(now it gets really interesting ...)
The output format has no bearing on the number of frames the animator chooses to draw for every one second of footage.
I disagree. Although it is possible to tell a computer to play back a video file at virtually any frame rate (up to about 60 fps), this doesn't mean that it's a good idea to produce in just a randomly chosen frame rate. Why?
1. Film and TV still have a fixed playback frame rate. 24, 25 and 30. That's it. Add to it the quasi-standard of youtube's 15 fps, and you're done.
2. Although there are tools to convert frame rates, animation content suffers extremely when undergoing this surgery. Everybody knows how an animated movement can be destroyed when frames get entirely dropped during conversion - it might happen that of all things just the keys get deleted!
Converting the other way round (less to more fps) does something different, it adds frames which get composed of existing ones, with a strange double exposure look.
3. Since classy animation extremely depends on the clear definition of every drawing, those artifacts are not acceptable. It would be liketor andom lyspill blank sallo veryo rtext - the mean ingis stillt here,b utwhob others toread itany more? *g* It destroys the "illusion of life", at least spoils it significantly.
4. 3D CGI stuff is a bit easier in this respect because (within the predominant style) visually it is more like life action footage. Animation has sharply defined corners, outlines and fills, which make every artifact stand out prominently. The technology simply does not care for animation. All its filters and codecs are designed for life action reference footage. Procedures which work pretty well for life action (and most 3D CGI) wreak havoc with drawn animation, as everybody with a DVD player can experience: digital noise reduction swallows outlines, pulldown procedures add or subtract frames and destroy movement, and so on.
(4a. As long as you're still within your animation software, it is possible to avoid some of these. Because you're able to re-calculate playback frame rate (in AS and elsewhere), it is possible to render for different fps without losing keyframes. Once the material is a rendered video file, you can't avoid the described effects.)
5. Film and TV producers don't accept material which isn't "broadcast quality". In animation this means, a frame of film or video must hold exactly one frame of animation - no double exposure, no dropped keys. Changing between 24 and 25 fps is no problem there, conversion between 30 and 25 is, but still is accepted - otherwise there would be no american animation shown in european TV, and vice versa.
it simply means to animate at 50% of the total frames per second of the intended output format
It's not that easy. Sure, on the first level it means just that: reduce the amount of drawings. But "animate on twos" means more: it means to be able to switch to
ones at any moment, and back again to
twos and whatever.
6. The playback frame rate (fps) only defines the smallest possible increment in time. (Since I come from film, it is 1/24 of a second for me.) An empty X-sheet gives me a raster of 24 frames per second - and within this raster I am completely free to plan my animation. I could do a part
on ones because it is a fast and important movement which needs some impact. I could do other parts
on twos, mainly dialogue, because it is movement at normal speed and doesn't need further definition. I could even do some parts
on fours if the movement is slower than normal, or put an entire character or parts of it on
hold.
It is very much like a musical composition with its use of notes of different length. It is a rhythm on the smallest possible scale, the single frame.
Funny enough the few attempts to create drawn animation entirely
on ones lack a certain dynamic. My favourite example would be "The Thief and the Cobbler" (see
http://thethief1.blogspot.com/ for some more information) which was done on ones and, IMO, suffers from over-animation a lot - it's just too smooth to be interesting.
7. Drawn animation isn't life action. It is an abstraction of real movement as much as it is a visual abstraction of the real world. As well as the artists decides to leave out separate hairs in favour of just an outline filled with colour to represent a hairdo, as well he decides to leave out several levels of movement in favour of the few he choses to express his idea.
This goes as far as a 2D animator is able to create absolutely impossible movements, up to the level of not really showing any movement at all, but just a sequence of frames of wild strokes which in projection deliver an idea of motion and emotion without actually creating any illusion of movement at all.
8. Since any decision about how many drawings and where to place them in time is an artistic one (or should be), your statement
Some of these concepts, such as "animating on the twos", are outdated and quickly becoming obsolete as we've been moving away from film to an entirely digital system
doesn't hold any truth. Au contraire: any software which has less options than my previous way of work is a software which restricts me - and I will not use it. Fortunately AS doesn't do that: I could work on twos if I like to, and I know the trick how to switch to ones eventually, although it is not really comfortable. In worst case I always could re-arrange rendered images in another application to create any sophisticated timing.
Of course it is possible that you will never run in any of these problems in your whole professional life because you work in 3D CGI. My position is different: not only I come from film, I also have a background in experimental and underground film which is quite different from your average animated TV series. And I worked in a professional studio right then when they made the transition from painted cels and rostrum camera to digitally coloured and composited animation.