Machine-readable index: https://sozai.app/llms.txt · data API: https://sozai.app/api/
