We know that anime shifted from an analogue production process to a digital one.
I can’t find a very simple, understandable account of the shift, of the sort which one could gently link when a friend refers to anime animators ‘no longer drawing by hand’. Upon questioning, many fans who attend to production info, most of them more knowledgeable than I, don’t seem to know much about the nature and timing of the shift either. So I set out, in foolish ignorance, to write such an account.
I’ve tried to keep what follows clear and brief. I’ve also tried to exclude my views on the relative strengths of different times and techniques. I’m no expert on the topic, and this post is a call for facts, additions and corrections as least as much as it’s as a summation of them.
Before the shift:
- animators drew line drawings on paper
- the line drawings were photocopied onto cels
- colourists painted the cels with physical paint
- the cels were photographed together against a background, using a physical camera, onto physical film
After the shift:
- animators draw line drawings on paper or directly onto a tablet; in both cases, they still use their hands
- (any paper lineart is scanned to create digital art)
- colourists paint the art with digital paint
- different layers are combined together against a background—effectively, photographed—on a computer
The new process very closely resembles the old process, and indeed I’m told a lot of the software used in the new process takes the old process as a governing metaphor. Both are about crafting layers, then making those layers interact during photography.
This means 2D animation in anime is still hand-drawn, forays into automatic interpolation of in-betweens aside. Fans sometimes discuss the shift as though it meant abandoning hand-drawn work, but in fact it only meant abandoning hand-painting.
For some of the shift’s technical and aesthetic implications for animators’ experiments from roughly 2000 on, see this very handy discussion.
I haven’t mentioned ‘lighting’ or other effects in either bullet-pointed list. As far as I understand it, in both pipelines these sit, at least for a simple account’s sake, within the photography/composition part. If you want to know (much) more than I can offer about photography in the cel era, go read this, and see also this note.
* * *
When did the shift happen?
During this post’s initial reception, no one quibbled with the previous section, but generous people offered helpful suggestions on the problem of dating the shift. Here, note, we leave straightforward things to travel into the land of induction and report. This part of the post still changes as I get new info; consider it work permanently in progress. Fortunately, this part is also the less-vital part.
We conventionally date the shift to ‘around 2000’, with some entirely digitally-coloured titles appearing in 1999 (e.g. the entertaining workplace satire Dai-Guard) and some cel-based titles poking into the new millennium (e.g. Mazinkaiser, 2001, and, I believe, the 2003 Astro Boy). Sazae-san, the final anime to go born-digital, reportedly started to shift in 2005, and aired the last anime episode using any cels at all on 29 September 2013.
Kraker suggested from memory that most of the industry switched more-or-less in concert, in a fairly narrow time-frame. This sounds likely given (i) how many relatively small companies there were, (ii) how interwoven their dealings often seem to have been, and (iii) how collaborative they can sometimes be: I recall seeing Aya Suzuki say in an interview that multiple companies shared the same transport when sending materials back and forth to sub-contracted studios in other countries.
We can complicate the simple date of ‘around 2000’, however. First, some productions jumped early. On Twitter, ehoba kindly reported that the book アニメ制作者たちの方法: 21世紀のアニメ表現論入門 (2019, ɪsʙɴ 9784845918089) points to the sixty-fourth episode of the fourth GeGeGe no Kitarō adaptation as either a or the moment of inception for digital colouring. I’ve looked up episode 64 and compared it to what came before and after in the show; it certainly does have digital colouring, and seems to be the point at which the show as a whole shifted to digital colouring. It aired on 6 April 1997.
The February 2019 issue of Animage apparently had an interview with Tatsuya Nagamine in which he reportedly that Toei Animation as a whole, not just Kitarō, made the shift at this time. Fmod91 pointed out to me that the seventh episode of Toei’s Be-Bop Highschool OVA, released 11 December 1998, uses digital painting too, which fits this.
Second, the use of computers for aspects of the photography/compositing stage has a longer history. Some elements of digital photography probably played a role in anime production for the first time in the second episode of The Yearling / Kojika Monogatari (1983). This factoid comes from Clements’s Anime: A History, and Renato Rivera Rusca generously chipped in to back it up, and Luis Alis pointed me to this column which also mentions it. Reportedly, there was some interest, even this early, in using computers to automate in-betweening. But the tech required to do this in the early eighties cost an eye-watering amount, making the process a provocative demonstration rather than a viable route forward at the time.
(As an aside, I’m tickled by the foresight this project displayed. Like anachronistic oddities such as Project Cybersyn, it strained its own time’s machinery to breaking-point, but anticipated things which would one day come to pass.)
1983 is also the year that 3DCG was put to work for a few scenes in Golgo 13—by the same tech outfit, in Clements’s account. 3DCG would see further use in the Lensman film the following year. 3DCG work is, obviously, not 2D animation, but in the making of those movies the 3DCG content must have gone from digital files to physical film for distribution somehow.
Moreover, by the late the 1990s staff could add digital effects to anime otherwise made using cels and analogue film. As with Golgo 13 and Lensman, this hybrid process of physical painting and digital compositing presumably still resulted in a final physical film output. (If there wasn’t an analogue infinite-resolution film output, then 1990s anime would’ve started out, and been stuck, at low digital resolutions.) Xaryen has very reasonably suggested to me that this was achieved by printing digital images onto film, and then using analogue techniques—I imagine either an optical printer (expensive) or double exposure— to combine them with other elements.
These hybrid processes laid the groundwork for the shift away from all physical tools. They didn’t in and of themselves constitute that shift, but they shouldn’t be forgotten!
* * *
So much for what I (think I) know: really, this captures what an interested amateur can rustle up using the English-language internet in not too much time. Further corrections or additions would delight me, especially if sourced, and I will happily edit any such useful suggestions into the post.