An Interview with Jay Roth, President of NewTek's 3D Products Division
by Wes Beckwith
Jay Roth , president of NewTek's 3D graphics products division, joined NewTek in February of 2005. Roth brings 25 years experience in the 3D visual effects and graphics industry to his oversight of the expansion of the 3D product line and the advancement of technical innovation in LightWave 3DŽ, along with sales and marketing efforts relating to NewTek's 3D graphics offerings.
Roth was previously a major competitor to NewTek: he was co-founder and former CEO of Electric Image, Inc., and under his guidance that company achieved numerous milestones including development of the market-leading professional 3D application for the Macintosh, creation of the world's fastest renderer, and stunning visual effects for the destruction of downtown L.A. in James Cameron's Terminator 2.
Roth's experience in the visual effects industry began in 1980 working at Roger Corman's New World Pictures, with Academy Award winners James Cameron and Robert and Dennis Skotak. Roth has over 40 feature film, television and commercial credits, including CGI effects for MacGyver, Entertainment Tonight, The Stepford Wives and X Files.
Can you share your perspective on LightWave 3D, as someone who has primarily been a developer and user of other 3D packages?
Well, in some areas LightWave leads, and in others, it's behind the times. In regards to workflow, I think there is tremendous room for improvement. LightWave used to be very straightforward to use, but over the years, as complex features have been introduced, the workflows have gotten convoluted in many cases.
LightWave has been on the market longer than any other application currently available. Some of LightWave's features have led the industry, such as subdivision surfaces. However, the industry has also established standards, which need to be addressed in the LightWave workflow. Doing so will help make LightWave easier to use for those who have experience with other packages, and will also allow the LightWave user to benefit more directly from books and training materials which cover topics such as rigging, modeling and animation.
What do you feel should be the major goals for the development of LightWave 3D?
We intend to put innovation back into the market, but first we intend to create a codebase that allows us to do that. We have stabilized our system, and are now capable of adding more and more new features in the product. These new features are intended to be production-reliable, meaning that they really work, and are not meant to merely wow at tradeshows. Sometimes these new features will be rather flashy; but always, whether flashy or just day-to-day useful, they will be reliable and solid. As we go, more and more innovation in LightWave will become apparent and undeniable.
How will making LightWave innovative be achieved? Are there specific plans that you can share at this time? If so, what are those?
As noted above, innovation is our target goal. I think that the market will be pleased with what we do down the road. Reliability is our other goal, as most of our customers make money with LightWave, and use the product to make a living. To that end, we intend to keep LightWave always approachable, yet upgrade LightWave, and give it a more modern feel and workflow as well.
While many have called for a "fresh start" that involves rewriting the application from scratch, NewTek has stated that it is taking a progressive "stage-at-a-time" approach to redesigning LightWave from the ground up. Why does NewTek feel this can work? What are the benefits to using this approach?
This is called a "parallel changeover" in the development world, and is often considered the best approach to take when you have an actively used product in the marketplace. If we were to have followed what others have done (max, wavefront->Maya, xsi), we would go dark for a long period of time, forcing users to wait years for the results. We do not view that approach as practical, given market conditions, and the fact that our customers have been very loyal to us. So, by using the parallel changeover model, we can make available new technologies earlier in the process, far earlier than they would have been with any other approach.
What has been done to date?
Not sure what I can talk about here :)
All of the code has received a variety of revisions. Some of these changes don't necessarily have a short-term meaning to the user base, though it all will in the mid-term. Re-organizing the code structure is making the development and debugging process more efficient and reliable.
The drawing and user interaction systems have already received significant upgrades. These upgrades offer more much more performance and interactivity than previous systems, and are based upon modern methods and drawing engines. If you can work within the system faster, you can produce better results, more quickly.
The modeling systems have received significant functional enhancements, and work is underway to improve workflows. Most of the modeling functions available in other products are now available in our modeling environment, and we have also added some that no one else has. We are also improving the SDK which will allow 3rd parties much more efficient access to core data structures, with far less waste in the process.
Some significant amount of modeling functions are now accessible inside of the Layout portion of LightWave, which will significantly improve workflow, and offer potentially new ways of working and animating within LightWave. This core change was a primary factor in crossing the threshold to the new series.
The rendering system is being re-worked from the ground up to take advantage of modern data structures and algorithms. Initial improvements, available soon, should result in a minimum of 2x increase in render speeds for certain types of shaders, etc. All portions of the rendering engine are going to benefit from this initial re-working of the core. Note that the memory footprint will likely increase; we need to increase precision within the system to offer more accurate shading results.
How does it seem to be working so far?
Overall, things seem to be working very well. The response from the development group has been phenomenal, and our roadmap to the future looks very aggressive and innovative. As we add more to our staff, that roadmap becomes even more so. I am very excited about the upcoming release, as well as the foundations that we are laying for true innovation in the future.
For NewTek's white paper, "LightWave's Future: The Development Model," which contains a longer version of the interview above, link to:
http://www.newtek.com/lightwave/lwfuturedev.php
