 I'll start. Thank you all for being here. My name is Nika and I am a 3d generalist and animator from Felicia, Georgia. And I'm here to talk to you about puppets. So we've heard about the giant stuff like the shows and the pipelines and stuff. But now we need to get infinitely smaller and talk about puppets and the real ones too, because I am representing a stop motion animation studio called Phantasmagoria. This is our logo and this is what we do. We are based in Felicia. We are a new studio. Just had like one year anniversary just last month. So this is the first short film that we're working on and we're also a very young company. It's also, I don't know, it's called English. It's a social factory, I think. So we basically teach kids, the students, and then the students once they graduate, they make films. So this is the first one. And as you can guess, we're also a very young company. I'm also actually one of the older ones. And we like to not only have this traditional medium with this all traditional means, the stop motion is very traditional and handmade, right? That's why we love it. But we also try to integrate and adapt to all new technologies that we can. And that's why I'm here. Because we decided the directors were very adamant from the start to use 3D printed replacement parts. Sorry about that. The font got mixed in the conversion to Linux, okay? So if you don't know what is a replacement part in the stop motion animation, it is one way to create the illusion of motion. As you know, the stop motion is one frame is each shot. So to create a motion, you can deform the object and basically sculpt the animation. Or the second way is to have them sculpted beforehand as separate objects for each frame, as you can see on the picture, and just replace them and attach to the puppet one frame at a time. And usually, traditionally in like before, I think like 10 or 15 years, it was done also by hand. So people will just create every replacement part by hand that was sculpted. And that was, I think, one of the most time consuming aspects of stop motion animation. It could take months to years if you have like full-length animation. So once there was a chance to integrate new technology and once the 3D printing became a thing, the big studios quickly adapted to it and created the workflow for 3D printing replacement parts. The first film I think that used this method was like 15 years. I think the first one was maybe Coraline, if I'm not mistaken. So the Leica studio is using it, the Ardman is using it, the big studios. But for the smaller studios, it was not yet, it was available but it was not yet feasible to do it because it required some cost, you know, because the printers costed the money but now as it's becoming cheaper, the commercial printers are becoming cheaper and there's more software, more information, it's becoming good. So we have now reasons why small studios can use it. It's not widely adopted yet but the many studios are already using it. So it's cost and time effective if you do that. It's easier to produce and reproduce which is important so if something happens and you lost the face in our case or some hand maybe and something got corrupted, you need the copy of the one you had, it's much easier to reproduce if you have them as 3D objects and you're printing them and they can create much better animation because you can preview the animation in 3D and print them instead of creating by hand and just eyeballing what animation should look like. But there's reasons why you shouldn't do it and the most of it is that there is just no experience at all. There is nothing you can learn from, there's no documentation. The big studios don't really share this information and for smaller studios it's not, like when we were starting this and we're googling and searching for what we could find, there's almost nothing next to nothing. There's some videos where studios are talking about how they are using it but it's mostly bragging. It's not what they're doing, it's like oh we're doing this and we're cool but they don't teach you anything, they don't give you any information so that's why one might refrain from using this method because it requires some time to learn and adapt but we still decided to do it. Directors are a big fan of this method and they wanted to implement so we did this and the motivation for this presentation is that all the steps that we took, the growing pains and the things we had to figure out, the mistakes we had to make to figure out the answers, all that stuff maybe not every studio has to go through this, not every artist has to learn from the scratch so I want this to be first case maybe of sharing this experience and information and we're going to get technical, you're going to get technical tips so that the other studios and artists who will use this method will have something to base their work upon. So we're going to talk about four things basically, first we're going to talk about modeling first of motion so what's the difference, what unique and intricate parts it has compared to the regular modeling and then talk about the same thing for 3D animation. Mostly we're going to talk about custom tools that we created because as you can imagine because this is a very unique and new workflow, there's not tools available yet that you need so we needed to create them and lastly and also we're going to talk about the tools that we found and basically changed to fit our workflow because Blender community was very helpful in every way in direct and indirect way we found some scripts and add-ons and stuff that we could use in our workflow and lastly we'll talk about how we organize and it's a fun part. Okay so first it's modeling so we have the workflow basically is that we are two people in the 3D department the studio is very small we're just 20 people and we just have like two floor studio it's very small and we are two people working on 3D basically the 3D department it's me and Lasha who's sitting there and our pipeline basically is that he's modeling the sculpting those faces and creating basically ricks and then I animate them and I send them back to him and he prints them so we're going to go through this workflow and this is a modeling which Lasha does I'm not doing this but I know something so I'm going to talk about so the faces at first are modeled in a very regular way they are just sculpted we apologized the multires shrink rub everything and we're using references from our second puppeteer Mariam and we have this particular style because our film is set in at the start of the 20th century it's from I think 1919 to 1920s so we wanted to be more stylized in the painting style that they liked at the time especially in Georgia but then we need to create some adjustment so that when we print this we can use them on puppets so the first one is that we need to split the head in middle because we're not printing entire head for every frame we just need the head, the face, the back part is static we just printed one suite, attached to the object it's part of the main puppet so we need to split it and have some mechanisms to reattach them to have them not visible the seams you know on the face and we do some stuff for the first we are using bullions to cut out necks so that we can attach it to the rest of the puppet and also to have some movement in the next so this is the theme of the modeling for it we constantly need to think about the real objects we have a 3D object for the modeling but we need to adopt it to the real objects that the puppeteers are creating and they're very small and we need to be mindful of for example how the neck moves and how the head moves and also the animation how is it the neck deformable, what material is it using and all that and how much gap should there be there's a lot of trial and error because we're dealing with millimeters and we need to print and see if it works and print if it works and so on this is what it looks like from the down I don't know and then we need some mechanisms to attach eyes which is also the real object so we have these eye holders that are attached to the back of the head and the black part means the real object so the eyes are put into that and they are also static I can be created in street printing but we're using the real objects and Lasha is doing this I don't know what it's called the material but the way we like the ocean stuff that is in the eyes is creating them and that's not very possible to do in 3D and also we need first of all the eye holders everything should be as seen as possible because we need to put a lot of stuff in the head we need to feed a lot of stuff that we can and also it's not very visible here but if I can get one in this part we have a gap because we are attaching of course eyelids which are also real objects they are molded plastic and we need to leave some space so that the eyelids can be attached and just go up and down to create a blinking and this is what it looks like this is our face with head without a face and eyelids we're using as also motion artists the sticks to pull down eyelids and create blinking effects and as you can see the next one I'm going to talk about is this the mechanism that we have up there which is the most important one it is magnet holders I have made some mistake on this I have balls twice so don't mind that but what we are using is that we Lasha has created this object which is attached to the face and also to the back of the head and we are using two magnets on the sides so that they stick and the one ball that just holds it tightly so that it doesn't move around but this is the most problematic part in creating this because the creating faces it's relatively easier but making them retractable that was the challenge because this is basically a rig now this is a stop motion rig and first subject was that the first problem was that the subjects the magnets are as can you imagine if the head is 40 millimeters and can you imagine how much that is how much lengths that is so it's very small it just we were at the point we just couldn't find any smaller magnets on that it just doesn't exist at least you cannot buy it so we had to adapt to whatever sizes the size objects that we could find and the biggest challenge was not attaching the faces but removing them again because the animators has specific requirements how this the stop motion animators I mean how these faces should be removed and attached again because basically for every frame and there needs to be some you know you shouldn't struggle with it so as I wrote down it should be easy to remove by intent and hard to remove by accident which means that when you want to remove it you should be able to just remove but if you're like for example shaking a puppet or if you're moving the puppet it shouldn't fall down so we need magnets that are holding real tight but also you can just tweak and it gets removed so that was the big challenge that's I think the something that we spend most of our time on in pre-production we I mean me and Lasha but Lasha mostly so because that was a big challenge we have this there's two images missing but it's okay so it shouldn't you shouldn't have to tilt the face when you remove it because on the head we have attached hair and if you remove it by this this motion that you might mess with the hairline we don't want that you might also mess with the eyelids we also don't want that you also shouldn't have to wiggle the object to remove it because it can shake the puppet and shake the head so that if you like if you're doing the facial animation while the head is in a motion which is most of the time when we do this and you want to remove the face and you're shaking this object to you know remove it all the positioning that you worked on in a previous frame for the head is lost and you have to basically resculpt the pose and that's a terrible thing so they should be removed there should be a picture here but they should be removed with the forward gesture she'll be able to just hold it and remove and also there's some things that you shouldn't have to touch the head back of the head that's the less you touch it the better so it should be removable very easily and you also shouldn't have to use force because if you use force the puppet shakes again and just working on this was very very intricate and this is what our studio looks like all the time we have these piles of faces from testing and it was basically for a couple of months Lasha was adjusting something by one millimeter or just one degree in angle and just printing again it doesn't work again it doesn't work again it's just a linear it's just something that you cannot plan and you cannot learn from us basically because based on puppets you need based on each puppet and each mechanism you have each size of puppet you have and the style of course you need to create your own rig and mechanism and this is not one time thing also because now it's getting easier because we know what our puppets need to do and Lasha has some expertise on that already but for each puppet because they have different head structures and the size and everything and these rigs have to be created again and again so that is the most time consuming part before the animation starts so now for the animation this is what I do so one time one of the directors of the film approached me and said so we created this object we were planning this project to have replacement parts and everything and we we want to animate them now and we don't know anyone who can animate and whoever we found asked us to export the object so that they could animate in Maya and they couldn't do it and it turns out problem was the multires modifier but they couldn't do it and she asked me if I know how to export to Maya because I know Blender and I said I don't but I know how to animate in Blender and I got the job and ever since I've been working on this it's been nine months now that we're working on this project and this is what I do I'm going to talk about first how it differs from the regular animation that you are all used to I'm guessing and then I'm going to talk about the specific challenges that we had in this project and in the stop motion workflow that you might have and how we solved them so the first how it differs what is unique about animating for stop motion so we have basically two traditional 3D animation pipelines one is just regular city animations CG animation it is when you're just creating film in 3D that means you have ability to choose whatever frame rate focal length the angle for each shot you can create lighting on your own maybe not you specifically as an animator but you know the city you can and you have the VFX animation in which you do not create any of those because it is already provided to you you have information about frame rate the focal length and the camera sensor and everything and lighting already exists in use HDRI basically you have two workflows one which you create the footage and one in which you already have the footage and the stop motion animation leaves in the weird middle in which you're working on real footage but it doesn't exist yet it's going to exist it's going to be a real footage but it doesn't yet exist so I have written that you must anticipate those things and what that means is that you mustn't when you're animating you mustn't think that what you're seeing is what you're going to get you must anticipate that the photographers on the set and animators on the set directors are going to have some decision making choices to make some limitations and all of this is dependent on a lot of things in stop motion it's not very straightforward you cannot just decide what focal length you're going to use because you have some weird limitations for example this is one shot this is the right one is what it looks like I know you know one we should but this the left one is what it looks like on the set we are dealing with very small puppets and we have to do weird things to make them recordable especially in this shot what was the problem is that we wanted to have depths of field very hard depths of field and have this bokeh and everything and the cameras the photography cameras to be able to detect bokeh and have the focal lengths on that scale you need to really get close to the object and sometimes it creates a problem so for example how does the animator go in and to remove the face and attach the face so things like that it just pops up always and we have to be mindful of that and we have to pre-plan it we have to know what we are shooting exactly down to the camera position down to lighting positions down to where the animator goes everything because we have to and me too the animator has to animate according to that information because for example in this case this is a lighting case we have in this particular shot lighting that comes from above and this is the case for most of our shots that we have outside because the film is set during the night and most of the time the lighting comes from above which is from the street lamps and the moon and what happens is that the eyes and eyebrow line is very visible and lit all the time but the mouth is hidden so if I don't have that information for example in this shot I have to create annoyed face she heard something from the side and she has to look annoyed and if I don't have any of this information I don't have information which side she's looking and how is the lighting setup in this scene what focal length we're using how far we are from the object I might animate something that has like little mouth movement but we print it and it turns out that it's not visible because first I animated the leap that is on the second the other side and we can see it it might be that I didn't put enough weight on browse and the mouth is shadow so it's not very visible and turns out I should have animated a browse more weight of the animation weight of the emotion she'll have been put more into the browse and it has to be it can be that I the motion of the leaps for example or browse even or anything is so small that from this distance it's not visible so if I know that this shot is like like a lit like this and we have shadows like this we have focal length and the distance like this and I know she's looking on that side what I will do is I will animate most of the expression on the side that is visible to the camera I will put the weight of the animation to the browse and for example we have this thing that she's she's constantly lifting one of her eyebrows it's quirks of the character and I know that if she's looking on this side I need to animate the brow movement on the brow that is going to look away so that it creates a direction it creates a sense of direction towards the towards where she's looking so this information has to come to me and I have to coordinate with directors and photographers and animators and have all this information before I start working because that was the case when we first start doing that I will just create something we'll print it and we'll just animate and we'll look at the shot and see how emotion is visible at all we can't see anything so what we do is we do extensive planning this is the case for sub motion animation usually but in this case it's more amplified we are basically creating each shot in every medium that we can we have the animatics we have we should reference so that we know what camera we want to use and what focal length we want to use and of course this is for the animation of the sub motion animators as well we also do blockings a lot we have photographers days before shooting sometimes even earlier than that we have them just go and see each shot how it is created so that we know where the camera can sit because we can't just put anywhere we need to have certain locations locked down we need to have certain lighting locations locked down so that we can know that okay we want this lighting is it possible and if we know it is possible and if we know that okay we are doing this then I will start doing what I do and the second limitation we have is mesh deform limitations we cannot do squash and stretches as much as in the regular 3D animation because if for example I stretch face a bit too much now it doesn't attach to the back of the head anymore because we have a mechanism there and we'll create this gap and the magnets will be misaligned they will not click and we have to be more restricted to the types of animations we do we have to be more conservative in that case and this is again the same case like if I stretch it you can see maybe not too well but you can see this is stretched the magnets will basically might fall off or not even fit if the face is stretched so we have to freeze those parts that we know are part of the rig and need to be attached to the face and animate only what is left and it makes the animations more conservative but in our case I feel like it works because the film is as I said at the start of the 20th century it's going to be around 15 minute film I don't know if you said that so it is about love it's about poetry it's a dramatic film and it's about literary people it's about historical events so we do not want squash and stretches that much anyway we do have a character which is a young girl and she's in love so we want to have naivety and playfulness of young person in love so we need to find the balance of being conservative and more restricted in our facial movements but also have some life to it and this limitation is kind of not help us but we work around it and again with the eyes if I stretch eyes too much eye sockets we will have this massive gap we will have the back of the head visible basically and eyelids might not be perfectly aligned they might fall down which is the last thing we want happen sometimes but what I do is I animate with shape kiss this is the method I chose I chose it for I guess the same reasons why anybody chooses to do facial animation and shape kiss but they are more amplified the reasons in stop motion animation because this is the case where you will want to have hand crafted animations to have each face sculpted to not have any repetitions to have them you know I want them to feel like they are sculpted they are made by hand and they are basically and I don't want them to be mechanical so each time she smiles it's just one character I'm showcasing with each time a character smiles or something I want it to be crafted from the scratch to have diversity I don't know how to say it in English but to have that diversity in emotions and I'm using face sets to freeze out certain areas as you can see the orange line is the part that just should not move that is part that holds the rig at the back and I'm also freezing eyelids and the rest of the face I just have split into shape keys so that I can use auto masking and decide if I want to animate with face sets or not you know sometimes when I want to open the mouse for example the easiest way to create openings I found out was with the shape keys and with elastic deform brush that is my favorite brush in sculpt mode and it is I don't know it's great it just looks like how it should feel so I'm using it to I think the only brush I use during the animation so now the challenges or the challenges that were specific to stop motion and how we overcome them the first one is the most basic one you're going to have how are you going to convert the animation to separate objects so that you can print them and this is something that we talked about even at the first meeting that I had with the studio that was nine months ago I think so it's very easy I have this animation which is more than four frames I'm lying here but it is a shopped animation and I want to have four faces from that so that I can print four faces and animator can just attach on frame one object one and frame two objects two and etc how do I do that well it's pretty straightforward if you think about this because once you have shape keys what you need to do is you need to duplicate the object apply the shape keys on the certain frame and it works like it applies whatever is sculpted on that frame to the mesh it's kind of works like if you don't know close simulations because it is internally also modifier I think shape keys so and I need to move it so at the time when I was starting this project I was getting into Blender Python I was new to it and I was very excited to be able to use it in the actual production to create something actually useful for me because I knew like from testing and even before that that these doing this is not a good workflow for example the first reason is that for example you might have 15 faces and doing these operations on 15 frames that is also one that is already 115 clicks 50 clicks and the second reason is that is very unstable you just miss click one arrow and you just jump on the some frame you don't know what you're doing and once you duplicate this object there's no way to know which object is what you created because they all look the same when you just do the whole interpolations of separate objects so I was using Python and I want to talk about how I approached Python Blender Python because it's going to be a love letter to Blender Python I'm obsessed with it and I create every time I see that I have something that needs three clicks for example I realize that it can be a one click because the way Python works is that also I need to do so for all friends but what I realized that at first I was doing notes and then I moved into Python and so I realized how similar it is people are talking about how like how notes are visual scripting right but for me because I started later the scripting felt like textual note editing and I realized that each line in Python because of how well written it is and how well documented it is each step you take is each line in Python so you can literally just write down all your steps and just write the operation for that in Python language and the number of operations you do is a number of lines in Python you also have like variables which is like user inputs functions which is like node groups it made a lot of sense to me and I started doing all of these now for duplicating you can just do object copy apply shape because it's shape you remove and move with location which is old location plus whatever number I input like seven meters and I'm going to do this for all frames which is for framing timeline which you don't know is the hidden word there for every frame in timeline but now I have 150 clicks shrink down to one but when you go on and on and the more and more operations appear than you need to do we're talking about now thousands of clicks that I just don't need to do and now it's more safe but this is the easy part now what was the challenging and when the real challenge came is how we define the variables now we have two variables one is object OBJ which is simple it's whatever I have selected but the second one is a timeline it doesn't show but timeline should be circled imagine it is what is a timeline so I need to define what is a timeline so which frames should be converted into animation this is what the operator looks like in the invoke window but the first one the easiest one is I have a frame range I specify from this frame to that frame but that was not enough I needed more and more and more operations as I went along and had more and more challenges the second one is on every ends frame now this I added because at first we were thinking about doing the animation on 24 frames but on 2s instead of doing the 12 frame animation which now with the 12 frame but at the start of the project we were doing it by 24 so I needed the ability to not print every not sorry convert every object into skip every frame and I added that feature and I also added more important keyframes only which is it's an easy name but it says what it does it just only creates objects for on the frames that you have keyframes which gives the ability to just manually just mark which frames you want to be a new object and specifically for this I created the the first operator actually that I actually wrote down in python was baked shape key action and I called the entire addon baked shape keys because it was the first one so I realized that when I was working and I created an animation and now I need to bake this animation because I wanted to have an animation on certain frames and turns out there's no bake operator I thought it was I animated with that in mind but I went into the menu and it's not it's just for bake action and the shape keys are not using action shape keys are using shape key action so there was no baking operation for that so I created my own because what is a baking it's just inserting frame for putting keyframe on every frame basically and turns out there is an operator for it called a sample something sample keyframes I think or something but I didn't know that at the time so I wrote my own but I added more features to it so that I have more control over which frames get keyframes so that I can use that first operate that over at the first just to keyframe the certain faces that I want to be printed out and then use this operator with the keyframes only selected and the last one and the most complicated one is delete duplicates which gives me the ability to escape unnecessary frames so what is an unnecessary frame and these are all are additive they do not replace each other more you click and more you change the more operations happen so this is the animation I have the same animation now it's seven frames and I have seven objects from that but you can see that when you look at the shape key values that object five and six are exactly similar they have the exact same shape key values there is no visual visible difference between them so I don't want to print them because that is a time wasted there is a resource wasted and more importantly what is wasted is the time Lasha has to paint those faces because each of the face that we print has to be painted by hand the faces and the eyebrows and leaps and the mustaches if they have them everything has to be painted by hand so the less objects we have to print for painting better for us so I need a way to detect those duplicates and avoid printing because if I have object five I will just leave that on the frame six and just animate with that so I can tell the sub motion animator that okay on frame five and six you can use the same object and the save all the time also from removing it so how do you detect that well first thing you can do is you can detect by you can write the script that checks the shape key values and compares it to the shape key values on the previous frame so if there is a similarity it just doesn't print anymore and that worked for a while but limitation of that is that it only can check the previous frame now you can write the script that checks for all the frames in the timeline you give but I wanted something different I wanted something more full proof and more versatile so I created a naming commission because for example in this case what happens if instead of objects five and six now objects four and six are similar you know we cannot just detect it with the regular script so I created the naming convention so the naming convention how it works is that each time the operator is ran and the objects are created the operator creates unique names for each object based on the shape key values basically I wrote some lines to convert the shape key values into a string which you don't know means text and then it gives this text as a name for the meshes and now the operator can detect that okay the object six has the same exact name as the object four so it does not print object six anymore it gives me the line in the console that says that you can use object four on frame six and the ability that it gives me also because it is based on the naming convention and not the script that it checks is that for example if I animated the first row which is more darkened if I animated the first row on like first frame so there should be numbers there but it's not showing I created the animation like two days ago that was on the frames like from one to seven or eight and now it's two days past and I'm working on this animation and it's different animation it's different emotion different expression but because they are using the naming conventions and each mesh is named in a certain way based on shape key values when I run this script it will tell me that okay on the last object you have the same exact shape keys as you had two days ago on object one so it checks not only in the timeline checks in the entire blend file so I can detect any duplicates that might exist and avoid printing them and that way it's a win for us and I can also print out those names and take them into a file in the Excel file manually check them and everything now this is a short one this is the how do we make shape keys non-distractable because I work the shape keys a lot I think this year I work the shape keys more than any person on this planet that's all I do I just do shape keys and there's some things that I want from it I want it to be faster I want it to be more non-distractive I want less of course which is messy and I want to be easier to take risks so I created some operators to help me with that the first one is split shape keys now what it does is that sometimes for example I create what I do is what is better is that each part of the face has its own animation its own shape key like mouth and the brows but in this one I accidentally sculpted brows on the mouth shape key and I don't want that because maybe the director tells me that they don't go together well maybe they need to be offset or something so I found this script that can split shape keys based on the vertex groups you need to create vertex groups you need to name them certain way and then it just like do something and I thought I'm not going to do that that's too many clicks I know it makes it sound like that I don't like clicking I'm lazy so I created a method I thought like Python can create shape key a vertex group Python can name them so why do I do that so I created the operator that just you click if I click why am I oh yeah so you you select the part of the face that you want to be split and you just press the button and now we have the same exact same exact animation but now one part of the face is on the one shape key and the other on the other the second one well it's not going on the next frame sorry it's stuck okay the next one is duplicate shape keys now this is easy we can all get what it means but now it's a challenge there because when I first use those operators I realized that okay I thought okay I duplicated shape key so I'm going to use the backup I can sculpt on that and be destructive and if something happens I'll just use the backup and I did that and turns out I lost the entire animation because I didn't just copy the shape key because that's impossible I created a new one and the entire animation was gone so I had to write a script that just copies the entire animation down to the last tangent position and also what was very important for me is that the shape keys are in the same place as the original one so that I don't have to scroll because I have a lot of shape keys I have dozens of them so if I have to scroll down and click this horrible arrow button to go up I don't want that and it also copies the shape keys and it creates the perfect copy for every operator that I use and the last one is a merge shape keys which is the opposite of the split shape keys I'm going to contradict myself and say that sometimes you want entire animation on one shape key because you can see that my f-curves are a mess because I moved one shape key, the keyframe and I come back and it just messes up so I can just merge those shape keys together and now I have one f-curve and animating with one f-curve is very easy I'm usually using this operator when I'm done with sculpting I know that I have final sculpts to create easier animating process to create easings and use that and the last part is post-processing or as I call it, a key mesh phase sorry, a key mesh phase so after I'm done animating there is a workflow that I can take up to my entire work day which I need to prepare each of the phases that I created for printing, I need to do three bullions on them I need to cut the face, cut the neck I need to attach the magnet holders and we can't do that beforehand because we don't want them affected by multires, we don't want to when you pull down the neck we don't want the hole to shrink and stuff and I also need to draw the number on the back of the hat so that when we print out we know which face is for which frame, which object is for which frame and the problem is that not only do I have to do that this is what my scenes look like sometimes I might have 49 faces which can go up to 20 million faces and polygons it's a mess, it gets very slow and it slows me down a lot and I need to manually move my camera one by one to do these operations and the bigger problem is that you're using solidify modifier to make those faces non-manifold and the solidify just basically extrudes along normals so when we have on the lips the lower lips normals look up and the upper lips normals look down when they are extruded, sometimes those extrusions go inside each other and create those artifacts that you can see here and I need to manually clean them the biggest problem was that when I'm cleaning them and I'm smoothing out on the faces I don't have ability to preview anymore the previous frame and the next frame because there's no animation anymore they're separate objects and I needed them to be animation again to have the ability to not just eyeball it and just move the camera side okay, does it look same? so I needed to keep the animation somehow and I realized what was the solution so KeyMesh is, if you don't know, one of the most iconic blender addons it's created by Pablo de Barro by Daniel Martinez Lara they are both associated with Blender, they worked here and Daniel still works here so it's an addon that is used for creating stop-motion animations inside Blender and I used that addon for my other stuff, my personal projects mostly and I decided to upgrade that and I called it KeyMesh 2 just to differentiate, it's the same addon but I rewrote it, I added some new features and I knew that it was bothering me it was bothering me for a while that I had two addons that deal with the stop-motion animation and I cannot use them both because one is for the 3D stop-motion and one is for the real stop-motion and how do I bridge that, it was a personal challenge for me how do I use the KeyMesh into my workflow and one day it was all sparkles in the head and I realized how I can use it if you don't know how KeyMesh works internally basically in Blender you have one mesh data associated with one object, each object is basically a frame, empty frame that holds the data that you give in the mesh tab, what KeyMesh does is that KeyMesh gives objects the ability to have as many meshes as you want and you can keyframe each mesh, on frame 6 for example you can have frame 10, you can have code, you can have as many meshes as you want and when it scrubs through the timeline you create this illusion of motion and I realize that I am already creating many many meshes, I'm creating 49 meshes on each time and I know on which frames they should be so what I just need to do is I just need to assign them to one object instead of 49 different objects and if I just tell them on which frame, which phase should go when I scrub through the timeline I will have the animation and with the KeyMesh it's very easy because KeyMesh uses custom properties, it's that easy so I just created an operator that assigns each of this mesh to each of them as a custom property and assigns them under one object and you can see in this case I have this shape key animation and I press shape key to KeyMesh operator I specify the frame range that I want and I think there's a waiting time here, there's a couple of seconds you have to wait when you do that I should have cut out this part but yeah now all the shape keys are gone because now they are applied there are separate meshes but I have this new UI that has all the meshes and I have the same exact animation nothing's changed but now they are separate meshes and that gives me ability to work on each of them separately without affecting previous or next frames and know that I can still have the animation and I can still print them as much as I want this is what the UI looks like, the top part is just for regular animation you can manually create keyframes you want and new meshes and I have this frame picker which I created which mesh I have on frame with this icon and if I use the dark end icon the pin icon I can just append those append what? append ok append the meshes on the certain frames and because of that I also have numbers of time used because each mesh can be used multiple times during the timeline so basically I created the way to instance frames this is a thing that was inspired by tune boom harmony and it exists into the animations now it's going to exist I know in grease pencil 2 so back again with the duplicates problem I have this duplicates on frame 4 and 6 how do I tackle that? well basically because I have the naming convention a shape key detects that the frame 6 looks exactly the same as frame 4 because they have the same name it will instead of creating a new mesh it will assign the mesh that was on the frame 4 on the frame 6 so I have instances now and if I decide for some reason to sculpt and elongate this face for some reason now on frame 4 and frame 6 they both get changed so now I get preview of the animation I can scrub to the animation I can keep whatever I want and also I do not have to work twice as much because on the two or more frames and this is how I use in a workflow it's one minute video sorry about that but the first I'm using the strings I have in the UIs so that I can name each mesh whatever I want it's not a mesh it uses custom string properties different but so in this case I'm just scribbling numbers on them and I can go through the entire timeline and I just have my camera in a position to go on the next frame I sculpt I go to next frame I sculpt I go next frame I sculpt and when I play the animation back which will happen soon it will play the animation with all those scribbles and I do this with the bullions too you can see I do this with the bullions on three of the bullions and if you use bull tool and apply the modifier you can just do this for each mesh for each frame and I can also now delete the bullions because they are in the same place it's much easier but I don't do that I don't trust bullions that much and creating separate objects from them is easy it's just the same operator I use the same slide here basically I realize once you have two separate objects of the operator down all you need to do is change the naming conventions because instead of like using mesh names you can use whatever naming conventions you want and it will just work the same for example if I ever decide to use rigged animation I can create naming conventions based on bone rotations and locations I can create one naming convention with the geometry nodes modifiers and their values and it's all the same it works all the same so I create objects and I send them to Lasha he prints them he does all the scaffolding and stuff in the software I don't know how it works and then we paint them there should be a video but it doesn't play it's okay we paint them each by hand this is Lasha by the way and we create stencils with the papers and masks so that we can isolate brows we can isolate the mustaches and leaps on each object and we paint them well not I say we and I don't mean me specifically but we now because we are printing so much faces and we have so much work to do we get new people and we are expanding we need more people who paint we need more people who print and we now have a team that does this job and for organization I shamelessly stole the idea from Pinocchio from that they use pizza boxes to store the faces and in Georgia yeah in Georgia we like pizza but we have our pizza shaped food one of them is called Lobbiani which we love a lot and we order it all the time on the set so we just use its boxes and we organize them and the last one is that I create spreadsheets so that I can keep information like on which frame what is has to be used what is hold what is a transitional frame I name them all and everything of scenes as shots everything and I just give this information to the stop motion animators and they create this once again so this is it if you have any questions you can ask me now or you can just admire the work that's that's what I had to say so if there's no questions I didn't see any okay okay yeah okay so the animation development part is on board that's good to hear so there was a question back right yeah well actually no that's a good question I don't because most of the time when the shapes are used it's when I have like going into the motion and going out of the motion for example the character smokes a lot and when she's exhaling and leaps go in you know then the certain shape and then goes back to the regular shape it's just the same frames you know it's just going in and going out so if I change going in it doesn't matter going out it still looks the same so I don't have that problem so I've never had that problem yet yeah is there anymore okay no we can use remesh because we need to have multires on it and we cannot get rid of it yeah okay is that it I will we're out of time we need to prepare this for the next presenter so thank you very much this is our information so if you want to get in touch if you want to get in touch with studio you can do if you're studio especially we would like to have studio friends so contact us or if you have any questions if you have any follow-ups you can contact me on any of these social media and don't forget to follow I need more followers thank you very much