29.09.2008 14:00
[Sonstiges]
The RISEN-Trailer - Making of, Part 2
2) You are Producer for RISEN, the game – how did manage to make the trailer and produce the game at the same time without having to forfeit sleep?
Well, the nice thing about my work is that if I do my job (the organisation) well enough beforehand then I end up with having less stress afterwards. And this worked reasonable well for this project. It was a lot of work to prepare the trailer production, but as soon as we had the coloured storyboard it was quite easy to communicate the direction in which the trailer was supposed to go. And that synchronisation of the vision is normally the main part of my job.
3) Post-production is a very important thing for almost any film project – what exactly did you do here and where were the problems?
Post-production was done by Ettore again and I can't even tell you how much fun it is to watch a professional doing his work. Until this project I never even imagined how much you could get out of a video with the means of post-production! With the various techniques from colour grading to motion blur and depth of field he made the video look almost like a movie. Because the trailer needed to be ready for the exhibition and we were running out of time in the end, we spent almost 48 hours without sleeping in the post-production studio. But I think it was well worth our while.
4) Your name sounds familiar. On what games have you worked as well?
I had the luck to work together with Piranha Bytes for a while now. Our paths first crossed with Gothic 3.
The trailer itself was then created at VIRGIN LANDS, a famous studio for 2D/3D graphics and animation. Tobias Weingärtner, co-founder and chief executive of VIRGIN LANDS took the time to answer a few questions as well. He also provided some pictures from the production phase – which is uncommon but it demonstrates quite well the high standard on which VL operates.
My name is Tobias Weingärtner. I am co-founder and CEO of Virgin Lands since March 1997. Between 1994 and 1996 I was involved in the creation of graphics for several projects. For example, “The Black Eye 3” from Attic or “Archimedean Dynasty” (german title: “Schleichfahrt”) from BlueByte, two classics from that time. In 1996, my two colleagues Lars Wagner, Volker Jäcklein and I created created a lavish short film named “Virgin Lands”. We won two important awards for that film right at the start. The name of the film then served as label for our company as well...
1) What was the most expensive scene in the trailer? And why?
The most difficult part certainly was the final scene in the trailer – the rising of the ruins of the temple. Only the tracking shot was created by hand; the complete effect animations were pure physical calculations. After the definition of the tracking shot we simulated how the cracks in the ground developed. With that simulation we determined how the bigger fragments of the ground had to fall. And in relation to that, we created and calculated the sub-levels: Smaller debris, roots, pebbles and finally sand and dust.
After that was finished we focused on the surrounding landscape. The entire major vegetation was animated by simulating wind and tremor as the temple rised. The wind setting further more allowed us to animate the grass layer in the foreground.
All of that was intensive work because we had to aggregate the several calculation stages which each depended on previous computations – the complete scene had to be done step by step. Because prior to the rendering itself we needed to finalize the animation calculations, it took about 12 days to create this animation.
Gallery: Temple shows the structure of the scene. In total, there were many more layers at work here.
The pictures show the entire layer structure; although the entire scene is also built from several tabs from left to right.
2) What role did compositing play in the creation of the scenes? Is it faster to render the complete scene in one approach or to create the parts separately and join them later?
We wanted and had to split the scenes for several reasons. The complete trailer was created in short time. Many works were done in parallel. That means that backgrounds were finished before the characters for that scene even got modelled and animated. Since the background layers require the most rendering time we created them beforehand. When the character models and animations were done we rendered them as well. In many scenes we put the characters on different layers; up to 14 different layers for character animation for the town scene. Many character layers also had sub-levels for hair and clothing.
After all the backgrounds and characters were calculated we joined the entire footage in a compositing software. There we fine-tuned the colour tone of the layers to each other and put effects like motion blur or depth of field onto the final scene itself.
Because we divided the footage we could split the rendering times into smaller portions. We also gained more flexibility in the creation of the scenes. Both were strong arguments to do things separately. Besides that, there was a third point: the computing power and performance. Because individual scenes were so complex it would have been impossible to render them completely with meshes including textures, animations and effect in one single pass.
All renderings were divided into backgrounds and characters. After that, they were reunited in a compositing software.
3) How far could you utilise the in-game assets for the trailer?
For the character creation we used the prepared material. The foundation for almost all faces were existing real time data. Those were then refurbished with the appropriate 3D tools (Zbrush, 3D Studio Max) and worked out explicitly. Existing body meshes served as graphic basis but had to be recreated in the most cases because they never were meant to be viewed from such a close distance as in the trailer. For the clothing animation the meshes required to follow certain rules which also made a re-modelling necessary.
We had a few backgrounds to give us some graphical orientation. But we created all scenes independently from the existing material because the technical demands between real-time and cinematics simply are too different.
4) What does the studio pipeline at Virgin Lands look like?
We work separated by core points. There are graphic artists with focus on character modelling environmental modelling, character animation, effect creation, staging (textures and lighting of the scenes) including render supervision and compositing with 2D post processing.
That means for creating a trailer there is not any one person who solely works on a given film part all by himself. Instead, a scene is the combined work of a dozen of graphic artists. The different competencies and the work with different tools for the individual parts require that approach. And further more, it leads to a consistent result without having different, individual graphic styles in different scenes of the production.
5) What do you have to present to get a job in an animation studio like VL? (We have many fans enthusiastic about graphics so that question is certainly of interest.)
Artistic skills, technical understanding and a big motivation. Watching a final movie allows to estimate the artistic work. But the technical background to get those results should not be underestimated. Scenes like the rising temple require the creative competency (how shall the scene look like) and the technical understanding (what parameters do I have to tweak) to go hand in hand. Otherwise, you won't get any consistent result.
technology: the (difficult) path to the final result...
LINKS:
Of course, there's a link to Virgin Lands:
http://www.virgin-lands.com/vl.php
As well as one to Michael Paecks Cliffhanger Productions:
http://www.cliffhanger-productions.de/
And about Ettore Pizzetti:
Ettore Pizzetti has been into making movies for 8 years now. After having graduated from the SAE college in Vienna (Digital film) and the NYFA New York Film Academy (Directing, Cinematography) he participated in the making of more than 13 movies, 4 of them as a director. His movie “Coney Island” has won the “best guerrilla documentary” award and is still used as a training video at the NYFA, and the movie “Lucky” has won the “best cinematography” award at the Brooklyn film festival.
Quote from the Internet Movie Database: http://www.imdb.com/name/nm2018615/
Well, the nice thing about my work is that if I do my job (the organisation) well enough beforehand then I end up with having less stress afterwards. And this worked reasonable well for this project. It was a lot of work to prepare the trailer production, but as soon as we had the coloured storyboard it was quite easy to communicate the direction in which the trailer was supposed to go. And that synchronisation of the vision is normally the main part of my job.
3) Post-production is a very important thing for almost any film project – what exactly did you do here and where were the problems?
Post-production was done by Ettore again and I can't even tell you how much fun it is to watch a professional doing his work. Until this project I never even imagined how much you could get out of a video with the means of post-production! With the various techniques from colour grading to motion blur and depth of field he made the video look almost like a movie. Because the trailer needed to be ready for the exhibition and we were running out of time in the end, we spent almost 48 hours without sleeping in the post-production studio. But I think it was well worth our while.
4) Your name sounds familiar. On what games have you worked as well?
I had the luck to work together with Piranha Bytes for a while now. Our paths first crossed with Gothic 3.
The trailer itself was then created at VIRGIN LANDS, a famous studio for 2D/3D graphics and animation. Tobias Weingärtner, co-founder and chief executive of VIRGIN LANDS took the time to answer a few questions as well. He also provided some pictures from the production phase – which is uncommon but it demonstrates quite well the high standard on which VL operates.
My name is Tobias Weingärtner. I am co-founder and CEO of Virgin Lands since March 1997. Between 1994 and 1996 I was involved in the creation of graphics for several projects. For example, “The Black Eye 3” from Attic or “Archimedean Dynasty” (german title: “Schleichfahrt”) from BlueByte, two classics from that time. In 1996, my two colleagues Lars Wagner, Volker Jäcklein and I created created a lavish short film named “Virgin Lands”. We won two important awards for that film right at the start. The name of the film then served as label for our company as well...
1) What was the most expensive scene in the trailer? And why?
The most difficult part certainly was the final scene in the trailer – the rising of the ruins of the temple. Only the tracking shot was created by hand; the complete effect animations were pure physical calculations. After the definition of the tracking shot we simulated how the cracks in the ground developed. With that simulation we determined how the bigger fragments of the ground had to fall. And in relation to that, we created and calculated the sub-levels: Smaller debris, roots, pebbles and finally sand and dust.
After that was finished we focused on the surrounding landscape. The entire major vegetation was animated by simulating wind and tremor as the temple rised. The wind setting further more allowed us to animate the grass layer in the foreground.
All of that was intensive work because we had to aggregate the several calculation stages which each depended on previous computations – the complete scene had to be done step by step. Because prior to the rendering itself we needed to finalize the animation calculations, it took about 12 days to create this animation.
Gallery: Temple shows the structure of the scene. In total, there were many more layers at work here.
The pictures show the entire layer structure; although the entire scene is also built from several tabs from left to right.
2) What role did compositing play in the creation of the scenes? Is it faster to render the complete scene in one approach or to create the parts separately and join them later?
We wanted and had to split the scenes for several reasons. The complete trailer was created in short time. Many works were done in parallel. That means that backgrounds were finished before the characters for that scene even got modelled and animated. Since the background layers require the most rendering time we created them beforehand. When the character models and animations were done we rendered them as well. In many scenes we put the characters on different layers; up to 14 different layers for character animation for the town scene. Many character layers also had sub-levels for hair and clothing.
After all the backgrounds and characters were calculated we joined the entire footage in a compositing software. There we fine-tuned the colour tone of the layers to each other and put effects like motion blur or depth of field onto the final scene itself.
Because we divided the footage we could split the rendering times into smaller portions. We also gained more flexibility in the creation of the scenes. Both were strong arguments to do things separately. Besides that, there was a third point: the computing power and performance. Because individual scenes were so complex it would have been impossible to render them completely with meshes including textures, animations and effect in one single pass.
All renderings were divided into backgrounds and characters. After that, they were reunited in a compositing software.
3) How far could you utilise the in-game assets for the trailer?
For the character creation we used the prepared material. The foundation for almost all faces were existing real time data. Those were then refurbished with the appropriate 3D tools (Zbrush, 3D Studio Max) and worked out explicitly. Existing body meshes served as graphic basis but had to be recreated in the most cases because they never were meant to be viewed from such a close distance as in the trailer. For the clothing animation the meshes required to follow certain rules which also made a re-modelling necessary.
We had a few backgrounds to give us some graphical orientation. But we created all scenes independently from the existing material because the technical demands between real-time and cinematics simply are too different.
4) What does the studio pipeline at Virgin Lands look like?
We work separated by core points. There are graphic artists with focus on character modelling environmental modelling, character animation, effect creation, staging (textures and lighting of the scenes) including render supervision and compositing with 2D post processing.
That means for creating a trailer there is not any one person who solely works on a given film part all by himself. Instead, a scene is the combined work of a dozen of graphic artists. The different competencies and the work with different tools for the individual parts require that approach. And further more, it leads to a consistent result without having different, individual graphic styles in different scenes of the production.
5) What do you have to present to get a job in an animation studio like VL? (We have many fans enthusiastic about graphics so that question is certainly of interest.)
Artistic skills, technical understanding and a big motivation. Watching a final movie allows to estimate the artistic work. But the technical background to get those results should not be underestimated. Scenes like the rising temple require the creative competency (how shall the scene look like) and the technical understanding (what parameters do I have to tweak) to go hand in hand. Otherwise, you won't get any consistent result.
technology: the (difficult) path to the final result...
LINKS:
Of course, there's a link to Virgin Lands:
http://www.virgin-lands.com/vl.php
As well as one to Michael Paecks Cliffhanger Productions:
http://www.cliffhanger-productions.de/
And about Ettore Pizzetti:
Ettore Pizzetti has been into making movies for 8 years now. After having graduated from the SAE college in Vienna (Digital film) and the NYFA New York Film Academy (Directing, Cinematography) he participated in the making of more than 13 movies, 4 of them as a director. His movie “Coney Island” has won the “best guerrilla documentary” award and is still used as a training video at the NYFA, and the movie “Lucky” has won the “best cinematography” award at the Brooklyn film festival.
Quote from the Internet Movie Database: http://www.imdb.com/name/nm2018615/
