I Spent 7 Years Learning VFX for This #vfx #digital #digitaleffect

Seven years ago I made my very first VFX shot. And it's not exactly a masterpiece, but since then I've improved a lot. And it's finally time to remake this disaster. But what I didn't expect was that this time would be even harder than the first. So join me as we head to the real Loch Ness Lake in Scotland and do this the right way. I'm kind of nervous. Do you want to know why? You see, Erik is an expert when it comes to water simulations. And prior to making this video, I had only done the Tona tutorial. So it's not exactly my best work. Are you ready? Uh, I think so. Let's do this. I'm excited. It's really awkward. I hate watching this, but it's okay. I love your actor. Yeah, it's my... Oh! Yeah, that's the... It's really not bad at all though. That's actually insane. That's your second project? Wow. Whoa. Where did that come from? Ambitious. That's crazy ambitious. I think this is probably like the best work someone's ever done after doing just the Tona tutorial. This was my first time doing animation, water, rendering, like everything. But since then, I've learned a lot. So I think it's finally time to remake the whole thing. And to do that, we're going to Scotland, baby. Let's go. Take your own. We just landed in Edinburgh? Edinburgh? Something like that. And I'm so excited. So you might be wondering, why Scotland? Well, there's one thing Scotland is known for. Or two things. Loch Ness. Home of one of the most famous legends ever. The Loch Ness Monster. Also known as Nessie. For decades, people have recorded strange sightings in the water. Dark shapes, blurry photos, and even sonar readings. Most of it, probably a myth. But that's kind of what makes this place so exciting. And after a four-hour drive from Edinburgh, we finally arrived. And it was beautiful. It's perfect. We started shooting straight away. And after a lot of running shots. This is the only exercise I get. We somehow ended up spending 30 minutes on this one close-up of me starting. Oh, it's impossible. That didn't look that bad until you gave up. Which I didn't even end up using. So you want me to say my perfectionism gets in my way? Yep. We then filmed some reaction shots of me pretending to pet Nessie. And finally, Eric borrowed the shoulders of our very kind friend Luke. To get some higher angle shots. And I don't know, at that moment, I just felt so grateful to be surrounded by these incredible people that like genuinely cared about this silly project. Because when I made the original version of this, it was completely different. Back then, it was just me, my friend's little brother, who I basically had to bribe into being the actor. My mom wrestling with the green screen. And all of this was happening at this tiny lake near home. At that time, I was so used to doing all of this on my own. And I didn't really have anyone around me who shared the same interests. So I think 16-year-old me would be so, so happy if she knew it was only temporary. That eventually, one day, she'd find her people. And with that, we had everything we needed. No, but what's that? That's... It's a wrap! Unfortunately, though, we didn't get any actual sightings of Nessie. Which would have saved me a lot of work. So since Nessie refused to cooperate, I had to do what every VFX artist does best. Edit in post. Last time, I used this ready-made 3D model, which honestly looks pretty solid. But I wanted more details this time. Plus, I didn't really want it to look like a monster. You know, that's more Eric's style. I wanted something that felt like a friend. Something that looks adorable. So since I figured out last time with Aslan that sculpting isn't really my biggest talent, I found these incredible guys that actually know what they're doing. This is Marvin and Till. And they have worked on some of the biggest movies in the world. And they were even part of creating Baby Groot. It doesn't get cuter than that. I immediately knew they were the perfect fit. So I gave them a pretty detailed description of what I wished for. Long neck, detailed skin, a ridge-like structure on the back, slim face, pointy ears, cute eyes, and dinosaur-like textures. A few weeks later, the model was ready. And I was blown away by the quality. It's exactly as I imagined it. It's now time for my most favorite thing in the world. Animating. There was a lot that had to be animated. Swimming motions, facial expressions, jumps out of the water. Basically everything. Honestly, I considered giving up before I even started. And just hiring someone to animate it for me, like Eric usually does. Lazy much? Hey! But then I thought, I should at least give it a try first. And just see how far I get before I throw out my computer. Luckily, the model already came with a few built-in animations. Like swimming and idle motions. So instead of starting from scratch, I kept the subtle movements from the original animations. While adding more specific actions on top. Like lifting the neck or reactions. And of course, the whole jumping shot. Which took a while to get right. For the facial expressions, I studied a lot of reference footage. I then added a bunch of shape keys to the face. And keyframed different parts depending on the emotion I wanted to show. And to make everything feel a little less stiff. I added some subtle randomized motion to the head as well. It was definitely a challenge to make it feel natural and dynamic. But after a full week of work, all the shots were finally done. Or at least good enough. I'm pretty nervous about the water simulations coming later. You know, considering last time. But fortunately, if all goes horribly wrong. I've got some practical assets up my sleeve. You see, Erik and I are physically incapable of going on a normal vacation. Even at his family's cabin, we somehow ended up working. We managed to convince his entire family to go out late in the evening. And help us fill rocks splashing into the water. Which will actually be perfect for this project later. But then we realized, whoa, this is kind of fun. So we just kept going. We filmed fire, snow, rain. And Erik was a little too dedicated. And after building up all these assets, we thought, why keep them to ourselves? So we made a website where you can get them too. But I'm a perfectionist. So finding a domain name was a real challenge. We thought of vfxstore.com. But it just didn't feel right. Until we did this. There we go. No fluff. Plus, it looks super clean. And we're not the only creators to figure out that a .store domain is the way to go when you launch your store. It doesn't get more perfect than your name .store. And MrBeast, iShowSpeed, Colin and Samir, ZachKing, and many more are already on board. A .store for a store is also known to get twice as many clicks. Which means more sales. And a clear, intuitive domain name just feels more premium. If you're a creator who's already got a store or thinking of starting one, don't miss the boat. Secure your channel name .store before someone else does using the link in my description. Thank you to my friends at .store domains for sponsoring the video. It's now time for the water simulations. Which I've actually never tried again since that first time. I think I've always tried to stay as far away as possible. But now it's finally time to face my fears. And I'm actually gonna try on a new software for this. Houdini. Houdini is known for being top tier when it comes to all types of simulations. So much so that it's used in pretty much every movie. But I've never even opened Houdini before. And I definitely don't think water simulations are the most beginner friendly. How bad can it be? So I started with a beginner course. Which actually was pretty fun. And I was quite impressed with myself when I had managed to create this. But then I started the real water course. Which felt like going from learning 2 plus 2 to quantum physics. It was clearly not for beginners. So I'd only done one water sim in blender seven years ago. So after having done two thirds of the course. And still being almost as clueless as the day I started. I humbly asked Eric to give me some private tutoring. Which after seeing all my frustration for the past few days. He quickly said yes too. So let me explain what I ended up doing. Firstly I exported the Nessie animation and camera from blender. And put it straight into Houdini. I then set the boundary for the simulation. Converted Nessie to a collision volume. Built a water simulation network. And hit simulate. Okay. We needed way more particles. So I cranked up the quality. Which made the simulation take seven hours. But it led to this really nice simulation. I then created a white water source from these particles. And simulated the white water. Which is the foam bubbles and spray particles. I turned the particles into a mesh. Which is a whole process in itself. But it took another five hours for my poor computer. Well where I'm from that's called slavery. And after all that hard work. We were still not done. So I set up a light and hit render. Which took another 12 hours. It was at this point I realized that doing nine water shots. That all take days of simulation and rendering. Was a really bad idea. But luckily the render turned out awesome. Now that this jumping shot was done. I figured the rest of the shots with less movement. Would be much easier to pull off. But oh boy was I wrong. I did a similar process for this shot. Where she slowly emerges from the water. But no water came up with her. Just like last time. But this time I would not give up. I found out that there's actually a setting. For the stickiness of the collision. So I animated that higher at the start. And turned it off at the end. So the water hopefully would slide off much slower. So I kept playing with the setting. And even set up a second water simulation. On top of the main one. And after four days of work. I ended up with something usable. Ignoring the fact that she kind of looks like Elrond here. I'm actually quite happy with this. But it had now been six days. And I had only completed two shots. At this rate I'd be lucky if I was done by Christmas. So I locked in. And over the next few weeks. I painfully worked through all nine shots. Which each somehow came with its own challenge. But eventually three agonizing weeks later. I was finally done. But before we keep going. And get to see the result of our hard work. I need to fix a major problem. I just can't keep working in this office. My desk is way too small. And could collapse any minute. And there's no curtains. Making it impossible to work while the sun's out. So I've been using this green screen for some mild relief. Until Erik pulls it over to his side. If you've been wondering what we usually fight about. This is it. On top of that the walls are screaming for some decorations. So today I've finally had enough. It's time for an upgrade. So let's start with a clean slate. The first thing I need is a new desk. What about me? Uh I guess you can have one too. Nice. But we're still missing some key details. Okay you can do it. Easy peasy. And it was finally done. And the timing could not have been better. Because I just got something very special in the mail. I'm so so grateful for all the support I've gotten on the videos. It really means the world. So yeah I'm so happy with how it all turned out. I hand a big thanks to Okiwood for sending over the desks. And accessories for all of my mess. So definitely check them out. Now it was finally time to put everything together. I started by importing the Houdini render into Nuke. I then used the Speed AOV to automatically isolate the areas where the water was moving. But Nessie looked super flat here. So instead of bothering to figure out why. I went back to my already well-lit blender scene. Rendered Nessie separately and placed her on top. I also rendered out her shadow and reflection. And mixed everything together. As a final touch I color matched Nessie so she fit better into the environment. For the close-ups I skipped the full water simulation. And used particles that looked like water droplets instead. And after a few adjustments I finally hit render. While waiting we went on a little side quest. And brought home a new member of the family. Say hi to Nessie. You know you've been working on a project for way too long when you name your dog after it. But honestly working on this project has meant a lot to me. And I already know that this dog will mean a lot to you. Even though she's a bit crazy. Anyways once the render was done I made some color adjustments. And added this dripping water acid on top for extra realism. For the shot where I touched her we needed way more space. Otherwise we would only see her nose. So I extended the image in photoshop. And then rotoscoped out my hand and placed it over Nessie. But as you can see it didn't really look like we were interacting. So I added some shadows. And slightly darkened her nose where my hand touches her. And as a final touch I added some dripping water here as well. For the jumping shot I masked out the edges of the water. Made some adjustments. And layered in a few of the acids we made earlier for more realism. And it worked like a charm. You know it's a small thing. But those extra details really help sell the shot. And that's exactly what we wanted with our store. High quality acids that are super easy to just drop into your own videos. Whenever you need that extra layer of realism. And a little secret. I created the discount code MARIA30 so you can get 30% off. Don't tell Eric. Finally I composited the rest of the shots. Did some color grading. And after two months of work it was finally done. Whoa. It's crazy to think that this whole journey started seven years ago. When I knew almost nothing. And had to learn everything from scratch. There were so many times I just wanted to give up. But I didn't. I just kept reminding myself that if other people can learn this. Then so can I. And even now with years of learning and months of work on this project. It's still not perfect. And honestly I think that's okay. Because nowadays with AI. Perfect can be generated in seconds. Did it just eat my bomb? But it's the struggles. The effort. And the imperfections. That makes it human. That makes it valuable. And I'll always choose imperfect realness over soulless perfection. So with all that in mind. Let's finally reveal the results.