Making of Lensha by Loocas Duber, Czech Republic (Europe)
Rendering
This was a nightmare! Of all the things done on this project, rendering was the most annoying and difficult part. Not that it was hard to setup the renderer, nor it was difficult to create all the materials or to light the scene, but it was almost impossible for Mental Ray, which was only used for rendering the girl’s head, to render the scene out on my machine. I have only 2GB of memory and with such aspects as: very large textures, very large render, area shadows, blurred reflections, sub-surface scattering etc... it was almost impossible to render it out. I had to tweak the settings numerous times, I had to change a few things, cheat certain effects, and especially divide the render into several passes and later to compose them back in Photoshop. As I said already, I chose Mental Ray for the head, because it has simply the best 3S shader for skin materials over its competitors. No matter what are its limitations, this was something I just had to use Mental Ray for. The rest was rendered in finalRender Stage-1, which is 99.9% my renderer of choice. Very fast, flexible, reliable and for everything, except for realistic skin :( it’s just perfect.
So, when everything was rendered out in individual layers, as you can see on the attached image, I took all the layers into Photoshop and started composing them back together. Just to note, I used four lights to light the scene, more or less only one of them was the main source coming from behind the camera, the rest were just fills that you can barely spot, they served for fine-tuning the lighting provided by Global Illumination. The composition was placed in front of a photograph that I found on the internet and used it to match the camera and place the objects in front of it. Aside from the problems I was having with Mental Ray, the rendering process was quite flawless and, again, straight forward. The hair on the ski-cap was created using built-in Shave & Haircut system, as well as the hair of the character.
Final Touches and Conclusion
Once I composed all the layers together, it was time to fine-tune and color-correct the entire image. Photoshop is an unbeatable king in this field and I did really enjoy this part tweaking the colors and values of the render to get to a point I liked the best. Once that was done, I collapsed it all together, added my signature and had it printed :) It’s really rewarding seeing your work, after so much time spent on it, hung on a wall, nicely framed :) I really recommend printing your own work once you’re done with it. That’s also why I work with such resolutions, because I just love all the detail you can see on the picture. Also, the artwork looks somehow different, more alive when printed!
To summarize the project, aside from the animation, which I’m still yet to do, it was also a great experience for me to work on it, you always learn something new or at least you train your skills with every single project you do and believe it or not, it’s quite relaxing working on something own after spending endless hours working for clients. Since this was some sort of a side-kick from the training DVDs I did, it also demonstrates what I said on the DVDs, it’s entirely up to you where you take what you end up with learning from my DVDs and you can create portraits, short animation tests, or even entire characters, even though my DVDs aren’t about them.
So, thank you very much for reading and I hope you’ll enjoy the image at least as much as I enjoyed creating it :)
Related Links
|
|
|
