hs anyone tried a "hybrid" technique using switch layers sometimes and "cut paper" PNG's @ other times as an animation technique?
I've heard folks warn me off of using AS as a drawn animation medium. I imagine it must get too memory heavy.
to elaborate, im drawing objects, characters, and backgrounds in Painter or Photoshop, then importing them into AS. To show geometric changes, i use switch layers, though at other times i intend to use bones to animate different "parts" of an object or character.
hope that im not biting off too much here. Im up to shot four of an approximately 5 minutes of filming and so far so good. But I am noticing that large geometric changes still require many, many, switch layers.
has anyone tried this?
any pitfalls to avoid?
advice to a beginner/intermediate AS user?
thanks for any insight/advice you may be able to provide.
-edg
hybrid drawing and cut-paper technique?
Moderators: Víctor Paredes, Belgarath, slowtiger
-
synthsin75
- Posts: 10276
- Joined: Mon Jan 14, 2008 11:20 pm
- Location: Oklahoma
- Contact:
I'm betting there are some shortcuts you could take, but I'd have to see an example of the intended 'geometric changes' to know where to start. Even a couple of stills illustrating the initial and end result you're aiming for.
There is a tutorial somewhere that shows how to do frame-by-frame directly in AS.
There is a tutorial somewhere that shows how to do frame-by-frame directly in AS.
-
synthsin75
- Posts: 10276
- Joined: Mon Jan 14, 2008 11:20 pm
- Location: Oklahoma
- Contact:
http://my.smithmicro.com/tutorials/1952.html
It's vector based, but convincing freehand frame-by-frame. Probably won't help much with the number of layers needed though.
It's vector based, but convincing freehand frame-by-frame. Probably won't help much with the number of layers needed though.
example
got it, this is all PNG image based, hence the dilemma.
try this link for a small example. try to watch on a fast connection or it'll certainly look "choppy" and appear to have fewer frames than it does.
It's actually animated at 24pfs. and the original looks pretty darned smooth.
Please let me know any thoughts?
http://www.myspace.com/edgendron
then click on "videos"
try this link for a small example. try to watch on a fast connection or it'll certainly look "choppy" and appear to have fewer frames than it does.
It's actually animated at 24pfs. and the original looks pretty darned smooth.
Please let me know any thoughts?
http://www.myspace.com/edgendron
then click on "videos"
Frame by frame with images is going to be a TON of work with AS. It isn't exactly designed for that purpose. You can certainly do it but it's not going to be efficient. AS is designed for "reusing" what you create. for instance you use an image of an arm or leg. Instead of drawing a new picture for each frame of that image you rotate or animate the image. If you create a vector shape, you don't redraw it for each frame, you move the points or use bones to move the points of the shape around over time. That is the strength of AS, bones and point motion.
From your sample video it would appear that are using "layers" of moving images. This would work in AS very well. Using a series of image sequences to do turns or rotations could be done using switch layers and actions (an action with keys for the sequence).
I just did a quick test with a switch layer and about 5 frames. I created two actions, one that cycles through frame by frame from switch layer 1 to 5, and another that goes from 5 to 1. You could then insert that action reference at any point in the time line to repeat that sequence without having to key the switches.
You could automate creating these sequenced switch layers by using the image sequence import script. It brings in sequentially numbered images into a switch layer automatically keyed for frame by frame display of each image.
You probably could also create sequential image switches INSIDE a switch and do even more things with it. For example a series of image layers that display one at a time in a switch layer. Have a bunch of different ones inside a switch. Create actions for each different sequence and insert them anywhere in the time line. Actions will save a ton of time. Using that with bone binding you could do all kinds of cool things like for instance an animated hand opening and closing or any other position as image layer sequences in a switch within a switch.(hope this makes sense).
The trick to make it easier is to "automate" the process as much as possible and "reuse" as much as you can. Plan the animation out very carefully ahead of time. Using images this way is probably going to need a bit more planning.
-vern
From your sample video it would appear that are using "layers" of moving images. This would work in AS very well. Using a series of image sequences to do turns or rotations could be done using switch layers and actions (an action with keys for the sequence).
I just did a quick test with a switch layer and about 5 frames. I created two actions, one that cycles through frame by frame from switch layer 1 to 5, and another that goes from 5 to 1. You could then insert that action reference at any point in the time line to repeat that sequence without having to key the switches.
You could automate creating these sequenced switch layers by using the image sequence import script. It brings in sequentially numbered images into a switch layer automatically keyed for frame by frame display of each image.
You probably could also create sequential image switches INSIDE a switch and do even more things with it. For example a series of image layers that display one at a time in a switch layer. Have a bunch of different ones inside a switch. Create actions for each different sequence and insert them anywhere in the time line. Actions will save a ton of time. Using that with bone binding you could do all kinds of cool things like for instance an animated hand opening and closing or any other position as image layer sequences in a switch within a switch.(hope this makes sense).
The trick to make it easier is to "automate" the process as much as possible and "reuse" as much as you can. Plan the animation out very carefully ahead of time. Using images this way is probably going to need a bit more planning.
-vern
-
synthsin75
- Posts: 10276
- Joined: Mon Jan 14, 2008 11:20 pm
- Location: Oklahoma
- Contact:
Nice looking. It has a real vintage feel to it.
I don't really have any useful advice since I don't work with images much. But as a general rule, the more 'peices' you can break the image into the easier it is to animate by bone and layer manipulation.
Complex motion, like the wing flapping, probably can't be done by anything other than switch layers. But I could see animating everything else by bone and layer manipulation. Or at least very minimal switch layers.
Great work so far.
I don't really have any useful advice since I don't work with images much. But as a general rule, the more 'peices' you can break the image into the easier it is to animate by bone and layer manipulation.
Complex motion, like the wing flapping, probably can't be done by anything other than switch layers. But I could see animating everything else by bone and layer manipulation. Or at least very minimal switch layers.
Great work so far.
I think you could do simple "wing flaps" with bone warping on the image. Scaling the bones would produce a "pseudo" perspective on the image making the wings appear to flap up and down. The overlapping "undulation" might even be done using a couple of "wing parts" with several bones.
The key is reusing the images as much as possible rather than replacing them with a completely new image. That is probably why some people advise you to avoid AS if your technique is geared more towards "cell" or frame by frame animation. AS just isn't really designed to handle frame by frame cell animation that well. I'm not saying it can't do it, but if that is your primary technique and you want frame by frame pixel level control of each frame you might want to look at other options.
-vern
The key is reusing the images as much as possible rather than replacing them with a completely new image. That is probably why some people advise you to avoid AS if your technique is geared more towards "cell" or frame by frame animation. AS just isn't really designed to handle frame by frame cell animation that well. I'm not saying it can't do it, but if that is your primary technique and you want frame by frame pixel level control of each frame you might want to look at other options.
-vern
From that short example I think it's possible to work like that in AS. It only will be a lot of work - but this doesn't prevent Yuriy Norshteyn from creating amazingly 3D-like animation with cut-outs, for example. I've learned a lot about animation in the past which was done in the most time-consuming way possible - animating and colouring on paper, then cut out every character and glue it on cel (Michael Sporn), or cutting characters out of very fine tissue paper and animate it in numerous layers on a lighttable ("Twice Upon a Time", John Korty), or even tissue paper and feathers like Te Wei. Also have a look at JibJab's work, they sometimes use image distortion to good effect.
I think that AS will be capable of most if not all scenes you're planning in your style. Just don't expect it to be very fast when scenes get very complex. Maybe it's a good idea to roughly lay out scenes in vector first, just do check movement and timing, then add the image layers for the final result.
Careful planning is essential in restricting the workload. I'd like to see a storyboard of what you're planning, that way I could spot difficulties easier.
I think that AS will be capable of most if not all scenes you're planning in your style. Just don't expect it to be very fast when scenes get very complex. Maybe it's a good idea to roughly lay out scenes in vector first, just do check movement and timing, then add the image layers for the final result.
Careful planning is essential in restricting the workload. I'd like to see a storyboard of what you're planning, that way I could spot difficulties easier.
im a big fan of alexander Petrov ever since i saw did "the cow" (among other things) and actually tried the paint on glass thing. i was thinking "this is like stop mo, it'll be faster than cel animation"
it wasn't. it took just as long as cel. It seemed to take forever. (it was gorgeous though) anyhow i had to scrap that idea since i realized that id never complete that movie at that pace. paint-on-glass may be THE most laborious animation medium that ive ever attempted. (im also a big william kentridge fan but same applies to his technique - takes too long)
a little later, i did a film (called Fishbowl) in a live action/cgi mix. Sometimes, i'd have to show people scenes that werre only half shot. so how are they supposed to know what the UNshot scene will look like? I *very* quickly and very roughly sketched in certain shots and made them move minimally (an animatic) to give the very basic impression and fill in the gaps. it was meant as filler, nothing more. (ive never felt that they were special) Yet people RAVED about these drawings and encouraged me to animate in that style. I couldn't understand it, but it happened repeatedly.
so I finally gave in and now Im trying to do something like that in an abbreviated digi-cel style. I don't draw every frame, nor is it as choppy as southpark. Im not doing vectors bc i like the "painted look" as opposed to the flatter looking vector style. (perhaps that'll change)
so basically what im looking to do is speed up the process as much as possible. it appears that the largest geometric changes (such as a slowly spinning dancer) would cause the most work. whereas a fast spinning dancer would be ok. Im trying to do this with a good organic look but I really want to avoid swamping myself with work since im only one guy as opposed to the entire disney animation team.
I *have* considered vector layers as augmentation such as moving highlights (perhaps) or for things like that but i haven't tried it yet. maybe even semi-transparent vectors.
it wasn't. it took just as long as cel. It seemed to take forever. (it was gorgeous though) anyhow i had to scrap that idea since i realized that id never complete that movie at that pace. paint-on-glass may be THE most laborious animation medium that ive ever attempted. (im also a big william kentridge fan but same applies to his technique - takes too long)
a little later, i did a film (called Fishbowl) in a live action/cgi mix. Sometimes, i'd have to show people scenes that werre only half shot. so how are they supposed to know what the UNshot scene will look like? I *very* quickly and very roughly sketched in certain shots and made them move minimally (an animatic) to give the very basic impression and fill in the gaps. it was meant as filler, nothing more. (ive never felt that they were special) Yet people RAVED about these drawings and encouraged me to animate in that style. I couldn't understand it, but it happened repeatedly.
so I finally gave in and now Im trying to do something like that in an abbreviated digi-cel style. I don't draw every frame, nor is it as choppy as southpark. Im not doing vectors bc i like the "painted look" as opposed to the flatter looking vector style. (perhaps that'll change)
so basically what im looking to do is speed up the process as much as possible. it appears that the largest geometric changes (such as a slowly spinning dancer) would cause the most work. whereas a fast spinning dancer would be ok. Im trying to do this with a good organic look but I really want to avoid swamping myself with work since im only one guy as opposed to the entire disney animation team.
I *have* considered vector layers as augmentation such as moving highlights (perhaps) or for things like that but i haven't tried it yet. maybe even semi-transparent vectors.
Slowtiger has a FANTASTIC idea there.
You use fast easy to render vector layers to do the "animation" then swap out with images. The trick would be to put the vector INSIDE a group or bone layer and animate the GROUP rather than the layer. Then you just stick the image layer inside the group and the animation is intact.
Another option would be to animate the vector layer directly and then use a text editor to change it to an image layer keeping the layer keys intact... much more work and kind of pointless I think... maybe.
The reason I think this might be a great trick is based on what Slowtiger mentioned about JibJab. Lots of "new" animators often use them as their inspiration for doing cut out style animation... the thing is... JibJab animation looks so good because the ANIMATION is fantastic. Doesn't matter if it is simple image cutouts or not... the movement is so "alive" and well done. That's what makes it look so fantastic. If you watch their stuff it is so deadly simple in layout and structure but the animations they can produce are top notch.
So using rough vector shapes to define the "outlines" of what the images will be will speed up the animation preview process considerably. Lots of image layers will really slow down performance.
Another benefit is you can do animation AND work on the images at the same time. The vector animation can also give you feedback on how the images might be done differently or "better" to suit the style.
----
Off on a tangent... wild crazy idea... a script to "swap" layer key frames. Maybe I can just use my bone animation saver and use it for layer animation. Save and load the animation from any layer onto any other layer. that would be cool for those doing image animation like this. they can draw simple minimal vector layers for placement and then swap out for images later.
-vern
You use fast easy to render vector layers to do the "animation" then swap out with images. The trick would be to put the vector INSIDE a group or bone layer and animate the GROUP rather than the layer. Then you just stick the image layer inside the group and the animation is intact.
Another option would be to animate the vector layer directly and then use a text editor to change it to an image layer keeping the layer keys intact... much more work and kind of pointless I think... maybe.
The reason I think this might be a great trick is based on what Slowtiger mentioned about JibJab. Lots of "new" animators often use them as their inspiration for doing cut out style animation... the thing is... JibJab animation looks so good because the ANIMATION is fantastic. Doesn't matter if it is simple image cutouts or not... the movement is so "alive" and well done. That's what makes it look so fantastic. If you watch their stuff it is so deadly simple in layout and structure but the animations they can produce are top notch.
So using rough vector shapes to define the "outlines" of what the images will be will speed up the animation preview process considerably. Lots of image layers will really slow down performance.
Another benefit is you can do animation AND work on the images at the same time. The vector animation can also give you feedback on how the images might be done differently or "better" to suit the style.
----
Off on a tangent... wild crazy idea... a script to "swap" layer key frames. Maybe I can just use my bone animation saver and use it for layer animation. Save and load the animation from any layer onto any other layer. that would be cool for those doing image animation like this. they can draw simple minimal vector layers for placement and then swap out for images later.
-vern
Vern, the credit for this goes to you as I was suggesting just a different way of testing some animation before going into the labourious clean up process. I can't even imagine what you think the result will be.
eg: I don't know how much independent animation from Europe you know, especially the "painterly" one. I have seen loots of it, from stuff from the 50's up to recent attempts. I spotted some conflict in most of those movies, the conflict between movement and painting.
A typical scene from those would show an elaborate painting of a human holding some flower in his/her hand. This image would be shown for some time, and to make it intereresting it would be animated in some way. Sometimes the whole image would be re-painted, sometimes only moving parts, sometimes dissolves would be used, sometimes straight stop motion. In any case, the whole thing wasn't a bit more convincing or emotional than a still image, and even worse when the flower was drooping, or some tears forming in the face. I often sat through the film with the nagging question "Why was this animated? Why not just be one painting? Or a short poem?"
(I'm going onto thin ice and highly theroretical now - you've been warned.)
I think there are some parameters to be adjusted which maybe are impossible to be lined up to a successful result.
- First there's the story, the idea, the Something that wants to be expressed.
- Second there's the emotion which shall be invoked through a certain style of painting or drawing.
- Third there's suspense and tension, everything dramatic which develops over time.
- Fourth there's the need for movement in some way, if only to justify the use of film instead of a slide show.
In some way this very much shouts "theatre!" to me, especially the kind of modern tableau stage work like Wilson. But what works nicely with living actors and a live audience gets pretty boring fast on film. Everytime I had to endure a scene with an "important picture" for some time to "get the mood" or whatever the director of that film intended to blow me over the head with, I got angry and felt insulted. An image doesn't get stronger the longer it is shown on screen. A mood or idea isn't transported any clearer the more work is spent in the painting. And a movement doesn't get more meaning if it was slowed down and smoothed endlessly.
I'm still waiting for that "painterly" animated film which gets me as much involved into it as those in other techniques. I avoid to blame the medium for the effect, but maybe it's just because painting and animating are complete opposites to each other. Please, somebody prove me wrong.
eg: I don't know how much independent animation from Europe you know, especially the "painterly" one. I have seen loots of it, from stuff from the 50's up to recent attempts. I spotted some conflict in most of those movies, the conflict between movement and painting.
A typical scene from those would show an elaborate painting of a human holding some flower in his/her hand. This image would be shown for some time, and to make it intereresting it would be animated in some way. Sometimes the whole image would be re-painted, sometimes only moving parts, sometimes dissolves would be used, sometimes straight stop motion. In any case, the whole thing wasn't a bit more convincing or emotional than a still image, and even worse when the flower was drooping, or some tears forming in the face. I often sat through the film with the nagging question "Why was this animated? Why not just be one painting? Or a short poem?"
(I'm going onto thin ice and highly theroretical now - you've been warned.)
I think there are some parameters to be adjusted which maybe are impossible to be lined up to a successful result.
- First there's the story, the idea, the Something that wants to be expressed.
- Second there's the emotion which shall be invoked through a certain style of painting or drawing.
- Third there's suspense and tension, everything dramatic which develops over time.
- Fourth there's the need for movement in some way, if only to justify the use of film instead of a slide show.
In some way this very much shouts "theatre!" to me, especially the kind of modern tableau stage work like Wilson. But what works nicely with living actors and a live audience gets pretty boring fast on film. Everytime I had to endure a scene with an "important picture" for some time to "get the mood" or whatever the director of that film intended to blow me over the head with, I got angry and felt insulted. An image doesn't get stronger the longer it is shown on screen. A mood or idea isn't transported any clearer the more work is spent in the painting. And a movement doesn't get more meaning if it was slowed down and smoothed endlessly.
I'm still waiting for that "painterly" animated film which gets me as much involved into it as those in other techniques. I avoid to blame the medium for the effect, but maybe it's just because painting and animating are complete opposites to each other. Please, somebody prove me wrong.