Growing as a VFX Artist

[Music] hi everybody welcome to my talk i'm very excited to be here i want to talk today about a few things firstly for beginners or those who want to start adventuring games as vfx artists i wanted to talk about my mistakes about my failures how i started and how i got here and then for veterans and really for everybody i wanted to talk about how i kept growing and staying motivated and passionate like on the first day of my job and i also have some practical examples for you what would like to achieve in this talk is to inspire curiosity in you about all wonderful things you can find and learn and how this can be applicable to your career and your job and i would also like to spark passion and love for what you do a few words about myself i'm a fx artist with 15 years of experience in games and currently i'm working at its software on doom eternal

as lead vfx artist and i have a short video from previous doom doom 2016 of my work [Music] [Music]

[Music] i'm also a big fan of tiny potato meme from emily's diary so this tiny potato will accompany us through the presentation i was born in poland country of potatoes cabbage and a lot of other good food and as a child i didn't have barbies or legos but i have had crayons and colored pencils and this is what i've been playing with i was drawing flowers horses and i loved drawing explosions i got my first pc at 17 and i spent my life savings on it and i was blown away by creative freedom in cg at the time it seemed to me limitless and i created my first website i worked on my first 2d

game and first racing 3d game back in poland and i quickly realized that to get access to all the knowledge i have to learn english and there was no online tutorials and or internet as we know it now so the books were the only source of knowledge for me and they all were in english this is one of my favorite and the first book i got for myself about vfx uh deconstructing the elements by by pete draper and to do all the tutorials in this book i have to get myself a dictionary and this is this is how i went this is how it started and there are some other books i recommend for anybody who is or want to be a vfx artist um my parents didn't wanted me to be a poor artist generally so they didn't like that idea so they sent me to a college that was miles away from my house to become a master engineer of environment protection science but commuting time i was three hours one

way and that meant six hours a day i was sitting on a train and i thought i could do so many cool things on my new computer if i only have time for that so after a year i transferred to a different college to save on time and i studied material material science instead and my parents didn't know about it because they were working shifts so they didn't realize that i've been coming home a little bit earlier and soon i started contracting and earning some money so i was thinking maybe i should drop off the school completely but i didn't and i'm glad i didn't because later on higher education proved to be invaluable for immigration everywhere i went starting with england singapore canada and united states higher education was pretty much requirement for me to get job and live in those countries additionally i convinced all my professors that i can do very good quality visualization of burning process in industrial stove and they agreed for

me this to be a big part of my master thesis which was a lot of fun to do actually um next day after my master's thesis degree i flew to england for my first first series interview for 3d generalist position and i had my demo reel but in my demo reel i didn't have any of my contract work i did in poland or even my master thesis i worked on a personal project for a while i thought that was the better piece of art of anything else at the time i had and i thought i had so much fun working on it because there was no restrictions on it nobody told me there was a deadline there was no art director art direction there was no performance limitations or strange tools so it was great and at the same time i was learning english very hard for a few months to prepare for that interview but it didn't go well it went actually quite bumpy because i got so stressed i completely froze i couldn't say word on that interview

and fortunately my demo really saved me and i still got hired and and with that company i shipped my first triple a title and from there my doors were opening everywhere for me of course it wasn't as simple as that it was a lot of stress because i was still struggling with english learning new tools proprietary tools and maya and i felt like an idiot asking everybody for questions and giving and receiving feedback was terrifying because all artists were more experienced than me and they were all better than me and the only thing now when i'm looking looking at it from perspective i know that i was stressed because i was learning and growing and that's that's the process um so first i was hired to be a 3d generalist but of all the things i most wanted to make the effects so i learned particle flow fume effects fluids in maya krakatoa thinking

particles reflow you name it and i did tons of training for filming cg because there was not much training for vfx for games at the time and i was wondering how far i can push real-time vfx to look better to look like in film to look amazing and i realized that one of the things i can i have control over regardless what set of tools i'm using is the texture especially if it's animated flipbook texture when everybody else was using real footage to create animated flipbooks i was rendering textures from realflow and film fx simulations i thought they are art directable and unique so no other game would have the same fire as game i'm working on which was great i thought they are higher quality noise free i could render easily alpha mask or any other mask and maps and it was much easier to retime if i needed them to have more or less frames to play slow

motion or faster so i had a huge success with that and then one day i got a dream job congratulations and now what so i was finally hired for a senior vfx artist for amazing triple a title i was excited happy and the world was just amazing at the time and because i got used to the pressure and fast-paced development so i was pretty happy but then with time with couple of years especially during downtime i got bored and i grew more restless and frustrated i was just thinking i can't just make particles all the time it's it's it's kind of boring i'm kind of done with it and i felt like i needed a challenge but they didn't want to quit my job so what i did i took part in some cg competitions started doing more personal projects again and i learned new skills and software and i also connected with other artists and all that combined meeting new people learning new software and starting new projects it felt exactly

like starting new job but without quitting the old one which was fantastic but then at some point i also took it too far i shipped three three triple a titles in three years that was great for my portfolio but not good for my health and passion for games i still wanted to do more but i couldn't uh looks like too much of anything even a good thing can be a bad thing this is why i wanted to talk about passion workaholics and bernard and how can they be related firstly passion is treated very seriously seriously in games it's part of the job description and i look closer at what actually means and from many definitions from the internet i put just to just for contrast something that you are strongly interested in enjoy sounds like a pleasant thing but then strong and badly controllable emotion every day all the time is probably not

the healthiest thing and from my experience there is no really bonding worry between the two and one can turn into the other at any time some better metaphors passion is a powerful driving force it can spark a fire in you that's often difficult to extinguish by sweater battle and passion is like ocean it can take you far places but can also drone you if you don't respect it and i made that one up but the point is the point is that passion can be like a double double edged sword it can help and it can hurt so how do i deal with it i'm trying to be careful how i fuel my my passion and it's like a hobby so i would invest money and time in it but i wouldn't wouldn't invest all the money all the time because that would turn into obsession and i need to know when to stop in time to always get back to it excited the next day not be tired and bored of it i use pomodoro technique to know when to

take breaks from work and i remember to work that work in entertainment industry supposed to be fun because if we don't have fun making games then players will not enjoy playing them i think back to the time when i started and what was exciting about my job back then because i feel still the same the same way about my job now and i focus on what's important of course this is much easier from manager perspective because i see now the bigger picture and i prioritize and let go and come to terms with fact that artists never finish just abundant when we run out of time i try to work smarter not harder i'm trying to use all the knowledge tools and ask others for help if this is what makes my job more efficient and part of this efficiency are procedural workflows and this is when i heard about houdini many many years ago but i started to learn houdini just couple of years ago just because it is

very hard most of you probably know this picture and think it's a great joke but it's not it's true it's hard i also had to justify spending on expensive pc and i thought about it as investment in my passion and career and honestly i wouldn't be here if i didn't get that pc couple of years ago harder though hardest one water was finding time in busy day-to-day living and i started learning houdini in evenings after work and on weekends and then i was looking for good sources of information i started by joining side effects forums art forums and various houdini facebook groups and from there i found tons of information learning resources and tutorials but the best tutorials for me are those that are easy to follow and they don't take eight hours to complete and there are a few examples and tag master and nipping john farmfield houdini wiki and people now ask me quite often what do you start what tutorials do i begin

with if i want to learn houdini and i do send them those tutorials the list from the previous slide and some of those people do come back to me after couple of months with rendered images from those tutorials showing that they follow the tutorials and that's a good start but this is not where i would stop personally when i learn i need to apply the idea from a tutorial to some different concept to a different scene or combine few ideas into another completely unique scene and that way i'm sure it's something i can take also credit for but also it's good to take the same tutorial and take it further add to it plus it change it a little bit so this is one of the examples i have here for taking a tutorials router and i started with this simple uh tutorial from intagma it takes about five ten minutes it's on and tag my patreon from the their vex 101 series and as a result you get a

lot of boxes that spin and as they spin they create a nice wavy motion so i thought how can i place it what else i can do with this so i added another layer of noise on this and that already made the pattern more more interesting and then i thought i don't want all my boxes to have the same color uh i would like some boxes to be brighter more glowy uh depending on like here depending on angle of the box and on on the velocity of spin velocity so at that point i changed the color of the boxes and i did the test render and i realized would be cool to have some additional smoke swirling around those boxes and so i sourced smoke from the floor from under the boxes and i sourced velocity from the boxes themselves and i simulated pretty standard simulation in dopes in houdini dynamic

operators and from there i did another test render and i realized that my nice glowing bright boxes are covered by dark smoke which wasn't that great so i wanted to make the smoke brighter where the boxes were brighter but i didn't want to re-simulate the whole thing so what i did i scattered bunch of points around the boxes and i transferred i made volume from attribute to get the density where the points were red and then i multiplied this new created volume with the volume from previously simulated dope and that gave me the same smoke but only in areas where boxes were bright and i renamed it to be heat that way i could shade it as a fire and the smoke from the simulation i could shade uh density from the simulation i could shade as a smoke and that way i achieved the effect i wanted and on the boxes i had the material that as i mentioned the red color was

controlling the emissive and blue color diffuse and the material was reflective refractive had wrong corners and i rendered the whole thing in 4k in redshift that was one of my first renders in redshift and i was very surprised how much harder it is rendering 4k than hd it was very interesting experience and i added crazy soundtrack by tiny cat massage and that was the final result [Music] so that was one of the first videos where we thought that the music matches with visuals very well and after that we wanted to take it further with visuals and sound integration where we wanted the visuals animation or particles to be completely autoreactive

triggered by sound so i started by looking at this tutorial by atom on vimeo he has a sphere and a box moving to a tune and he shows how to use chops channel operators course to modify them to make them usable for animation and this is the reason you get and then i also really like the technique by spencer lauders from cg workshops introduction to effects using houdini he shows how to dissolve an object into particles it's a very cool technique he shows so i wanted to combine those two and make something unique and what i changed here i used midi instead of wave files because i wanted to have access to separate nodes i wanted more nodes not just one or two but i want to have five or six

and i created separate emitter for each note and i spawned a particle from relevant emitter when node was happening and i copy a sphere to each particle and the size of the sphere was relevant to the amplitude of the node and this is how it would look like at the start and then i created second particle operator network second pop network to dissolve spheres on collision and that part is exactly like in spencer's tutorial and i'm just adding some additional velocity custom velocity for particles so they go up uh above the floor not under the floor but this is the final result [Music] so this is all going great but i had to

do some adjustments with my working working routine because it was very frustrating to walk away from a problem very late at night and my head was still spinning to try to solve problems in houdini when i supposed to be asleep long time ago and so i had to limit learning who they need to every second evening and only part-time weekends but i did save tons of time on rendering because we do have two computers at home and my husband's computer is never used at night and so we start rendering every second frame to our network drive the only problem we had was that we were drawing too much power from our office room so computers were randomly shutting down this is why this is why we had to get this really nice thick long yellow cable to draw power from another room and after that it was working perfectly fine and on the evenings when i was not

working on houdini i started painting and that actually proved to be much more relaxing activity before going to bed so where do i take the ideas from really sky is the limit anything art station instagram pinterest dreams films and i keep a list of ideas to pull from i start with a simple idea because if the idea is too complex it's less likely to project is less likely to be finished and the message is less likely to be conveyed for a viewer but i don't start with the first idea because the first idea is usually in the most boring one so i try to add to it plus it twist it change it a little bit and sometimes i will try to tell a story and of course there's a question do you need to model rig animate and make all those assets to make actually a scene especially if it's like a storytelling any answer is not always i looked what other artists do and this is one of my favorite artists renderburger you can find him on instagram and he does a lot

of cool videos and this little animation model and animation surely are procedural and it's a nice little idea you can also get free models from 3dscans.com and that's just one of the websites where you can get them this is where i got this little dragon from and yeah those the quality of those models really really good you can do a lot with them and you can also buy very very cheap models i mean sometimes five ten dollars from cg trader or unity many times they come animated rigged animated and i don't exactly know where render burger took this head model from but the point is not about the head the point is about the idea and caption which is mind-blowing and also if you want to if you know people who want to upgrade these skills as animators or modelers it's very good idea to team up

so when working on a project i implement ideas fast to see what has potential and i move to the next concept if something doesn't spark or if it's technically too hard if i can't get it done because i'm discouraged in any way or it goes too slow i might stop and the point is to keep going keep progressing and so for example this shot i made just in a couple of days because the previous idea i've been working on for weeks was just technically too hard and i might come back to it at some point when i know more about it so how do i know when the project is finished if no one tells me when is the deadline i'm trying to focus on what's important i fix only five to ten things from my first test render and i'm trying to finish one project a month if i can't deliver something clever or sophisticated in that time i render something simple and rendering does take time unfortunately and definitely teaches

patience and none of those projects ever feel ready or good enough to be shared online but i know that the perfection is the enemy of dawn so at some point i need to move on and the goal at this point is to progress and learn not to make something sophisticated of course feedback loop is part of the learning this is why it's important to share the work as meg black said the work of art is not complete until it's shared and you'll be surprised what people respond to there's nothing frustrating more frustrating than trying to solve a problem that i used to have in the past but not remembering how i did it this is why i take tons of notes and you can find my notes on github and i keep i keep all the old files and i make the same setup sometimes multiple times to remember i also make sure tools presets and hdas to recreate setups quicker and this is also

why houdini is such a great visual learning tool because it's not just easier to debug but also easier to memorize and i have here another example to show you that learning math in houdini can be a lot of fun so idea here was that i wanted to create a procedurally a circle just using attribute wrangle and points and connect them and make a circle so this is the equation that let me do it and that was cool but i was thinking what would happen if i start modifying this equation and changing things in it so i got that when i multiplied x and y position by angle i got spiral and then when i add a changing variable like a time to it and i got animation and instead of multiplying when i add to x component i got a wave and this and that and the point is that there is no such thing like math police

playing with equations and changing variables and numbers and just seeing what happens is a lot of fun so how is this all relevant for me projects in houdini branched into learning other software speed tree redshift substance mega scans github neutlino and much much more learning logic math and scripting improved my technical skills debugging and solving problems taught me patience and critical thinking i learned how to divide complex problems into smaller chunks and tons more but the good thing is it only got easier from start how growing makes me happy happier i feel that i'm becoming a better artist i like satisfaction and feedback loop i feel like i'm staying current with new technologies and techniques which is very relevant to any aaa developer i'm widening horizons it's easier for me to connect with others from different disciplines it's easier for me to talk with animators programmers designers or

lighting artists i'm finding things i like like motion design abstract cg and music videos which i would wouldn't have a chance to discover otherwise so next time you have some time see what other artists do what excites you and inspires you open your favorite software or software you're not familiar with and play with it maybe you even like it personal projects are on your rules you are an art director you deciding what to do how to do it and in what time frame nourish your passion never stop learning and help others to grow and remember tiny potato believes in you you can do the thing so i wanted to thank everyone who believed in me especially my husband sean and mark here for invitation to talk at gdc and entire team at each software for support thank you

i also supposed to remind you to fill out this survey after we finish with this presentation thank you if anybody has questions do we have time for questions five minutes okay yeah we can have some questions then anybody anybody no no i have a question do you have times that you like to design like are you do you design at night or in the morning or do you set certain parameters for yourself yeah i try to make it a habit generally it's it's it's very hard to make it like at random times i would say so lately it's been like every second evening so i would do it you know like mondays wednesdays fridays and then on the weekends for four hours a day and and generally with the ideas i don't

spend too much time thinking about them i mean i have a list of ideas but they're very rough it's just whatever i watch on the internet or movie i just put on my phone and it's just like few sentences saying what the idea would be and then when it comes to doing it actually i'm trying to implement it roughly really fast so if i'm afraid that well i don't know if i can pull it off technically i try to start with technical side and if i can't pull it off then yeah i either have to ask somebody for help or drop it and try it later and if i'm worried about well can i pull it off how it would look like actually this is where i would start then i'll just you know create camera whatever i need to see in the camera and see like is this actually working from those angles are people going to be able to understand what i'm trying to show here and uh yeah again like if it doesn't work i move on very fast those videos i make this is maybe i don't know like one fourth of the ideas i'm working on but many of them i just drop or never finish

because i don't know how thank you uh you mentioned keeping track of all the stuff you're trying to remember github i was curious is that like uh shared or is that a private thing for you it's a i mean it's a my github account but it's uh it's completely shared you can get access to it just google it and you find it yeah you'll see all my notes with all the grammar mistakes and everything it's all there yeah my other question is where did you find tiny cat massage music oh that's my husband he says okay i figured it had to be somebody new he makes music yeah he makes music for all my videos yes that's right and yeah it's his music is i think on itunes and everywhere you can definitely find it hello hi so i was always interested in houdini but i never started to learn this piece

of software and i'm wondering the level of math required to be successful in using houdini it it all depends so some people for example use only shelf tools and you can be pretty successful with that but it's the same for me it's the same like when i was using max and maya because i know both and i'll be using max and i'll be like doing something cool with particles and destruction and at some point i'll get stuck like i can't for example instance lights to particles you know like something crazy like that i want to do it i can't do it and i can't do it without some crazy python scripting thing and i find that actually in houdini yeah you will get stuck if you use shelf tools and at that point you do want to know something more about it but it's actually easier because you don't have to know python it's easier to find out what you need to do to achieve this kind of thing so i find myself less stuck in houdini and the only thing that limits me it's you know my knowledge about how to do it and there's a huge

community and i never seen anybody you know get stuck for a long time people usually answer questions when you post to forums and so on so it's it's very easy to get unstuck from things like that if you really need it to be done okay basically because uh you mentioned several pieces of software to learn to be successful and i know that there's a limit you can't learn everything but you mentioned you know both max and maya which is not too common i think so starting to learn let's say houdini which is quite yeah i'm wondering if if it's possible to learn this software because i'm a more of a moderate guy asset work in asset creation so in your personal opinion is it possible to do both and be good at these things or using houdini requires the whole

attention of the people so with creation of assets there's another thing that if if in company you're working and you have like a technical artist who knows houdini they can create tools for you to use in maya or macs or even in a game engine you're working in that you don't need to know like everything what happens behind the scenes behind the curtains all the math you just get sliders that are relevant to you it's similar thing like riggers and animators riggers provide a rig and animators just use the rig to create animation and that's that's completely valid way of working and and many studios do that that they have only a few people who know houdini from inside out to create those hda assets and then everybody else you know modelers animators everybody else can use those assets to create ivy to create buildings but they don't have to know inside out everything about it

okay thank you very much bye hello thank you for the presentation i've seen you don't use compositing at all in your workflow is it because it's an um home setup or is it for you in your opinion too much when you're working alone to try to integrate this kind of pipeline you mean realflow compositing workflows manually sometimes you can you can get more flexibility to re rework lighting colors without reading or re-rendering everything yeah yeah i mean for those videos i create i use just after effects and i do use filters of course and it does help but they don't render multiple um layers or anything like that to try to relate your renderings yeah i usually i usually because redshift is so fast i usually just rely whatever it gives me out of the box uh and you know save on time that way

because i i'm trying to still mainly focus on learning houdini and of course composing is a whole new area to learn and i did that in past in maya or max when i was running v-ray uh heavier it does give very very good you know layers and you can composite it later it's a great workflow definitely but with redshift i'm just rendering pretty much as it is and then i just use filter to correct the colors and it's done i don't composite anything else on top and you know it's much harder if you if you do a lot of work with transparency like fluids plus particles and you want you know motion blur on top of that and trying to composite this this is like advanced class for me so i would need to you know look up some tutorials for that definitely they can do pretty nice things thank you cool all right thanks very much everybody