 Okay, so welcome everyone, I'm Michael Aranander and I will be talking about Hitchhiker's guide to Blender CG animation Okay, so this is my first time in Blender conference, and I'm so happy it's Amazing to be here and meet all the people that love Blender. So yeah Okay, so first of all, why did I name this talk this? Well to know that you first need to know like who who I am So I will first of all tell a little bit a short a brief story about myself here So I've been using Blender like it's a hobby for eight years roughly and I've always had a passion for like a huge passion for Warhammer 40,000. I don't know if anyone knows what that is, but Yeah, this basically means that whenever I did something in Blender I just wanted to do anything that I thought was fun to do So therefore I made these Warhammer figures. It's like small tiny figures and gave them life basically We were using Blender So then if we now Fast forward seven years then yeah some stuff I released looks like this Those are my figures now that have become 3D molds and Animated so I'm super happy to see this and be able to share like my passion for something in this way And but then the thing is that I've always had this as a hobby Except for recently where yeah but and anyway That means that I didn't have much time To work on this very often But still it's like I had basically kind of like a maximum. I calculated this I had a maximum of eight hours per week to work on Blender stuff and I still wanted to like produce very often or be able to at least like every year finish some big project and Make something cool and they also felt the like I Wanted to make something better for every iteration or for every time I uploaded something and that meant that I had to like Try and reach any limit I could find so Yeah, other than just Making my videos or like yeah, I'm actually a full-time student other than being this Taumich So that's what that's what like eats up a lot of my time But it's also like many other things like Actually teach Blender each even for an association in my university KTH University by the way and Oh, yeah, also for the Swedish Museum of Technology in Stockholm. So Yeah, I'm very excited to be doing a lot of things with Blender So yeah, now that you know that basically I had a very like tight time constraint Then you can kind of understand perhaps that I had to like try and find these short cuts So I had to like basically find some some yeah fast route to the target like yeah, jump on to some Some spaceship and the flying to space Yeah, so here I am all alone Yeah, but I have a dream. I want to make these animations So what I did was I started off by Yeah, I mean, let's say that I wanted to make this thing here that we see It requires a lot. It has lots of details and I Mean it requires many different workflows and it even has cloth simulation. It's like many things Now since yeah during these many years that I've been using Blender I've been like slowly learning how to use all of like the tricks needed to like Yeah, the slow way basically to learn one thing at a time but If I would like with my current experience for example try and maybe figure out How long would it take? to create this Then perhaps it would look like this if I was a bit optimistic Eight hours. Okay, so how would that go four hours with assets? Is that possible? Well, I mean if you really do it as fast as you can and try to hide as many details as possible then maybe you can do both the character and the gun there and The lightning and all of that within eight hours and this wouldn't be one week worth for me but As it turns out, it's good to plan ahead so I Yeah, I would at least like double the time probably Somewhere around doubling it to two weeks worth of work Although What the actual time to make this would take is like about yeah, 30s. Yeah, it's like one month so it Takes a lot longer than it might like then then it might appear Even for someone who's very fast at making things But yeah, it's also a matter of how well you want it to be like and how many times in the future want to reuse this And since this character is like something I wanted to be able to reuse many times I tried to make it as good as possible kind of now There are many tricks that I learned during my years working with projects I kind of uploaded like every every year I uploaded one big like video at least and That meant that I had many tries to like try and find out which method works very well How can I organize myself for is there any workflow that I should be using in order to? achieve my end goal as fast as possible and Yeah, I basically found a lot of different things that you can use to cut cut down the time it takes to create anything But I chose three things that that that I think are especially good to use and that I want to highlight especially now they these are also like very Very good for a blender 2.8 especially But they are basically okay, so how to work smart? If you can keep things easy to iterate Then you can very easily both adjust something and create new things even That are similar to each other Another thing is linked to scale and I will go into that and availability. Yeah, but let's start with this iteration process so Something I started using a lot was that I tried I realized that I had to try and cut down on Cut away time that Would be repetitive otherwise are redundant So I tried this to look up a lot of methods to work non-destructively And so yeah, so in the end I decided that okay my pipeline or my workflow should basically be that If I haven't created something like the thing I want before then I should Keep it as non-destructive for as long as possible up until Some point where I figure that okay now It's probably not worth it to keep it non-destructive Yeah, and that's also partly with tax modify limitations and such But at that point I save this last version And store it in some way that it's easy for me to know that okay if I need this again Then I know where to find it But and then what I do is that I just finish the thing I actually was gonna do make out of this And in that way I can produce Various types of buildings pretty fast or various types of other things and one example That want to show is I made the some kind of pipe library Or yeah, I actually have many pipe libraries, but this is one of them that Has like many straight lines and it was supposed to be like some rusty old Scene or assets for some rusty things and then I what I did was I had like a long pipe with many like segments and made even many versions of this pipe and I didn't need a pipe that was that long so instead. I just cut cut off with booleans the pipes hife And then just changed the like which part of the pipes that are gonna be cut off and so I like produced Like 15 of them. I think it is in a matter of seconds And then I did like yeah, I basically tried to iterate things so I did like some railing and Try to destroy it a little here and there for variation And yeah, it went very fast to make things this way and I also kept the old files So in case I want a larger library or want something similar then I know exactly where this file is saved and You can see in the outline or here that in the end in the bottom. There is a creation Thing that the creation folder that is unticked and in there I have like all the things I did for creating each of these assets very rapidly So by having it easy to iterate then you can get like Variation and create things very quickly. I will actually come back to this later, but Let's for now go to link to scale maybe my favorite Topic and we have actually heard a lot about the linking in the various other presentations here So that must mean that it's very important, right? So what do I want to mean with link to scale? Yeah, well you can see on this picture here that Basically, I can use for example different furniture or different sub assets that I have created from like a resource Library or yes, that I have maybe made in the past and then I can combine these to create some Larger asset and then by using linking it means that I'm just using references to these resources So it's both very easy to replace things. It's easy to like manage Many large amounts of things and you can also like make different iterations also pretty quickly with us And you also don't even need to have finished all your assets this way You can create them or make proxies quickly and then finish them later so but then after you have finished one building you can go further and Make you like the multiple ones that you then combine into a big city So in this way, you are kind of scaling everything up You're going from the base resources and then just building on top of that With like with this library hierarchy Up until you reach the point that you feel this is what I want or this is exactly what I need And that goes very fast and to make just because of the iterations and the ability to Yeah to link everything in and replace them so easily So Yeah, we can kind of see a little bit how the library structure is looking right here and because there is this big Folder that is called library if you look at the top of the outliner kind of second to the top and What I basically use this big folder for is just to import it for example to some File where I want to use these pieces And then you can start them play around with it and place them wherever you want So it's kind of like to easily like package it and send it over and What something I did with this them was to I don't know if it's too dark, but I made some kind of sci-fi scene that is like there I think it has like 200 of these pipes and floors and railings and everything and The file size is still like very small. It's like almost an empty blend file So I noticed that also by using the linking you also save a lot of storage room and Also, you can then have like Much you can have like it goes so fast to save for example that you can have out to save on like much much more often and This yeah, this helps with speed and in many ways also performance if you hide some stuff then Yeah, then it will basically not be there Now, yeah, that's file size there now here comes an example that was That turns out very crazy, but I started with like building some walls here and Wanted to like see okay, let's make some kind of rock Yeah Super crazy city with like almost gothic architecture So yeah, I iterated these walls in some ways to create these kinds of things everything is linked in Then I wanted to build a city with these so yeah, here you can see Very it went very quickly to make these just because I Yeah, I'm not showing the exactly middle step there but I combined these buildings into like a building complex and then I just dropped many of them onto the landscape and Yeah, you can see that it's became like massive And I mean it didn't even lag on On even my small computer that I'm presenting this on so even though it had like I Think it was five million polygons at this point It started to lag if I looked at everything though, but yeah, but it's because it has to render stuff So but as soon as you don't look at it or hide stuff then it's like Yeah, no performance issue, but here comes the question then can we do more and I was like hmm. This seems powerful. Can I do something really crazy with this? And I did I recruited an army so here is a soldier that I made for my movies and Yeah, I recruited I think it's 200 of them 300 of them here and Actually, each of them I looked it up and it was like in total 60,000 polygons, I think per soldier and each of them has like ice strands like hair particles and also I actually modeled my hand Just because I didn't have a hand model and I was like, yeah, I want to try and model it Just for experience and then I all of them basically have my left hand them So I think there's like 400 left hands Now I still Found that okay It's lagging a little bit or if I move around my camera and such all of these soldiers are also rigged and Have like cycles walk cycles and such But as soon as I pressed render then it went super fast still to render it was like almost No difference when I scaled things up So I actually used this. I don't know if it yeah, you can see it Okay, I used this in one of my latest like releases trailers for one of my series and Then I added like 2,000 soldiers So and the actual thing that took the longest render was not the soldiers It was the volumetrics that cost so many Small particles, but then yeah, I tried using like the a id noisers and all of that But yeah, it was the noise that was the most annoying in the scene not the actual amount of soldiers Also to create this I couldn't display all of them and move them around because that would lag way too much Instead what I did was since every like block of soldiers was linked from a collection I just replaced it with or like I moved out the soldiers from the collection and added a cube with the exact same size And then I just in this file them Managed like where should I walk well? They should walk over here and I knew the speed also because I like the pay the speed of their footsteps and such So I just added these or moved these blocks and then I switched it back to Soldiers and then I pressed render and I got like a full animation with these marching 2,000 soldiers But that was also not the limit. I wanted to see like how far can I reach? So I tried adding a lot more you can also see I added some of my like big robots To this scene and yeah, I think I did this at about 10 or 20,000 soldiers now I think is yeah, three billion trees is what we're at and I was like, oh why it still works to render just fine But that was not the limit either I Tried adding 100,000 soldiers And yeah, here's how it looks and With the city also and I even add some cloud system that I was playing around with them Yeah, so there you see 200,000 of my left hands Yeah, very mighty look indeed. Okay Now How long did it take to render this them? Yeah, you can kind of see this the statistics here. I have actually two 1070 Graphic cards But I rendered this in ultra HD So if you want HD and for example have one of these graphic cards, then this would take like a half an hour Which is still it's like a bit long, but it's For being like 42 billion polygons. This I think is pretty amazing still You can also see that the memory is like above five gigabytes for the VRAM than I say used But I was also using the noisier so one can like scale down the tile sizes in cycles Eve would not be able to handle this very easily But with cycles you can still render pretty impressive scenes them Using these methods if everything is linked It's mostly like variety and textures that seem to scale up the like memory Since due to everything being linked if I would duplicate this and make like the twice as many soldiers It it seemed like it would have been only going up by half a gigabyte RAM So yeah, I could probably increase it even further So if you want to create huge armies Then this is a good way or if you want to create buildings like big cities Yeah, but them Let's see then a little how This kind of these kinds of two things these two things like the iteration part and the actual link could lower Time a little from going from this one month Monstrosity to just maybe a week's worth of time So, oh, yeah before first we need to go through availability a little Yeah, I've I decided to add this point as well because I realized that both the iteration part and the linking parts required like a very well established structure For you to be able to quickly like import the things that you know that you need So what I did was Yeah, I tried to distribute all the files I have in Folders that would make so that it's kind of like the same Length to go from the top down to the resources that you want So like a balanced binary tree if you want more technical Speech for that, but yeah, you can kind of see how I did it here with yeah I realized civilizations and creatures and decorations and all that Would work very well So here is actually how long it kind of took me to make my latest trailer And this is not from entire pure scratch exactly And the reason is because I was using library You can see that this has library value here so because of I had these libraries with also Easily modifiable pieces and also linkable pieces. I was able to still like go from like Yeah, I want to make a trailer what a new trailer for my next episode And then from that point and two months forward then I had like 50 seconds of video completed Which means that I had nine scenes that I had to build and Like animate everything and such and you can see that the asset creation time is very small because of that And when most of the work that I spent was like trying to build the library further actually Because there was like cycles and such I wanted to have these animations and poses saved for later use So anything I made here is like Open for me to use in the future as well And by thinking this way then I realized that Like everything I do will add up to future stuff. It's like just as you Develop yourself as an artist. You also build your own bank of tools kind of Simultaneously even so let's look also now One year back or like one year before this one at my older projects So yeah, I had made a long like Five-minute long animation, but I feel it's like one of my favorite ones And that one took so long. It's like eight months For me it was like oh, I think by the end I almost felt a headache, but I loved it and not the headache but this animation But there I only had like a hundred hours of library value and It was not very well distributed in like save files So I didn't really know where to look for stuff that I had done in the past So it was like I spent a lot of time making new things and I even updated files that sometimes Had like a duplicate somewhere else, but they were not linked So then it was like I had to either try and import things or try and move things over or try and make remake it quickly and That was a big time sink So I realized that okay. I'm after this project. I must establish some kind of structure here And that's what I did and that's what worked very well, and that's why I also want to present it to you and how much like I value the that I Started to organize my work Yeah, because most of the time here like even walk cycles or places where I could have used walk cycles I made by hand just because I felt that if I would try and make a walk cycle Then I didn't really have know exactly how to do it So it felt like I would spend equally much time learning how to do it as it would take making it manually And yeah, it was probably not the best choice than to do it manually now in after thoughts but Yeah, it's still like a learning experience and you know And rendering as such also you can if you want to cut that Then there are so many farms Especially like community ones like this frame here. I'm pretty sure this was rendered on a community farm like sheep it I think So, yeah, there are people. I don't know. Maybe someone here even that has even helped me render one of these So, yeah, and thanks for that in in that case But yeah Other than that Yeah, these are then basically the three things I want you to carry with you After you leave this presentation room, but if you keep it easy to iterate it goes fast to make new things and Similar things. Oh, yeah, I was gonna come back to something actually Regaining that and it was that the character that he saw turning his head the robot guy He was actually made with the same base That the soldier was made with so I actually made the soldier like I just changed the armor pieces a little So it went much faster than working from scratch Yeah, easy to iterate then link to scale as you've seen examples good examples of and Then keeping things easily available those are my three main things So, yeah, thank you for Yeah, if you have any questions or anything I'm called how much unlike all social medias and I'm happy to respond to anything So it's like, yeah, yeah, just write to me. You can come and ask me talk to me whenever if you have any question Yes, thank you