Nervo.tv
03-22-2008, 01:29 AM
Nervo (http://www.nervo.tv) is an animation & design firm based in Portland, OR. The images and information below are related to a TV spot we have recently finished for TIMEX. Here is a link to the actual animation (http://nervo.tv/index.html?sect=5&proj=timex).
---x---
In our first collaboration with Arnold Worldwide, Nervo has created this TIMEX spot featuring the new Ironman IControl watch for Ipod. The spot, featuring a graphic runner, the watch itself and an iPod, was visually sewn together by a graphic sound wave and light effects that crossed the whole picture from beginning to end.
The song "Clear Island" by band "Liars" was edited by Nervo's Music Director Darrin Wiener who has also created all of the sound effects. Music played a big role in this project as all of the animated elements, including the runner's movements, had to follow the pace and structure of the song. Because of this it was very helpful to have an in-house musician who we could closely collaborate with as all of the different parts of the spot were being roughed into an animatic.
At the end of the spot, below the TIMEX logo, we've incorporated a short but special live-action element. We utilized a pool of magnetically charged ferrofluid to represent the energy of the run and the song accumulated at the end. You can see a clip of the raw material here (http://www.vimeo.com/806709).
http://www.nervo.tv/press/timex/timex_watch_frame.jpg
* This image is one of the many that were created for the same project, but was actually made at an extremely high-definition (11.000 pixels wide) to serve for various print purposes. You may comment on this image on Flickr here (http://www.flickr.com/photos/nervo/2350936458).
Abstracting 3D
With a style focused piece such as this, it is important to plan well in advance a work flow that addresses such needs. The style frames required a very stylized runner passing by trees, jumping a man-hole and going through a city, with lots of particles and sound waves around!
http://www.nervo.tv/press/timex/timex_all_styleframes_small.jpg
* Style-Frames designed at Nervo. You can comment on these and view other frames on Flickr (http://www.flickr.com/photos/nervo/sets/72157604171699003/). View all of the Style Frames together at a larger resolution here (http://www.nervo.tv/press/timex/timex_all_styleframes_large.jpg).
As you can see, the style-frames above are relatively simple. The graphics are mostly duotone (Black and Orange), but achieving the look straight out of render would be very improbable and would require too much effort upfront in the creative process that can be time consuming. In order to work under tight schedules and keep up with a productive interaction with the client we opted to use XSI to generate the necessary layers and tune the final look in Adobe After Effects.
In order to give us the maximum amount of freedom and flexibility in post-production we decided to output many channels per pass. This is a standard process that we have adopted for all of our projects. It allows for each team member to focus on very particular parts of the process and to avoid re-rendering many passes for aesthetic purposes. It also keeps render time to a minimal while still allowing isolation of elements all while render at once.
http://www.nervo.tv/press/timex/Runner_passes_and_channels_small.jpg
* Example of some of the many render passes we created for the compositing process of the spot. Here is a larger version of this image (http://www.nervo.tv/press/timex/Runner_passes_and_channels.jpg).
http://www.nervo.tv/press/timex/channel_output_small.jpg
* Other example of how we separated the passes for further compositing. Larger version here (http://www.nervo.tv/press/timex/channel_output.jpg).
---x---
Use of MOTOR
Re-targeting an animation clip and dropping it straight on the mixer isn't something that works well without plenty of tweaking, and in the other hand, cleaning mocap data isn't fun either. Therefore, instead of using MOTOR to get the final animation, we used it as a reference of a real person running in 3D. Then we created an animation from scratch and applied it to our final character, which then made it stylizing and tweaking the character's look a simpler process.
---x---
Referencing Everything
Our pipeline for this project was entirely based on referenced models in XSI 6.5. In order to have a completely non-linear workflow we adopted the use of referenced models for this project, which is often how we work.
All of the different parts of the project, ranging from particles to cloth simulation were transferred using this pipeline. All models of the same kind, such as the trees for instance, had a default structure with generic groups and names, which allowed us to keep refining the models and advancing the layout of the scene simultaneously. At the end of the day, swapping one model for another involved just a couple of clicks.
The names of the models and simulation settings were tied by an internal tool that determined the name of simulation paths and their settings based on its version. Another advantage for working with reference models and being able to actively update models at the same time as the main scene was being animated was that in this way the Delta mechanism was responsible for keeping the updated Animation clips and layers, simulation settings, shaders, render passes while everything else was decided in the main scene. In the end the final scene handled about 3 million triangles + Cloth simulation + Hair + particles while rendering everything at High-Definition (1920x1080).
http://www.nervo.tv/press/timex/xsi_interface_small.jpg
* Screenshot from one of our scenes. Larger version here (http://www.nervo.tv/press/timex/xsi_interface_large.jpg).
The entire rig of our main character along with his simulation setting were stored in the final scene. The run cycles were animated in a separate scene and then exported as an action so the final scene received that action as an animation clip. We mixed the cleaned mocap along with the imported keyed clip for some additional personality to his movements. This kept all the elements separate so the final setup of the runner was independent of its own animation, as well as it simulaton settings, shaders and passes was independent of the model.
You can also watch two clips from XSI here (http://www.vimeo.com/810209) and here (http://www.vimeo.com/810526) as well as a video clip of the untreated footage of Ferrofluid that we shot for the logo end sequence here (http://www.vimeo.com/806709).
Even-though it may seem unrelated to the subject of this website, it may be interesting to some XSI users to see how we tackled the animation of the sound wave that appears throughout the spot. The image below shows an example of how we created the orange lines in After Effects.
http://www.nervo.tv/press/timex/soundwave_detail.jpg
* Expressions in After Effects. You can read more about it and see this image larger here (http://www.flickr.com/photos/nervo/2347972491) and other ones here (http://www.flickr.com/photos/nervo/2347972475) and here (http://www.flickr.com/photos/nervo/2348760292).
We hope you enjoy our latest project and that this post has been helpful.
From the team at Nervo (http://www.nervo.tv).
PS: If you have read it up to this point and want to see the video larger, follow this link (http://www.nervo.tv/media/motion/video/nervo_qt_timex.mov).
If you are interested in reading other articles on our projects take a look at "Let Us Kiss" (http://community.softimage.com/showthread.php?p=4183#post4183) as well.
---x---
In our first collaboration with Arnold Worldwide, Nervo has created this TIMEX spot featuring the new Ironman IControl watch for Ipod. The spot, featuring a graphic runner, the watch itself and an iPod, was visually sewn together by a graphic sound wave and light effects that crossed the whole picture from beginning to end.
The song "Clear Island" by band "Liars" was edited by Nervo's Music Director Darrin Wiener who has also created all of the sound effects. Music played a big role in this project as all of the animated elements, including the runner's movements, had to follow the pace and structure of the song. Because of this it was very helpful to have an in-house musician who we could closely collaborate with as all of the different parts of the spot were being roughed into an animatic.
At the end of the spot, below the TIMEX logo, we've incorporated a short but special live-action element. We utilized a pool of magnetically charged ferrofluid to represent the energy of the run and the song accumulated at the end. You can see a clip of the raw material here (http://www.vimeo.com/806709).
http://www.nervo.tv/press/timex/timex_watch_frame.jpg
* This image is one of the many that were created for the same project, but was actually made at an extremely high-definition (11.000 pixels wide) to serve for various print purposes. You may comment on this image on Flickr here (http://www.flickr.com/photos/nervo/2350936458).
Abstracting 3D
With a style focused piece such as this, it is important to plan well in advance a work flow that addresses such needs. The style frames required a very stylized runner passing by trees, jumping a man-hole and going through a city, with lots of particles and sound waves around!
http://www.nervo.tv/press/timex/timex_all_styleframes_small.jpg
* Style-Frames designed at Nervo. You can comment on these and view other frames on Flickr (http://www.flickr.com/photos/nervo/sets/72157604171699003/). View all of the Style Frames together at a larger resolution here (http://www.nervo.tv/press/timex/timex_all_styleframes_large.jpg).
As you can see, the style-frames above are relatively simple. The graphics are mostly duotone (Black and Orange), but achieving the look straight out of render would be very improbable and would require too much effort upfront in the creative process that can be time consuming. In order to work under tight schedules and keep up with a productive interaction with the client we opted to use XSI to generate the necessary layers and tune the final look in Adobe After Effects.
In order to give us the maximum amount of freedom and flexibility in post-production we decided to output many channels per pass. This is a standard process that we have adopted for all of our projects. It allows for each team member to focus on very particular parts of the process and to avoid re-rendering many passes for aesthetic purposes. It also keeps render time to a minimal while still allowing isolation of elements all while render at once.
http://www.nervo.tv/press/timex/Runner_passes_and_channels_small.jpg
* Example of some of the many render passes we created for the compositing process of the spot. Here is a larger version of this image (http://www.nervo.tv/press/timex/Runner_passes_and_channels.jpg).
http://www.nervo.tv/press/timex/channel_output_small.jpg
* Other example of how we separated the passes for further compositing. Larger version here (http://www.nervo.tv/press/timex/channel_output.jpg).
---x---
Use of MOTOR
Re-targeting an animation clip and dropping it straight on the mixer isn't something that works well without plenty of tweaking, and in the other hand, cleaning mocap data isn't fun either. Therefore, instead of using MOTOR to get the final animation, we used it as a reference of a real person running in 3D. Then we created an animation from scratch and applied it to our final character, which then made it stylizing and tweaking the character's look a simpler process.
---x---
Referencing Everything
Our pipeline for this project was entirely based on referenced models in XSI 6.5. In order to have a completely non-linear workflow we adopted the use of referenced models for this project, which is often how we work.
All of the different parts of the project, ranging from particles to cloth simulation were transferred using this pipeline. All models of the same kind, such as the trees for instance, had a default structure with generic groups and names, which allowed us to keep refining the models and advancing the layout of the scene simultaneously. At the end of the day, swapping one model for another involved just a couple of clicks.
The names of the models and simulation settings were tied by an internal tool that determined the name of simulation paths and their settings based on its version. Another advantage for working with reference models and being able to actively update models at the same time as the main scene was being animated was that in this way the Delta mechanism was responsible for keeping the updated Animation clips and layers, simulation settings, shaders, render passes while everything else was decided in the main scene. In the end the final scene handled about 3 million triangles + Cloth simulation + Hair + particles while rendering everything at High-Definition (1920x1080).
http://www.nervo.tv/press/timex/xsi_interface_small.jpg
* Screenshot from one of our scenes. Larger version here (http://www.nervo.tv/press/timex/xsi_interface_large.jpg).
The entire rig of our main character along with his simulation setting were stored in the final scene. The run cycles were animated in a separate scene and then exported as an action so the final scene received that action as an animation clip. We mixed the cleaned mocap along with the imported keyed clip for some additional personality to his movements. This kept all the elements separate so the final setup of the runner was independent of its own animation, as well as it simulaton settings, shaders and passes was independent of the model.
You can also watch two clips from XSI here (http://www.vimeo.com/810209) and here (http://www.vimeo.com/810526) as well as a video clip of the untreated footage of Ferrofluid that we shot for the logo end sequence here (http://www.vimeo.com/806709).
Even-though it may seem unrelated to the subject of this website, it may be interesting to some XSI users to see how we tackled the animation of the sound wave that appears throughout the spot. The image below shows an example of how we created the orange lines in After Effects.
http://www.nervo.tv/press/timex/soundwave_detail.jpg
* Expressions in After Effects. You can read more about it and see this image larger here (http://www.flickr.com/photos/nervo/2347972491) and other ones here (http://www.flickr.com/photos/nervo/2347972475) and here (http://www.flickr.com/photos/nervo/2348760292).
We hope you enjoy our latest project and that this post has been helpful.
From the team at Nervo (http://www.nervo.tv).
PS: If you have read it up to this point and want to see the video larger, follow this link (http://www.nervo.tv/media/motion/video/nervo_qt_timex.mov).
If you are interested in reading other articles on our projects take a look at "Let Us Kiss" (http://community.softimage.com/showthread.php?p=4183#post4183) as well.