Thursday, August 9, 2012

What's wrong with this photo answer:


Like I said, the interesting part of the picture has to do with how the Lynx system constructs the finish photo and it is all driven by one thing… ADVERTISING DOLLARS $$$$$$$.  Surprise, surprise.  In this case, those dollars are coming from Omega.  You can plainly see the like 15 inch tall Omega banner sitting next to the track in the finish photo, however, that banner is nowhere to be found on the track (almost).

We’re all very used to seeing pictures of things.  Cameras, computer monitors, our own eyeballs all see things in the coordinate frame below defined by position or distance (meters, feet, miles, etc.).

The tricky thing is that the Lynx finish photos use a different coordinate system in that TIME is on the horizontal axis, instead of position.  This change is confusing to us humans because we’ve seen billions of images the normal way (with dimensions of position), but only a handful of images with time as one of the dimensions.

To clarify a little bit further (hopefully) I’ll explain another way.  The Lynx cameras take very skinny pictures of just the front edge of the finish line at a very high rate (2 kHz or every 0.005 sec).  They then arrange these from right to left to create a normal width image of what was going on at the finish line as time progressed. 

This means that if an advertiser, in this case Omega had spent big bucks to have a logo at the line, in the finish photo it would only show about 1cm of the logo, repeated over and over left to right.  Someone pretty smart came up with a way around this problem and it is circled in the photo below.  

The white-ish strip in the middle of that box displays skinny portions of the Omega/London 2012 logo that change in time, synchronized with the Lynx camera rate.  This allows the final compiled finish photo to appear to have a banner across the back with the advertiser’s logo on it.  (A similar effect could probably have been achieved using a bright green stripe and adding the logo in after the fact similar to the rectangle ads behind batters at MLB games.)  I’ve also pointed out with arrows how they make the lane lines in the compiled finish photos.  Tiny pieces of black tape at the intersection of the lane and finish lines.  Clever.  Otherwise the whole field would just be all white.
Sorry if you found this post boring or tedious.  I just thought it was a cool example to point out how the Lynx system works and how creative advertisers are at finding ways to get their logos in front of our eyeballs.

7 comments:

Dan said...

Of course, someone had to design that logo advertising system and test it so that it would properly display in the finish photo. So much engineering, just for an advertisement!

The whole "time is on the x-axis" finally has me understanding exactly what we're looking at in one of these photos.

Nathaniel said...

I figured it was just "shopped" in all the finish photos or that it would be the default graphical layout for the program. Would have been cheaper too than designing their scrolling banner

c-rad said...

Fascinating stuff Mike. This reminds me of a documentary I saw about a time in our evolution when our technology exceeds our understanding of true intelligence, and it takes the skeptical eye of a police detective with a haunted past and a dependency on technology himself, to realize that intelligence - artificial or otherwise - is closely tied to emotion.

RM said...

C-rad: iRobot? AI? TimeCop? I'm blanking on this doc...

Dart said...

RoboCop for NES?

http://www.youtube.com/watch?v=84IhZuT7PPE

http://en.wikipedia.org/wiki/RoboCop_(video_games)

Dart said...

Wait, reality is really trying to destroy Skynet and the Terminator-infested future on Sega's 16-bit Mega Drive Entertainment System

"RoboCop (TM) VERSUS THE TERMINATOR (TM)"

http://en.wikipedia.org/wiki/RoboCop_Versus_The_Terminator

David Ploskonka said...

Adding another "this is awesome" to the comment pile. It does seem as though overlaying the picture on a standard background would be a more conventional solution, but this solution eliminates the need for post-processing, which might actually involve adjusting some of the colors in the background to make it look more "realistic," and might also open up the possibility of tampering with the result. Clever.