1998 Keynote

James F. Blinn

Microsoft Research

25 Years

1974   Boulder          1984   Minneapolis      1994   Orlando
1975   Bowling Green    1985   San Francisco    1995   Los Angeles
1976   Philadelphia     1986   Dallas           1996   New Orleans
1977   San Jose         1987   Anaheim          1997   Los Angeles
1978   Atlanta          1988   Atlanta          1998   Orlando
1979   Chicago          1989   Boston
1980   Seattle          1990   Dallas
1981   Dallas           1991   Las Vegas
1982   Boston           1992   Chicago
1983   Detroit          1993   Anaheim

Observations

Then

Calligraphic

 

Now

Raster

Observations

Then

Mainframe or Mini $$$

 

Later

Workstation $$

 

Now

Desktop $

Observations

Then

Few users

 

Now

Everybody's getting into the act

Observations

Then

Everything was hard

 

Now

Lot's of stuff is easy

Observations

Then

Had to be clever

 

Now

More brute force

Observations

Then

CAD

 

Now

Special effects

Observations

Then

Feasibility

 

Now

Refinement/practicality

Observations

Then

Application papers

 

Now

Rendering papers

Observations

Then

Talks gave whole paper

 

Now

Talk is an advertisement for the paper

Unsolved Problems

1966 Sutherland

1977 Newell and Blinn

1987 Heckbert

1991 Siggraph Panel

1998 Blinn

Unsolved Problems

Sutherland 66

Cheap machines with basic capability

Basic Interaction techniques

Coupling simulation to display

Describing motion

Continuous tone displays

Making structure of drawings explicit

Hidden line removal

Program instrumentation and visualization

Automated sensible diagram layout

Visualizing abstractions

Unsolved Problems

Newell and Blinn 77

Increasing scene complexity

Fuzzy objects (hair, clouds)

Transparency and refraction

Extended light sources

Antialiasing

Systems integration

Unsolved Problems

Heckbert 87

Parametrizing implicit models

High quality texture filtering

Antialiasing

Shadows without ray tracing

Practical ray tracing

Practical radiosity

Frame-to-frame coherence

Automating model culling

Smooth model transitions

Affordable real-time rendering hardware

Unsolved Problems

SIGGRAPH Panel 91

Managing scene complexity (Barr)

Tools for serious modeling (Brooks)

Large-scale user interfaces (Card)

Multimedia (Clark)

Automatic graphic design (Feiner)

Robust geometric algorithms (Forrest)

Efficient and physically valid light transport algorithms (Hanrahan)

Graphics Standards (van Dam)

Unsolved Problems

Blinn 98

What does "solution" mean?

(Solved) Problems

Nonphotorealistic Rendering

Image Based Rendering

The Tyranny Of The Frame Buffer

Unsolved Problems

Blinn 98

1 Novelty

Finding something that hasn't been done yet

Unsolved Problems

Blinn 98

2 Education

Keeping track of what has been done

Dissemination of new discoveries

Avoiding temptation to reinvent

Unsolved Problems

Blinn 98

3 Systems Integration

Using all techniques in one system

Unsolved Problems

Blinn 98

4 Simplicity

Unsolved Problems

Blinn 98

5 Pixel Arithmetic Theory

Pre-multiplied alpha

Local/Global alpha

Correllated vs. Uncorrelated edges

Combining with light reflection models

Unsolved Problems

Blinn 98

6 Legacy Compatibility

TV pixels

Calculations on gamma corrected pixels

Different quantization range

3D API's

Unsolved Problems

Blinn 98

7 Arithmetic Sloppiness

Pixel Quantization Problems

Texture filtering

Lighting Models

 

(Need criteria for how accurate we need to be)

Unsolved Problems

Blinn 98

8 Anti Aliasing

Unsolved Problems

Blinn 98

9 Modeling, Rendering, Animation Challenges

Spaghetti

Unsolved Problems

Blinn 98

10 Finding A Use For Real Time 3D

Have: Games and Simulators

Need: Business, Communication, Visualization, UI

Unsolved Problems

Blinn 98

1 Novelty

2 Education

3 Systems Integration

4 Simplicity

5 Pixel Arithmetic Theory

6 Legacy Compatibility

7 Arithmetic Sloppiness

8 Anti Aliasing

9 Spaghetti

10 Finding A Use For It

The Future

Dr. Blinn Predicts:

Faster

Cheaper

More Memory

More Realistic

Seamless Integration

Run on NT

The Future

Synthetic Actors indistinguishable from Real Actors

2000 April 23 3:00PM

 

The Future

Better display screen technology

Sprite based displays

Videos distributed in layered format

Holographic projections

Direct brain input

The Future

Replace Hollywood back lots

Theme parks become virtual

Siggraph 2003 held in Virtual Reality

Anything not on the Web will be forgotten

25 Years

We've created a new industry

We've changed technical conferences

We've changed the world

We have a new medium