Data Logging 2

One Mississippi, Again

I leaned on a few friends and colleagues for this round–four to be precise.  Everyone had two opportunities to check their internal clock against the Arduino’s on-board chip.  Assuming 500 milliseconds really are just that, here’s what I found:

All the Data

I adjusted the starts so that they all align, but they diverge quickly.  I then broke out each of the “subjects” into a separate graph to compare their first attempt to their second, after they were able to see it visually against the control.  The subjects were myself (MF), KK, GB and ML:

Some subjects were less precise in the manner in which they covered the sensor.  They were quite helpful however in the suggestions made for possible improvements to the test in the future.  KK for instance noted that it might be more fun if meter were tested at the same time.  The experiment is necessarily limited in its current design as one tires quickly of trying to keep an even pace without additional rhythmic complexity.  GB noted that there are already tools in the music industry for selecting samples that most closely match “the grid”.

Here is a summary of the variation from the control for each subject:

3 out of 4 showed marked improvement after seeing their first attempt visually.  My score was suspiciously consistent, and I’d guess that I won’t make a very good subject going forward, likely because I designed the experiment.  If anyone can bear to do this again, I’d like to add one or two more blinds to it before showing the visualization against the control.