Everyone knows, or should know how much I like MythBusters. Here is the problem. Below is a picture of Adam analyzing the motion of a an exploding water heater.
Actually, I applaud Adam for his creative use of the vernier caliper. Really, it is an example of "making things work". However, in this case there is a very nice alternative - Tracker Video Analysis. So, here is what I am going to do. This will be a short tutorial on using Tracker. I will use the same video from the episode of the exploding water heater. The goal - how fast and how high did the water heater go. Or as the MythBusters put it - x and y.
I don't get that either - oh well. Here goes the tutorial.
This part should be easy for you (Adam). For the rest of us, we have to do something else. Sometimes, you can find the video on youtube. If that doesn't work, you will have to capture it yourself from your DVR. Here is what I propose (and I have said this before): make a nice webpage with all your high speed videos for educational use.
Download Tracker and start that sucker up. The first thing you need to do is import your video. Just go to Video-Import and choose your file. Next, you might want to clip it. This is optional. Find the first frame of the video that you want to look at (the frame number) and then click the video properties icon in the lower right.
This makes the video start at the interesting part. Also, if it is a really long clip, you might want to make the step size greater than one. This will reduce the number of frames you look at. You can also set the end frame and check the frame time step. This could be useful if you are using high speed video. The video file should include info on the actual frame speed, but if it is in slow motion you should change this to the real value. I can't remember off the top of my head how fast your slow motion videos are, so I will leave this.
The next step is to let the program know how big each pixel is. Hopefully, there will be something near your object of interest that you know how big it is. Here is the first frame of the video of interest.
I don't see anything useful there. Oh well, if there was something, I could just click the tape-measure tool.
This will create a blue arrow on the video that you can drag around and define the size.
Now for the marking. Create a point mass and then shift-click on the object of interest in each frame. Here is what you get.
On the right, you should see a position time graph.
The x-motion is not that interesting. So, click on the "x" on the left of the graph and change it to y: position y-component. Next, right-click on the graph and choose "Analyze..." from the context menu. This will bring up another window.
Key things (as shown by the red arrows):
- Include the "fit" check box. This will allow you to fit a function to the data.
- Check the "autofit" box to have the computer fit the best function to your data.
- In this case, I fit a linear function to the data (because it was a very short time interval). It is actually not constant position-time data because of air resistance and gravity. However, this would work well enough to get the speed.
- The slope of this line would be the speed. In this case, the number doesn't mean much. I don't know the scale of the video. It looks like it is played in real-time speed, but I am absolutely sure (Adam would know).
There you have it. Not too difficult, right? Better than a vernier caliper. Maybe I will save a more detailed analysis of the other clips for another post.
