| Q&A |
 Stefan
Dachwitz
Software Development and Consulting
october 3rd 2005 |
| |
 |
"The
Simplifikator is a tool to create level-of-detail models
from complex 3D models such as buildings or vehicles" |
|
| |
|
| Q1 |
Please
give a brief presentation of your activities. |
| A1 |
I
am in the process of setting up my own small company. My aim
is to create tools for realtime 3D, with a special focus on
the requirements of walkthrough applications, where there
is a lot of static content. My personal activities are implementing
the Simplifikator, do the marketing and trying to keep up
with technology and research in my field. |
| |
|
| Q2 |
What
is the purpose of Simplifikator? |
| A2 |
The
Simplifikator is a tool to create level-of-detail models from
complex 3D models such as buildings or vehicles. It has applications
where current progressive reduction algorithms are not applicable,
because the models have more than one material and/or texture.
Since the algorithms of the Simplifikator are resolution-oriented,
one of its main applications are optimizing scenes for devices
with a
lower screen resolution, such as mobile phones and PDAs. Instead
of generating them manually, they may now be generated automatically. |
| |
|
| Q3 |
Is
it only dedicated to static models? |
| A3 |
The
first version will support only static meshes. I am working
on the reduction algorithms to make them more suitable for animation,
so later versions might show more flexibility here (no pun intended). |
| |
|
| Q4 |
Does
it fit the needs of optimizing scene such as 3D Walkthrough? |
| A4 |
If
the walkthrough engine is using level of detail models, the
Simplifikator will be a way to generate these automatically
intead of manually. |
| |
|
| Q5 |
How
does it work? What are the main parameters? |
| A5 |
The
Simplifikator takes a whole scene, or a part of a scene, as
the input, and computes one reduced mesh and one texture from
it. This reduced model looks identical to the original model,
when the models are rendered in a small screen area. The size
of the screen area - the resolution - can be set as a parameter.
The original geometry is sampled on a fixed grid, the number
of samples are controlled by the parameter Resolution. The new
geometry is very dense, so a reduction algorithm is used to
reduce the triangles in the mesh. There is a parameter that
controls the quality of the reduced mesh.
After the reduction step the textures are generated. Here the
quality can be controlled by how many texture samples should
be taken, and which texture sampling mode meets your desired
quality.
Expect a lot of improvements in quality and performance before
the product finally hits the shelf. |
| |
|
| Q6 |
It
handles VRML1, does it include the VRML files created by 3ds
max (from VRML97 exporter)? |
| A6 |
I
am working on the VRML97 import. What might be more interesting
to most users is that I am working on supporting the FBX format,
which will make the Simplifikator cooperate with many modeling
packages like alias' Maya(R) and 3ds Max(R). Import and export
of most FBX features will be supported as well as export of
VRML1, VRML97 and X3D. |
| |
|
| Q7 |
Have
you plan to take advantage of features like normal map to improve
the quality of the optimized mesh? |
| A7 |
I
am sure normal maps are a definite must-have, along with bump
maps (for those who still remember them ;) ). They are pretty
easy for me to generate, and they will definitely improve the
visual quality of the results.
Another map I will export is a "material map", which
contains information about the original model's lighting parameters.
It will be
useful when used in conjunction with a shader. |
| |
|
| Q8 |
In
which form is Simplifikator marketed? |
| A8 |
The
simplifikator is marketed as a stand-alone product. I am currently
looking for a distributor who is interested in distributing
the product online and in a boxed form.
The Simplifikator might also be available as a plugin for Maya(R)
and 3ds Max(R). |
| |
|
| Q9 |
What
is iWalkthru? |
| A9 |
iWalkthru
is a walkthrough system designed to make the creation of visual
appealing online walkthroughs easy. Its goal is to take any
complex scene and make it available on different viewing platforms,
be it workstation, office PC or handheld. The end user has the
choice to use either an X3D viewer, or the to-be-developed iWalkthru
viewer. Important to me is the compatibility with Web3D standards
to experience the walkthrough online, but maybe even more important
is the possibility to walk through the scenes on _any_ device
that has at least some form of hardware 3D support.
By
employing the algorithms used in the Simplifikator, the iWalkthru
software generates a multi-resolution database, which will allow
the viewing experience to scale well for devices with varying
display capabilities, and also with varying network capacities.
The goal is to cover a very broad range of display devices,
and thus make large scale virtual environments attractive for
mass distribution. I think that this goal can be achieved by
automating the content optimization and by reducing client requirements.
iWalkthru
targets applications where the value lies in enjoying the experience
of being in the environment, like virtual archeology, virtual
tourism and the like. I expect iWalkthru to be a perfect fit
for many projects in this field, especially if the original
data is too complex to be optimized manually, as it is the case
when real places get scanned in with lasers or when real places
are being reconstructed with photogrammetric methods. Given
the advances in this field, there will be a lot of content for
iWalkthru to work on in the future.
|
| |
|
| |
|
| |
|
