If you call yourself a vfx artist and you haven't seen the Third and the Seventh...
Well... You should have. This is an exquisitely created and rendered film. Beautifully abstract and beautifully artful.
The Third and The Seventh
The film is digital. That is to say that it was rendered, and is not simply photography.
It could be called a form of visual masturbation. A way of saying, 'Hi, look at me. I'm amazing. Bow in awe.' And I have certainly accused people of doing exactly this. But in the case of this film, it is clear that the artist, Alex Roman, has a powerful understanding of the design principles underpinning his virtual construction, and an incredibly deep understanding of Art. From colour, to depth of field, space, framing, mood, atmosphere, balance and meaning. And I say meaning tentatively because one is not completely sure how much intended meaning there is in the film, and how much of a straight form and design work that is it.
The film is clearly a showcase of talent, and that talent is prodigious. This is something that can make even experienced CGI artists stare in awe and wonder and learn from. It is a film that can be analysed, broken down and examined in minute detail and it will hold up.
In some ways, this film is the very foundation of what design CGI is. Without the trappings and burdens of other areas, this is a pure art and design work. Immensely beautiful, and unmatched. Principles are not just applied, they are deeply understood. They are not used gratuitously to cover up ineptitude, bu they have been studied, understood, and used with great care.
I gush on about it, I fear, but the truth is, the only argument that can be mustered against a film like this is that it's entirely focused on iconic visual production. But that argument pales in the face of a reality that is inspirational and incredibly nuanced. As the artist says:
"A FULL-CG animated piece that tries to illustrate architecture art across a photographic point of view where main subjects are already-built spaces. Sometimes in an abstract way. Sometimes surreal."
The compositing breakdown can be found here: Compositing breakdown
Sunday 16 September 2012
Week 5 - Blender Compositing
This is a fairly new topic to the Blender community, though it's been done in the past, rarely have the tools available been sophisticated enough to composite live footage with digital footage as easily as it is now.
Now with the new features such as motion tracking and sweet core rendering improvements, one can finally use real footage and comp it with digital without enormous amounts of work.
The above video is a good example of the breakdown when using the blender program. The video is breakdown of the following one:
While the actual composite would need a bit more work as there is a bit of telltale saturation and contrast issues with different lighting in the scenes, that's easy to rectify. The salient point is that now there is a free alternative to the current industry heavyweights that has actually started to be useful for vfx work.
Blender has long been the underdog in the 3D world thanks to a pipeline that is somewhat difficult to integrate into a studio environment. While this is still an issue, the gap is rapidly closing as various studios start to stagnate between their nearly releases with nominally decent upgrades sometimes, and other times not so much. Sometimes even a reversion.
This is in stark contrast to Blender's constant and astounding rate of improvement. Features and versions are updated on a monthly, and sometimes even weekly basis. Thanks to a thriving open source community, this is one of the few examples of open source software actually working. Not only has it stayed lightweight and usable, but has continued to mature and develop features that, while rarely completely rivaling their purist counterparts (Nuke is obviously a better tracker and vfx program, Zbrush a far better sculpting program, and so on), it still holds up against the 3D modelling programs, and the recent render enhancements, simulation additions and other massive packages has allowed it to serve as a decent, understated option for those without thousands to shell out on mainstream packages and plugins.
A look at the blender gallery proves that the software is firmly capable of standing on it's own, but it's heartening to see that in an industry where some of the commercial counterparts aren't innovating, at least the little kid in the corner is not only innovating, but also catching up to it's big daddy's in leaps and bounds.
The question that remains is, what happens if the little kid grows up to be a legitimate contender for the throne?
Now with the new features such as motion tracking and sweet core rendering improvements, one can finally use real footage and comp it with digital without enormous amounts of work.
The above video is a good example of the breakdown when using the blender program. The video is breakdown of the following one:
While the actual composite would need a bit more work as there is a bit of telltale saturation and contrast issues with different lighting in the scenes, that's easy to rectify. The salient point is that now there is a free alternative to the current industry heavyweights that has actually started to be useful for vfx work.
Blender has long been the underdog in the 3D world thanks to a pipeline that is somewhat difficult to integrate into a studio environment. While this is still an issue, the gap is rapidly closing as various studios start to stagnate between their nearly releases with nominally decent upgrades sometimes, and other times not so much. Sometimes even a reversion.
This is in stark contrast to Blender's constant and astounding rate of improvement. Features and versions are updated on a monthly, and sometimes even weekly basis. Thanks to a thriving open source community, this is one of the few examples of open source software actually working. Not only has it stayed lightweight and usable, but has continued to mature and develop features that, while rarely completely rivaling their purist counterparts (Nuke is obviously a better tracker and vfx program, Zbrush a far better sculpting program, and so on), it still holds up against the 3D modelling programs, and the recent render enhancements, simulation additions and other massive packages has allowed it to serve as a decent, understated option for those without thousands to shell out on mainstream packages and plugins.
A look at the blender gallery proves that the software is firmly capable of standing on it's own, but it's heartening to see that in an industry where some of the commercial counterparts aren't innovating, at least the little kid in the corner is not only innovating, but also catching up to it's big daddy's in leaps and bounds.
The question that remains is, what happens if the little kid grows up to be a legitimate contender for the throne?
Sunday 2 September 2012
Week 4 - An Interview from the Matte Painter in Star Trek
This post will be referencing this article:
http://www.cgsociety.org/index.php/CGSFeatures/CGSFeatureSpecial/star_trek_next_gen
In this CGSociety article, we get a short look into the process that the matte painter behind the Star Trek: Original Series remaster, Max Gabl. It's often interesting to hear the thoughts of the artists who are responsible for creating the scenery behind what we see in movies.
For those unfamiliar, a matte painting is essentially painted footage that is combined with the live footage. Matte paintings are frequently used and in particular in situations where shooting live is too difficult, expensive or actually impossible. While 3D renders are used a lot, matte painting has an advantage in terms of the time taken, and the pure ability to recreate. Especially in terms of terrain, where generators can sometimes create somewhat unexpected effects, and render times of any CGI can be very long.
In this case, the artist had a very quick turnaround, and discusses how he would create up to 3 planets in the course of a day, quickly brushing in 3 different layers of detail; the surface, clouds and atmosphere; using references from NASA images for realism. These paintings would then be projected or mapped onto a sphere and rendered in Cinema 4D.
What's interesting is that when he started on the series, it would take him 3 or 4 days to complete a single planet. Which is quite a considerable amount of time, but considering how a single episode could feature multiple planets for large number of shots, it was important to get a strong sense of realism and detail. The planets had to look just right, and he would liaison with the art director to ensure an absolute sense of quality. It's certainly impressive to see such an increase in speed and skill in what would have been a relatively short space of time.
Matte painters are frequently overlooked in discussions on a film outside of the effects industry. Their intense application of detail and frequent hand painted photorealism is impressive considering that they often knock out these immense paintings in the course of a day or three. In addition to this, they're often expected to be proficient in compositing in various programs, as well as 3D. Essentially meaning that they are not only very highly technically proficient, but also multi-talented across multiple disciplines and skill sets.
This frequently means that they know a lot of tricks and time savers to push high quality content out as quickly as possible. Like his process for backgrounds on a planet; where because the camera didn't move he was able to create them as 2D images with photoshop and the occasional 3D architecture render.
This is in contrast to his process for planets, which are animated and rendered in Cinema 4D. In addition to all this is the implicit rule in 3D.
'Don't make what you can fake.'
"For The Original Series, I usually created one high-resolution 2D image of every planet. We projected that 2D image onto a 3D sphere that was rotating, but only slightly, so as not to give away the fact that it was a 2D projection. For The Next Generation, you are looking at fully textured spheres that rotate quite a bit more, sometimes over 90 degrees, and that means much more surface has to be created."
http://www.cgsociety.org/index.php/CGSFeatures/CGSFeatureSpecial/star_trek_next_gen
In this CGSociety article, we get a short look into the process that the matte painter behind the Star Trek: Original Series remaster, Max Gabl. It's often interesting to hear the thoughts of the artists who are responsible for creating the scenery behind what we see in movies.
For those unfamiliar, a matte painting is essentially painted footage that is combined with the live footage. Matte paintings are frequently used and in particular in situations where shooting live is too difficult, expensive or actually impossible. While 3D renders are used a lot, matte painting has an advantage in terms of the time taken, and the pure ability to recreate. Especially in terms of terrain, where generators can sometimes create somewhat unexpected effects, and render times of any CGI can be very long.
In this case, the artist had a very quick turnaround, and discusses how he would create up to 3 planets in the course of a day, quickly brushing in 3 different layers of detail; the surface, clouds and atmosphere; using references from NASA images for realism. These paintings would then be projected or mapped onto a sphere and rendered in Cinema 4D.
What's interesting is that when he started on the series, it would take him 3 or 4 days to complete a single planet. Which is quite a considerable amount of time, but considering how a single episode could feature multiple planets for large number of shots, it was important to get a strong sense of realism and detail. The planets had to look just right, and he would liaison with the art director to ensure an absolute sense of quality. It's certainly impressive to see such an increase in speed and skill in what would have been a relatively short space of time.
Matte painters are frequently overlooked in discussions on a film outside of the effects industry. Their intense application of detail and frequent hand painted photorealism is impressive considering that they often knock out these immense paintings in the course of a day or three. In addition to this, they're often expected to be proficient in compositing in various programs, as well as 3D. Essentially meaning that they are not only very highly technically proficient, but also multi-talented across multiple disciplines and skill sets.
This frequently means that they know a lot of tricks and time savers to push high quality content out as quickly as possible. Like his process for backgrounds on a planet; where because the camera didn't move he was able to create them as 2D images with photoshop and the occasional 3D architecture render.
This is in contrast to his process for planets, which are animated and rendered in Cinema 4D. In addition to all this is the implicit rule in 3D.
'Don't make what you can fake.'
"For The Original Series, I usually created one high-resolution 2D image of every planet. We projected that 2D image onto a 3D sphere that was rotating, but only slightly, so as not to give away the fact that it was a 2D projection. For The Next Generation, you are looking at fully textured spheres that rotate quite a bit more, sometimes over 90 degrees, and that means much more surface has to be created."
Subscribe to:
Posts (Atom)