15 Replies Latest reply on Jul 29, 2016 10:47 AM by Roei Tzoref

    How would I achieve this effect?

    ryanb74751266 Level 1

      I'm looking to re-create the effect as seen at 1:10 in this video below.






      Any help would be appreciated, especially the part where each individual smile falls down differently. Also, when each smile appears, I thought at first it was just a scale in, but it looks like it has a different effect to it.




        • 1. Re: How would I achieve this effect?
          Rick Gerard Adobe Community Professional & MVP

          You have basically three or four options. The first would be to create the    using a single shape layer then and a bunch of repeaters. You would duplicate your single shape layer a couple of times then animate everything by hand.



          Another option would be to stack up a couple of shape layers with repeaters to build up layers then use Shatter to make the faces fall. This will require a fairly extensive knowledge of Shatter and custom shatter maps.


          Another option with Card Dance ( which will also require a significant amount of time to learn) .


          You can learn how to use these two plug-ins my typing the name in the search help field at the top right corner of AE.


          The next option would be to use a shape layer as a custom particle and use Trapcode Form or Particular.

          • 2. Re: How would I achieve this effect?
            Roei Tzoref Adobe Community Professional

            I would got manually on this sucker. nothing to it. take a look:


            you create two layers in photoshop or illustrator

            import these to Ae. make an animation for one of them - the "different effect" you notice - this is called an overshoot-undershoot animation and you need to use the graph editor for that. first you set your scale to 0 and after 10 frames to 100 like this:

            then you go to the graph editor and on the value graph you change your curve from this:

            to this:

            see this is an overshoot and undershoot waveform more about it here:



            this is my clumsy way of explaining how this works


            move the keyframes closer until you get the timing right and now you got this


            so we got the scale just right! now about the position - since we know we want many of these, we should try to work D.R.Y - Don't Repeat Yourself - we make one animation and duplicate it instead of adjusting every single one. let's add position to the mix and we will take smiley all the way up and all the way down and then some so we have room when we scale it afterwards.

            now we want to change the speed a little so we'll go to the graph editor on speed graph and adjust it like this:

            this is speed graph so at the start we want less speed and to gradually increase it

            now we got one fine dandy of an animation


            now we precomp this layer and call it smiley-animation and press the collapse button so we can now in 13.8 version enjoy the beautiful cropped boundaries of a pre-comped layer

            we drag the anchor point to the middle so it will be easier to position the layers an scale them afterwards

            now we duplicate like crazy by pressing Ctrl+D and dragging in the comp window with the selection rectangle and what not

            and now we scale some of the layers down a little

            and offset the layers

            now we got this amazing thing:


            oops... got some that didn't go down enough. not to worry - it's all in the same comp! we can go inside and drag it lower!

            what about laughing face? we are working D.R.Y. we are just duplicating the smiley animation precomp in the project window and call it laugh animation

            go inside and replace smiley layer with laugh layer

            and now in our main composition we replace some of the smiley precomps with laugh precomps from the project window


            now we are DONE: TAKE A BREAK!


            • 3. Re: How would I achieve this effect?
              Roei Tzoref Adobe Community Professional

              Jesus Rick, 5.4mb of a Png. way to bog down a thread

              • 4. Re: How would I achieve this effect?
                Rick Gerard Adobe Community Professional & MVP

                Just a screenshot from an iPhone 


                I would have used shape layers and repeaters. A lot more controllable with only a few layers.

                • 5. Re: How would I achieve this effect?
                  ryanb74751266 Level 1

                  Wow Roeil!


                  Thanks so much for laying this all out for me. I appreciate it!


                  I do have one question though. When I duplicate them, they all just follow the original down and I don't want that, I want them to travel on their own.


                  How do I fix that?

                  • 6. Re: How would I achieve this effect?
                    Roei Tzoref Adobe Community Professional

                    thanks Rick.


                    about shape layers - I am not sure about that. there are many things that could go wrong with this approach and frankly I don't think it would even work, but why don't you show us? here are a few complications with going the shape layers approach:


                    1. there isn't more control than doing this stuff manually. if you have one smiley face that's covering something important - you can just remove it. if you used repeater for it - you can't control a single one. so in fact it's HARDER to control like with all automated process - they are good for automated things but not for manual control.


                    2. the only way you could use shape layers is if you have the art in an illustrator file and it converts to shape layers accurately. if you have gradients, opacity, blend modes - it will not, so no shape layer


                    3. you need to work in a limited space in your timeline this is hard for many users, especially beginners.


                    4. let's say you made a scale and position once. you can't use the repeater manually like that and create a random pattern. repeater is good for repeating pattern. here's what happens when I try to disperse the layers.

                    so now I need to create more instances and this is very NOT intuitive and very NOT controllable


                    I hope to learn new things from you please show us how you would nail it.

                    • 7. Re: How would I achieve this effect?
                      Roei Tzoref Adobe Community Professional

                      Sure thing Buddy. you caught me in a mood and once I start, I have to finish.


                      what do you mean by "Travelling on their own"? if you offset the layers the will each move in another point in time. I am just making it like the reference but if you want something else please do give me a challenge. we are working with key-frames here so this is done manually. if you want this to be done automatically you need CODE for that, and that's another approach far more complicated that requires a great deal of understanding in scripts and expressions. this would be good only if you need to make 100 of those animations in 100 sequences and needed to automate things. the way scripts and expressions work in automating these types of animations is that they are only productive in the long run. you invest time and energy setting things up and then sit back (unless you messed it up and then you are feeling worse because you can't get that final 10% of exactly what you wanted to do because the code is not absolutely matching your scenario - which is the case MANY TIMES)

                      • 8. Re: How would I achieve this effect?
                        ryanb74751266 Level 1

                        Like when I duplicate the smiley layer, that new layer's positions follows that of the first one. How do I get it so I can easily change the position?

                        • 9. Re: How would I achieve this effect?
                          Roei Tzoref Adobe Community Professional

                          it follows the first one because they all have the same animation so if you offset one in time, the second, third, forth an so on "Follow" the first one. you can change the position SUPER EASY by just placing it in another place like I showed you. you take one layer, place it in another place and it falls down like the first one. if you want it to "follow" the other, just offset them in time in a way they are closer together and create a sequential editing (layers like steps one by one) and it will appear to follow each other.


                          what I am saying is that you make this stuff happen so it appears to behave in certain way. you're not creating conditions for movement but creating the movement itself. if you want to create conditions for physics for example: gravity, stickiness, bounciness and then throw all of the layers into this kind of mix - that's even another approach and that's a fair question this is not the way 90% of Ae works. for example "Motion" is a type of animation software where you have conditions for movement. you set speed and angles and the layer does it's thing, After Effects on the other hands works differently - you create key-frames and set manual adjustments and if you are smart about it - it can appear as things follow each other of do a certain type of physics.


                          having said that there are plugins and scripts that will create conditions so you could have the layers behave in a certain way. if you insist on going the physics way - you can try working with this plugin: Newton 2 - aescripts + aeplugins - aescripts.com  it is amazing but like I said with automated processing  - you will need to invest time and know how to set it up just right so it would behave in a certain way and you lose manual control over your layers, for the sake of the global condition you set for ALL of your layers.


                          if you had a 1000 smiley faces, I would say go for an automated process. (particle plugins or others). if you had a smiley face that is bouncing in a pin ball machine and requires elaborate motion that is based on physics I would recommend the Newton plugin.

                          • 10. Re: How would I achieve this effect?
                            Rick Gerard Adobe Community Professional & MVP

                            Roei, Repeaters will work just fine, you just have to know how to use them. For example:


                            One smily face shape layer, (I made two different colors of shape layers to give 200 smily faces so you could how the two layers interact) then added two repeaters for the group with 10 copies each, added Card Dance using two animated gradient layers pre-composed with Mosaic added to create a 10 X 10 map, 10 minutes production time and the faces are all different sizes and fall off the screen one at a time at different times and even different rates - a total of 8 keyframes.


                            Throw in Gradient wipe at the head and the smily each individual smily face can appear at a different time in a different position then fall off the screen at different times with a total of 10 keyframes, three nested comps, one for size and x and z position for card wipe, one for y position in card wipe, and one for gradient wipe to make the 100 faces appear randomly (Pre-comps are just duplicates of the first with slightly different settings) and two shape layers = 200 smily faces.

                            Screen Shot 2016-07-29 at 7.16.31 AM.png

                            You could use similar techniques with Shatter - gradients for shatter map and custom shape. With Card Dance or Shatter you could also add a Camera and fly through the faces because both Card Dance and Shatter will simulate 3D and use the comp camera.


                            I won't take the time but you could use Trapcode TAO, Form and Mir to create an even more impressive animation with just 3 layers, but this would require investing in 3rd party plug-ins.


                            My point is that there are many ways to achieve any single effect. Taking the time to animate and control 300 or so layers (a rough guess at the number of faces in the original) even using Roei's technique will require a lot more fussing and time than using a more efficient technique if you have taken the time to learn how to use the tools that come with AE. BTW, I never suggest something that will not work, and I never think that my way is the best way.


                            One last point - my second suggestion was to duplicate your shape layers and do everything by hand - the technique marked as the correct answer. 300 layers, I'm guessing thirty minutes to an hour for an experienced user. I haven't used Card dance for probably more than a year and the example took me about the same amount of time to create as it did to write and edit this post.

                            • 11. Re: How would I achieve this effect?
                              Roei Tzoref Adobe Community Professional

                              Rick this is genius work here. please note that I am writing this with respect to your approach and technique. I am here to learn.


                              here goes:


                              there is absolutely no argument that there are different ways to achieve an effect. the approach should be judged by the final result, the clarity of the technique, and time invested in the process. about your approach - I can't judge the result because I can't see it in motion, the clarity of the technique is messy and very very complicated,  and the time required is in question to anybody but a genius. you are complicating things immensely in my opinion. I mean what are we talking about here? - making layers sequentially appear and fall down in a fine fashion. there is no other control required here - no camera, no special kind of 3d effect. PURE ANIMATION. I will be so bold to say I would expect that out of 10 people you explain what you just did, only one would get it and that would be YOU. I mean just look at this:

                              just look how many parameters you have to deal with here this is crazy. you got 3 layers with 100 parameters - is that better than 100 layers with 2 parameters? scale and position! that's it! it's nothing! look at what you set up here:


                              Card dance/Shatter

                              3 nested comps

                              Fractal noise


                              Gradient maps


                              I mean can't you see this is insane? I am being serious here. and this is not mentioning that your graphic is based on an artwork, but not the artwork itself that could or could not be vector, and could be or could not be converted to shapes in an accurate manner - we don't know. also what about specific fine tuning? how does that work when all of them are controlled as groups by sliders? what happens when your client tells you - "hey listen that one smiley face at the bottom there - make him gone!" or "make that other one a bit different", "add some drop shadow to each one"? there is no substitute for manual control and as long as this is not 3,000 smiley faces - you are well off doing this manually.


                              I believe your approach in this example is great for you. and maybe a few rare Ae users, but I honestly believe that if you can't explain what you did in a clear, easy to follow manner, and if it's that rocket science complex for a pretty straightforward example - your are not teaching but showing off. maybe you are that good, but users need to follow this eventually and make sense of it.


                              and this should be on a T-shirt

                              BTW, I never suggest something that will not work
                              • 12. Re: How would I achieve this effect?
                                ryanb74751266 Level 1

                                Here is an example of what I mean. When I duplicate the layer, it "follows" the other one, and then follows it down. I want it to appear in its own place, and then fall down with the rest, not follow the first one.


                                How would I achieve that?



                                • 13. Re: How would I achieve this effect?
                                  Roei Tzoref Adobe Community Professional

                                  that's why the keyframes are in the precomp and not in the master comp. let me explain: you have one animation of an object that's scaling up and falling down - that's the animation - when you close it in a precomp (move all attributes) you are essentially rendered a closed animation sequence. now you can move it anywhere you want and it will fall where you put it and not follow the other. you are moving a closed animation of a scale and position movement. get it?

                                  now we precomp this layer (move all attributes - maybe I should have mentioned that) and call it smiley-animation and press the collapse button so we can now in 13.8 version enjoy the beautiful cropped boundaries of a pre-comped layer

                                  • 14. Re: How would I achieve this effect?
                                    Rick Gerard Adobe Community Professional & MVP

                                    Don't want to start a war here but I've taught this card dance technique many times. One time at the Post Production World Conference at NAB in Las Vegas with about 250 in the audience. I spent about 15 or 20 minutes doing through Card Dance and every body got it.  The technique is quite easy to understand and control as soon as you start manipulating the gradients. By the end of my presentation I took about 10 minutes to answer questions and encourage the the audience to experiment on their own projects. I then had a half dozen of the students come up and how they had used the techniques to create completely different and animations as the group discussed options. Then it was on to other subjects. Same routine, demonstrate the basics, get them started, answer questions while they experiment and check out what they have done.


                                    I'm not here to hand hold the hand of the person asking the original question, I'm here to suggest options will point to solving the original question and also offer suggestions that will be valuable to users of every experience level. Maybe I should have put animating by hand and using duplicates as the first suggestion instead of putting it second. Because of the other options that are available I think that using animate by hand to to 300 or more layers in a random manner is a but nuts. There are many followers of this forum that already know how to animate using your technique or slight variations of animating by hand, but I'm guessing that most of them would never think of using Card Dance, Shatter or 3rd Party effects to give them more options that can easily improve productivity on something like this specific project and be useful on hundreds of other completely different effects.


                                    To learn Card Dance there are a bunch of good community resources available using the Search Help field at the top right corner of AE. Creating a simple shape layer, adding card dance and adding a nested comp with the settings I revealed then playing with only the two values I animated in the nested comp and changing the values of the 5 values I modified in Card Dance will give you a good idea how the effect works. It would be a great learning experience for anyone wanting to try something other than duplicating a layer and animating by hand.


                                    In a real production environment the idea of filling a screen with happy faces and being picky about where every one falls is a little nuts. If you have a hero in the shot, then put that hero on a separate layer. Nobody's eyes can judge anything but general movement over time. If I have a client that would look at the effect shown in the video and then routinely say something like  "hey listen that one smiley face at the bottom there - make him gone!" or "make that other one a bit different" it wouldn't take me long stop working with that client or start charging him for any more than 3 revisions. That's how you stay in business.


                                    It would be a good Idea for any student to learn how to use multiple repeaters. I don't have hundreds of modified properties in my example. 8 settings in Card Dance modified to use the gradients, 2 properties in the gradient layer modified from the default settings and animated (change the values in the duplicate comps), and a color edit in the gradient fill for the gradient solid layer.


                                    Now let me throw you a different animate individual layers idea. This is an "insane" idea for a brand new user but might be fun for someone with experience that want's to learn expressions. It at least demonstrates some of the real power in expressions. Start with a single image layer with an expression that a controls scale, starting position, maybe opacity based on the in point and out point of the layer. Now If your expression is written correctly then all you have to do is position the layer, duplicate it, set the in and out point and repeat the process. You get something like this:

                                    Screen Shot 2016-07-29 at 10.17.59 AM.png

                                    The expression uses some sliders but I'll make it easy for you, here's a freebie: Dropbox - flyInBounceDropOut.ffx

                                    Here's the project file: Dropbox - GroupAnimation.aep (note: your browser may add a .txt extension the the aep file. Just remove it)


                                    Here's the expression for position:

                                    sif = effect("slideIn Frames")("Slider");

                                    sof= effect("slideOut Frames")("Slider");

                                    // find in and out point

                                    inTime = sif * thisComp.frameDuration;

                                      if (effect("Match SI SO")("Checkbox") == 0) {

                                      outTime = sof * thisComp.frameDuration;


                                      else if (effect("Match SI SO")("Checkbox") == 1) {

                                      outTime = inTime;


                                    // Fix  layer size when scaled

                                    sf = scale - [100, 100];

                                    xSize = width + (width * sf[0]/100);

                                    ySize = height + (height * sf[1]/100);

                                    realSize = [xSize, ySize];

                                    // Set Positions

                                    spx = 0 - realSize[0] + realSize[0]/2 ;

                                    rstx = value [0];

                                    rsty = value[1];

                                    epy = thisComp.height + realSize[1] - realSize[1]/2;

                                    // create movement

                                    freq = effect("bounce Frequency")("Slider");

                                            amplitude = effect("bounce Amplitude")("Slider");

                                            decay = effect("bounce Decay")("Slider");

                                        posCos = Math.abs(Math.cos(freq*time*2*Math.PI));

                                      nt = time-inPoint;

                                        y = amplitude*posCos/Math.exp(decay*nt*2);

                                    moveIn = Math.min(linear(time,inPoint,inPoint + inTime,spx  ,rstx-y));

                                    moveOut = Math.min(linear(time,outPoint - outTime,outPoint - thisComp.frameDuration,rsty ,epy));

                                    [moveIn, moveOut]


                                    This expression does not care what the original position of the layer is or what the scale is. It moves the layer to the left side of the frame at the in point, then fly the layer into it's original position (which you can animate if you like) and then just before the out point, based on the number of frames chosen, the layer drops off the bottom of the frame.


                                    You will find a couple of other expressions in the sample project. One changes the colorize hue for each of the layers based on a random number and the index of the layer, and the other randomizes the scale of the layer so I don't have to go in and manually scale anything. I have about 200 of these animation presets - this is the one that I give away for free. If you want to talk insane for a new user try explaining how that expression works for a new user.


                                    BTW, just for fun change the x move to a y move (if you can figure out where it is) to get the layers to fall into frame and then all off.

                                    BTW, just for fun, try and write an expression for scale or opacity that uses the in and out point.


                                    When you get a bunch of expressions for animations based on in and out point or layer markers start saving them as animation presets so you don't have to re-invent the wheel every time you want to do something.


                                    Here's the result:


                                    • 15. Re: How would I achieve this effect?
                                      Roei Tzoref Adobe Community Professional

                                      thank you Rick. I now understand you choose a rather special approach on purpose to show the possibilities and my hat goes off to you. thank you for sharing the techniques and files and taking the time to write your posts filled with useful information, I will look into it. until next time, which I assume will be very soon...