 Hello. My name is Alberto Petronio. I'm a vehicle concept artist in Clouding period games, Manchester. For those who don't know me, I work on a video game called Star Citizen. It's an online multiplayer space simulator, particularly based on the design of spaceships. And today I'm going to talk to you about how much I like to talk about myself, but also about spaceships. Do you like spaceships? That's good, because I don't know how to do anything else. Just to make a little sum about where I come from for my career. I started my career in Italy. I'm Italian. I studied product design in Turin in AAD. I graduated in collaboration with a company called Pininfarina with a thesis about space living and how to make that kind of environment basically living in a thing and a little more pleasant. Right after the thesis I started working as product designer. I've been there for around four tragic years. I moved to Germany for one year, the last happy year of my life. And then I did this choice of moving to England. She's a great place. She likes political stability. Going back in 2015, I entered the design world in a quite harsh way. For those who doesn't know Pininfarina is a car company, particularly famous because they were the designers of Ferrari till 2012. And I was in the extra section. We were basically doing everything that was not a car, including yachts, planes, trains, buses, bikes, and really a lot of different kind of objects. I selected some for you to express a bit what's my path. I had a presentation ready when I got accepted to the Blender conference and then someone told me no you cannot do that because we have the rights. So I have to improvise it a bit and I use my portfolio a lot to show you what is actually what actually led me to my final pipeline and how that can change in the future. And I think this might be an opportunity to give you my opinion about why personal work is fundamental for growing as an artist. So one of my first projects was quite a big project. I think Cantieri and Rossinavi were big partners of Pininfarina. We started doing this yacht literally in the first day I arrived during my internship of six months. 100 million euros yacht for a Swiss banker, 70 meters, three swimming pools. The main bedroom is bigger than my parents house and way bigger than mine. And about this I didn't do the final renders. I merely did the design for the exteriors and it was my first approach to the nautic world, to vehicles world actually and I might have started understanding that I was liking what I was doing. Other kind of vehicles I've done was more simple. This was actually for an electric bifuror of Dutch company called Evoluzione. It was presented in 2017 in Germany. I don't remember exactly what was the event but took quite a lot of prizes. I'm very proud of that. So the major of my city presenting it was a fun way of learning how to do new things in a much different way by yacht because you have to be more subtle, more city like and was nice to see it complete and something a little different which I would have never thought I would have ever designed was actually this project for Ventus. This is a snow cannon. So in product design and vehicle design we really love the environments and that's why we shoot snow to let rich people have fun. And there was a point in my career as product designer that I was questioning a lot about the kind of products I was doing especially after we had six months working with Shell. I started having a lot of doubts about myself and the ethical meaning of what I was doing. So I started studying concept art because I wanted to move in the industry of gaming or movies and leave this very marketing driven market behind me. But before I move on I'd like to show you one last portfolio piece that is really, it makes me really proud. Pinfarina is owned by the 20% by was owned by the 20% by Elenia Aerospace, a French company that worked for AIS at the European Space Agency. We like this had the possibility to talk with Paolo Nespoli, Italian astronaut, he's a veteran, three missions into space, the oldest man ever been on the ISS and we got in contact with him and asked him if we could do a low budget, which means one person, project and give him actually a pen to bring in on the ISS. We gave it to him and to his crew, one got lost, he's still there and the other ones are in Houston. So this is quite a good memory actually. Then we move on, I went to Germany, I work with a lot of you, I miss you, I love you, et cetera. My first project was actually working on the spaceships for DLC4X4, splitting data, learning something completely different, learning Blender for the first time almost four years ago, so I'm not a veteran, unlike many of you and I think I might tell a different kind of story because I really passed through Macs, through Maya, through Cinema 4D and Rhinoceros and I must admit that Blender, despite having an amazing feature, what really has that I like is that allow me to be in a stream of consciousness while I create without too many or almost at all recently issues with UI obstacles that distance you from art. These are some of the spaceships we did in that year and we did interiors. I really liked the creative freedom that there was in this studio as right after that I moved in a AAA game. There is no possibility in a AAA company to put your dog in the Christmas card. I love my dog. Then in 2020, COVID started, I moved to England in Manchester, working for what actually was my dream, Star Citizen. I was following the project since a lot of times. I thought that was the kind of design I want to work on and I went to compromise. They didn't hire me as a concept artist. They hired me as a 3D vehicle artist and I thought it was well enough because when you grow up in a country that have the 30% unemployment, you start thinking about the job as something that might not be a right but a privilege. You take what you get and you shut up about it. After two years there, I quite thought I did an error that I should have been more proud of myself and try more. After the second, the first year was passed during this project. It was the Starlifter C2 A2. The second year was much more technical, much more bug-fixed in these areas, LODs. I wasn't really happy. After a bit of insisting, I threatened to leave unless I could have gone in a creative position and come back to where I started from. I became a concept artist. Then there was the energy crisis. Then I threatened to leave again and I became a senior concept artist. If you want to take this from this talk, is that ultimatums work? I see a divorce in my future. In case any of you might be interested, be half of the company, I can tell you we are hiring in any fields really. If you have any questions about the 3D vehicle art or the concept vehicle art, feel free to ask any question. I can make a recommendation or give you the right email. I would like to focus on what's the meaning of the personal portfolio despite work. I started doing personal portfolio with Cinema4D and Kishit. Despite Cinema4D not being as bad as I find 3D products, there was still a lot that was lacking and that I couldn't understand because I didn't try yet before 2019 with Lino that taught me everything I knew. Thank you very much. How much I was missing not having the possibility of working, for example, with an integrated render engine inside a software or all the possibility to not keep switching between software, which is despite being very cozy and comfortable is also very economical because I remember spending around 6,000 euros for Cinema4D and now I can say it didn't worth it. So I did the switch. We worked close to each other for like a month every day, at least for hours a day just to teach me Blender. And after that, I never came back on Cinema4D despite I had a degree in science and I started doing more portfolio and personal pieces with Blender. Why am I focusing so much on this? I think and I really couldn't insist more on this point that you should risk everything you can in a safe space, in a safe space, your bedroom, your home and your console. You date after during your work. Every time for motivating myself doing more personal projects, I try to do something I've never did before. I never used Blender in this case. I never made a spaceship in this one. It was 2019. This project was a disaster. It was 12 gigabytes of file that so many stupid texture that were not needed. EV crashed all the time. I had to do all the renders in cycles. So the next one was actually learning how to do something properly optimized. How to learn how to make a spaceship in the way Star Citizen did it because I wanted to go there. So I studied how they work. I tried to make a portfolio specifically for them. And using their techniques like trim sheets that doesn't require baking is not iPoly is not low poly is mid poly with weighted normals. And with this kind of textures based on our horizontal sheet, you can exceed the one the 0.1 space and go over and make virtually infinite the possibility of resolution that you have and using one set of texture for 700 spaceships as the citizen of many of them and the computer of our players have limited space. So we cannot add texture like for a movie production. It's a little example of how they works. They can be reusable. You scale it down, you scale it up depending by how close you go with the camera for a player. I tend to use box mapping for the rough map. And I try to keep it on an object. So it doesn't repeat when you repeat the asset. You just repeat the normal and maybe the opacity that comes through the previous texture. And of course, a couple of modifiers to make your life easy, let it looks like it is iPoly but it's not. This is the weighted normal modifier. In Star Citizen, the production team uses scripted tools for 3D max, which are much more complicated. This modifier in blender simply get the largest size of the faces and keep them flat and let bend the normals on the smallest faces. What happens is that let's take the glass in this case. It is just one segment but it looks iPoly and makes your life much easier. Not just in production, but also in pre-production. If my art director comes and tells me we have to redo the glass, I would like to avoid suicide that week. So maybe with less polygon, we can help each other. Plus, I can also give the file to the 3D artist and they don't have to work too much or rework too much. In case you want, I have seen a couple of years this 5.4.3. Please don't sell it on Street Raider, it happened. It is for starting only but it's a good example of what's the pipeline in a new generation hard surface production company. So as I said before, let's try to learn different things. Let's try to improve the knowledge of the tools I have. It was locked down, it was 2020. First lockdown in Covid, I was new in England, new nobody. Let's try Grisp pencil. I had a beautiful tutorial from Yama Yurabayev, actually presented a couple of years in 2018, I believe, here and watched that tutorial a lot and I thought, this can be fun. Grisp pencil has a lot of strokes, a lot of points, it creates a lot of noises if used in this way. So I thought it might have been perfect for simulating details whether I don't even think about it. There is this video on YouTube, I will not show it all because it's six minutes and sorry, the camera is very wobbly but gives an idea of how versatile a tool can be and how I try to use different tools that are not done for what I am doing to expand and maybe I will not use it anymore but I will still expand the knowledge I have in blender and that time that might be, I will be on my laptop working from home and I don't have any science of Photoshop but I still want to make a couple of sketches but I don't have a pen. I can do that here. So I did all the strokes in 2D and then used the edit mode to just reset them and move them in space, having something completely, I mean you cannot do that in other ways in blender, maybe you can do that in gravity sketch but it's a good way of exploring different things. It's very confusing, I'm aware. Thanks. Going over this. After I did that, I actually haven't done much for a year because I was absolutely burned out by England, by the job, by the lockdown, so it took me a while but when I do personal project to motivate myself, I try to hit tubers with one stone. So I try to improve my portfolio because that gives you job opportunities, give you network opportunities but I also try to learn new things. So I wanted to learn animation and rigging, basic rigging because doing a lot of Macs is useful. Sometimes to show your art director, your designers, how that specific things might work in a video game as we have a lot of interactions in this kind of stuff is good to be aware of all the tools that can help you. Plus AMD was doing a challenge at the time so I hit three burns with a stone, I have a nice portfolio project, I have learned rigging and I have a 5000 CPU, 5000 Euros CPU. So I guess I won with this artwork and then got back on it, inspired to Horizon Zero Dawn, obviously. And it was a good opportunity to learn how to do different things, different kind of shaders, something less metallic-y, something more militaristic and then of course how to make a work cycle. I don't care much about production issues like on penetrations, this is just the possibility of myself, for myself to have a tool that let me explain my superior or Chris Roberts or my art director what something can look like when it is in game. Then everything gets scrapped and the 3D artist, the expert's one, the one that knows technical stuff and can bear with loads, animations, rigging, then they will take care of it. This is just to give them an indication. If you look at the skeleton it works but I wouldn't put it on movie. And sometimes I try to expand and go in different things, 24th of February, we were, all of Europe was kind of shocked by what happened in Ukraine. So I got contacted by an association in Lviv, Leopoli. They were asking for artworks to turn into NFTs for financing the right cross in Lviv and Kiev. So mixing different techniques, 2D and 3D, I came out with this that was done in alpha day for the 3D and the other half for the other paint. It was my only NFTs ever done but it was bit outside of my comfort zone but it still was useful to have this approach because I would have used it after a lot in the concept team on such citizen. Little note, I saw this on Reddit, a lot of people saying, oh that's not a Russian tank, come on man. Someone else commenting, well that's a Nazi tank, maybe he's comparing Russians to Nazis and I really appreciate that you think I'm that smart but I just don't know anything about tanks and I'd like to keep it like this. Some experiments with 2D as I'm using it a lot recently. I use Photoshop, sometimes Grispensil, I've seen or procreate but recently I think paper is still the best way and then I moved directly in 3D so I can check everything, every dimension, every size, every interaction. So that astronaut went really well on our station and I thought let's try to do something I've learned again. Let's do props. I thought this was quite successful. I really like this project. I did the gun as well. Also to improve a smaller scale of the things that I generally did for all my life. I cannot put animations in the same time. I apologize about that. So I tried some experiment with EV. I think that EV compared to cycles work very well and very fast on small things but increasing the scale it becomes a little problematic with compensations, ZDAPT, AO not working properly. So I just try to mix both of them and I think is one of the main point of using Blender, the fact that you can use the same nose both on EV and cycles and that's probably the main reason I never downloaded Octane. And now we arrive here. This is my last project. I presented it in GFD North in Manchester. There is a full conference based on it. You can find it on Vimeo. This is the kind of scale I generally work with, the kind of scale I like to work with and this was another project done for learning something. I needed to make better and more optimized my production for concept to give to other artists. So I think that I would suggest to people that want to go in this word and want to create models from scratch is please use references. A lot of them and well chosen. I generally use Puref. I have a big folder with a lot of them. I'm talking about thousands. And then each project has its own with some from the real world and some from other artists. We have Paul Pepera on the bottom, Star Citizen, Chris Foss and Paul Schadisson. I try to not put too many, to not confuse myself, try to stick and stay true to the project and I'll just show you some some of the shots that came out of it. So some is just 3D and studio lights. Some is HDRI and other paints. Some shot to give a better sense of scale. Some to have a more spectacular spectacular view. Some more suggestive. Sorry for my accent. I'm very nervous and every time I'm nervous I become more Italian than usual. There is some shot that are interiors. This was literally trying to create a pipeline for Star Citizen to get a better pipeline for Star Citizen on my own time. And this kind of project plays back because once it is on our station people will connect it to you. We'll ask you how you do that. You will meet people and despite just having one more piece and have learned something it's really useful especially in this community to get to know more people. And after that I actually got a lot of kitbush that I would have used in the future. These are some of the engines that you've seen in the previous shots. This was the main pusher. There's a bit of gribble. Got some part from Janubert. I asked for a mention. Not like people that research my stuff on CGTrader. Try to experiment with some animations done with EVs and with cycles. As you can see by the flicker this was done with cycle. Again this is not perfect but if I have to explain my art director in Alpha Day how these things work this is enough. It's quite simple shading, quite simple camera movement. Not very good at that but I don't care if there are issues because this is not production. Got some cut out without the wings. I worked a lot on the materials on this one because I really wanted it to resemble the spacey Starship and etc etc. This was 24 hours rendering on my secondary computer and this one shows you how all the system works. There are issues so it explains how it lands. I wanted a mechanical design that looks chunky and obvious but felt like a mecca. A horrible camera movement and explaining how the storage works. So at this point I thought okay let's learn something new. Let's learn geometry notes and I always had the dream of having the possibility of creating and giving to a computer the most boring part of the spacey production, the detailing. But detailing have to follow very particular rules so it's not that easy. So let's try and for understanding what to tell to my computer I have to understand how design works and why good design works in a certain way. So I asked Alex, a dear friend of mine, that used to work in my position before. No wax, you may know him. Can I use your picture to explain what's good design is? He said yes. So let's take his truck. A lot of people copied it. That's the demonstration it is successful. It has proportions, right? It has an iconic shape, very grounded, not very alienish. It reminds absolutely of a human design, very militaristic, works well. It has primary shapes, it has secondary shapes and it has a good distribution of details. Not evenly spread it, spread it on top, spread it on the rear, there are repetitions, everything works. Alex from this point of view is amazing. There is equing of angles. The more you put the more alien or design the spaceship will look like. The less you put the more grounded and human and industrial the vehicle will look like. And then there is framing. A little thing about framing, because it's often forgotten for people that do hard surface art, is that our brain synthesize things. A four-year child see a raccoon and say that's a cat. So we see a barrel and we say it's a cylinder. But things that are man-made have industrial process that require a lot of elaboration and evings. A lot of reference is actually not a good thing. It's fundamental. There is no chance to do good design and good concept design without having good references. These might be everything. But on the left you have what our mind percept is, a barrel or a cannon. And on the right there is using proportion, equing, framing. That's our way of elaborating. It can be whatever. It can be a hatch, a syringe for a video game. It's the one that they plant. Doesn't matter. But things are more complex if they are man-made than what our mind remembers them. I did an example of some of my design. There is a lot of framing around the radar. So I don't think about these kind of things when I do my design. But I've gone back and checked. Do I have primary shapes? Yeah, I do. Secondary shapes? No problem. I put negative shapes because why not? It helps to break the silhouette. It helps to feel that design is more believable. And then of course details. I try to check if there are flows because I have to tell the computer if there are flows. And there is kind of some variation in shape. It's not an horizontal line, this shape. And the distribution of details follows those flows. So I started this a month ago. It's not amazing. But it started taking all the stuff I was using for making my blockouts. You can see some here. I went back in this file, took the main parts that I liked the most, put them together, did exactly what Daniel did yesterday, but worse. And got all the huge library of kitbash that I have. Took Simon Tom's file from the Blender Studio. And this was the result. Fine with it. Nothing wrong. It can open your mind, but I thought it might be something better. Although I needed help. Thank you. Three years ago I met Lucas. He is much smarter than me, so he helped me a lot with the math. We met in the Blender conference, by the way. And he helped me a lot, basically taking my words and understanding what I wanted to do, and actually making it work. So I have a quick video of how it progressed, and how it worked in the beginning. I can actually accelerate it a bit. There was enough variety. I wasn't quite happy with the shapes that we could obtain. I tried to put some negative shapes as well, like Simon did. Had quite fun. Give the impression that there was plates. I put details. I used the bounding box to create a negative shape to actually have a bit of framing, so everything doesn't feel like just stuck on a surface, but it wasn't... It doesn't work perfectly, but it works good enough. Oops. No, what am I doing? Sorry. Then... Oh, God. Okay. I don't have to use the arrow. Then when I was trying to do the cables that are a fundamental part of my design, I encountered the first real problems, and that was a disaster. A little note. I officially stand here using Comic Sans, because I want to demonstrate against the high price of Alvetica. Thank you. We used the volumetrics to create a rough decimator. I scattered points on the surface, and I tried the cables. Of course, this is not the result I wanted. That's confusing, not necessary. That's where Lucas came and helped me a lot. We started with using the Hedge Path Finding node. There are many ways to use it. We came out with better solution after. This was our first attempt. I was already very, very happy about it. Not going to lie. I find out that if you tree on Gula Rise, the source mesh for scattering the curves on, you actually can have a much more veiny looking kind of pipes that might be very useful for make alien ships. I don't know. It's just a node for the future. Changing the seeds, you change the way those cables spawn, but there's still nothing special at that point. I mean, for me, yes, but let's deconstruct. We said that we need primary shapes, and I selected those. The most important is the one on the right, because it gives a different flow than just unboring flat shape. We have secondary shapes that spawn on the selected primary shapes, and we have a first layer of details that also use the bounding box of their volume to create negative shapes, and some extra grebel for making just surface noise. All selected by the spaceship I already had, and after I cut them everything, we started doing more complex pipes. He's actually Lucas work. He had the idea of to avoid the overlap of pipes on each other to use the data that we take from the segments that we use as pipes and remove them from the hull and plug the next pipe from that hole with removed areas, so they always stick to each other, because you keep removing geometry from the source mesh you scatter the pipes on. I then changed a bit of things. He went visiting me in England. I went to the toilet for number one. You cannot claim I've been in the toilet for 20 minutes, and I came back and there was UV mapping. Thank you. And these are the section and how each one works. So I said that I put bounding box for creating the negative shapes. The best solution it would be to create bespoke meshes. Didn't have the time to do that. I'm sorry. Details how the three works. This is the way of scatter details on faces. It's incredibly useful if you know design, because you can use the positive Z for scattering vents, because air, warm air, goes up. You can use it for scattering the details in certain areas, depending by the bending of the corners. So we don't put flat things around cures. We put the engines on the bottom, obviously, or create in particular engines like retro thruster for the front. Wings on the side. And this disaster for the decals, which will be probably one of the first things I will rework on, when I will go back working on this project. I want water tanks or fuel tanks to have a distinction between the top and the bottom of the ship. And of course some negative shapes to create a sense of plating. One thing I mentioned earlier that I would have liked to do bespoke negative booleans for creating framing. I did it for the vehicles engines, actually. So we have the positive part and the negative part in the same object with different materials. So I used a separate geometry with a material selection. And in this way we can have framing, because I did a slightly bigger negative shapes around the engine. And every time the engine spawns actually dig around the hull and creates this kind of framing, but can be better even with the engines. Because if they spawn on the front of the ship would be quite stupid to have part of the ship that are directly exposed to the trusset fire. So the red that you can see here on the right is the negative part that with the material selection will be used as a negative boolean to remove volumes from behind the engines. Same done for the these other kind of these other kind of vital engine to create space above them for the tanks and some work on the material. This is six minutes. I am sure you don't want to see this for six minutes. So we'll go a little faster and I will drink. So what happens here is basically the material is divided in four main parts. The base color, the decal color, the edge wear and the AO. It's quite simple standard standard workflow. It's quite a standard workflow, it's nothing special. So we're in the blender conference you know how to use material shader. The projection of the decals come from above. Yesterday actually watching Daniel Stolk I think might be a great idea to have randomly selected decals and randomly selected masks to create variations. I was trying to give this crease false vibe with this big huge chest decals. Putting some some platering in different projections using a mask created multiplying a noise with boronoid texture. This was literally me making experiments and not understanding how things were not working. So still in comic sense. Again what's the power of the blender communities that I can find that information I need with a bunch of clicks. So I did this just to have a mask to avoid the sense of repetitions between two different mapped plate textures. You can choose how much how big they can be, how spreaded, how evenly etc. Then I started making colors. Try to create some variation. Also in the rough map which is very useful for giving the idea of something painted over and not made of the same material. And that's it really. So the final result of the condition of geometry now is this one. Which can be better but I'm quite satisfied with. I accelerated a bit because every reshuffle of the seeds take around five six seconds. But the possibility that this can have in the future I think is quite amazing. So far the use in the industry would be probably limited to background shapes or something more useful for starting a project and giving your art director, your lead many different shapes and studying many different things. But this is surely going somewhere. You can also shuffle the pipes, changing the seeds of anything really. And then it can be used for quick iterations. I will accelerate this one as well because otherwise we go over time. This is me making a quick scene, multiply, change the seeds, volumetrics because volumetrics are awesome. Set your scene, couple of lights, make a gradient and have fun. To this kind of scene I also put a texture in the density so it looks more less even and motion blur helps a lot. And the result is decent. With different, there was too much motion blur, sorry about that. And something more calm. But yeah, this is it. So what happened next? I'm going to have a fucking holiday. But after I would like to work on the variety both of the shader and both of the shapes and having more controls will actually be better. And eventually creating another system that allows me to model the main shape without letting it be generated and just taking care of the detail. But that's it. Thank you very much. I'm not sure am I allowed to ask for questions. Two questions, three questions. If there are questions. Yes. Especially with your background in the more minimalistic case, to make sure that these ornamental details are still like tasteful. Well, there is a huge difference between concept art or concept design and product design that is that product design tries to keep something normal, like a bottle, and making it sci-fi. While a concept designer will take something not existing, like polygons, let try to let them be believable and more consistent with the virtual world. So while in product design I would try to do everything to hide a screw in the concept design, I would actually emphasize it. So that's one of the answers. But if I may give an opinion, I hate minimalism. And there was someone behind. Okay. Well then, okay. They don't have to work together. You mean rig animation and geometry nodes? I would use the geometry nodes more in an initial phase, while rig can be very useful. Recently I'm... What can I say about that? Done land vehicles for like rigging quickly the suspensions and make short animation on like how distant is the bottom of the car from the ground, stuff like that. Well this comes more at the moment for the situation it is now as open your mind moment like when you create slightly more elaborated blockouts. In Star Citizen for what it is the props, environments and hard surface area we don't bake. We use the bent normals, the weighted normals to create. The impression it is baked but that would request too much VRAM to have baked texture for every items we have because we work in modular things. So we try to keep the bespoke texture to the least we can, especially for organic shapes like bags, leather. But if we can use trim sheets it's the best solution to optimize the workflow. You have more draw calls but you have less texture on your VRAM. Well thank you very much.