Recreating a Masterpiece with LightWave
Artist Zsolt Ekho Farkas walks you through the process of creating Benczúr Gyula's 2D painting Budavár Repossession in LightWave
Posted: Wed 23 Apr 2014
First of all, this project started as a funny challenge from my wife, but became something more.
Short background story:
We were on holiday solving riddles, and there was a medieval painting [Benczúr Gyula's painting "Budavár Repossession"] in the booklet at the Hungarian National Gallery. My wife asked me if I could do it in 3D and I said yes, of course. It was soooo impressive that I decided to create the scene as soon as we got home.
During the creation I made a few bad decisions, but it was an extremely useful and instructive learning process. Seeing this kind of popularity (which I've never thought of), I decided to make another one. Now that I know the obstacles, and the parts that needed to be made differently, maybe a 2.5D version could look better, but I WANTED to make it in full 3D; for my wife, and for myself of course.
Below is a small making of Gyula's Budavár Repossession" using LightWave 3D:
First, I had to analyze the painting and figure out the character positions in the 3d space. It wasn't too hard but the scene is a bit crowded.
Next came the modelling. The main idea was this: I should make every character with a full rig, ready to move and animate if neccessary, but I didn't think of animating it at all initially. The rigging was easy thanks to LightWave's Genoma feature.
After the characters were fully rigged, I had to freeze the whole scene due to my computer's low resources (well, not so weak, but not an atomic reactor ) . I was however, able to handle 8-9 rigs easily, but 32? No way.
Detailed modeling, without displacement maps. The polycount reached the 8.5 million so I had to forget any additional sculpting. It was nice and detailed enough though, at least for me.
Then on to texturing. Usually the simpliest solution is the best, just a planar projection and voilá! Finished! But... nah. I had to paint additional maps, to enhance the original image. It was a simple planar projection, because like I mentinoed earlier, I didn't think of animation or any movement at all in the inital stage of this image. Why should I make unnecessary UVs right? As I mentioed earlier, this image was made for my loving wife and nothing more!
When the scene reached a poly count of 8.5 million, I had to forget about displacement maps, however the scene was looking good without them. Below is a test render with maps:
Below is another test render with the additional maps.
So okay, almost ready, what next?
And the idea came: I should make a small movement/ animation. But as I said before my hardware limited me because of the high poly count. And there was another mistake/ problem/ obstacle: for a proper animation I ought to remodel the WHOLE SCENE! When I started the project I didn't think of any movement. So one month modeling and just dump it into the trash? No way! I had to work with what I already had.
Here comes the fun part, I had to repaint all of the characters that were partially obscured. As you can see , it is a little bit crowded there. I used mostly the clone stamp tool, but had to repaint the ropes, gloves, boots, etc. It was a lot of work.
I can't say too much about camera movement, maybe it's too small, but this was my first project of this type.
So okay, what now? The movement is almost there, but the entire feeling was too simple. Then another idea came: lets drop in some smoke! But, due to my limited hardware I was unable to make a smoke simulation (oh, but it would have been be so cool).
Instead, I decided to make the smoke effect with layers of masks with the camera movement exported from LightWave.
There were 3-4 mask layers / act. But I could bump up the overall feeling with the depth maps and enchance the DOF, for example. By the way, I used the DP Kit DOF in LW.
Below is a depth screenshot:
But back to the smoke. You can get a lot of stock and free footage onilne, but I decided to make my own smoke footage. It was soo simple. I just grabbed my camera, tripod, and two led lights. The backdrop was my monitor screen (turned off). The smoke was made with my ecig. At first, the footage was too dark with a greenish/ blueish tint, but after a little correction it was fine enough to use.
Below is an image with the smoke layers.
It was almost ready when I realized online players can't play it properly without lag, so I had to re-cut everything. One and a half month's worth of work went to the dump. But after 140 tests I was able to reach a desired speed, with only a little choppiness.
Almost there, just a little correction and TADA!
Here's the final video again, so you don't have to scroll back to view it.
Below is a bonus picture - the origina hanging in the National Gallery of Hungary.
Be sure to visit the Hungarian National Gallery to learn more about the original painting!
Q&A PEOPLE HAVE ASKED ABOUT this project
Q. Would it be better in 2.5D instead of 3D?
A. Yes, maybe but I wanted to make it in 3D!
Q. Why 3D instead of 2.5D?
A. My wife challenged me to do it. and WHY NOT? It is FUN!
Q. Is this a half-hearted work?
A. I spent one month modeling, three weeks painting, and another few weeks in masking and effecting. This project was made especially for my wife and not the public really. It is full of heart!
Q. Why didn't you make your own composition?
A. I've done enough of my own compositions. This time I wanted to try something different and recreate a 2D masterpiece in 3D.
Q. Why isn't there more camera movement in the animation?
A. My computer couldn't handle the rigs, and believe me I tried the UV 32 freeze/sculpted mode), but this project was not made for animation in the first place, so I couldn't really create a lot of movemetn with the camera.
Q. Could you have done it in 2.5D?
A. I don't think so, but if someone else wants to give it a try and show the difference between 2.5D and 3D, I'd really like to see it.