benasselstine's Feed

04-05-2018 at 06:25 AM
0 Comment
Rate this Entry

Prolonged 1UP flashing

Quote Originally Posted by benasselstine View Post
I wanted to analyze the "prolonged 1UP" issue a little more, so I downloaded the high quality version of 1.05M from:

And I plotted the brightness of the flashing 1UP again.

The 'speed bumps' and 'potholes' have gone away entirely because these recordings are not video of a TV playback. The 1UP flashing is extremely regular which makes it even easier to spot irregularities, and do statistics.

Here's the chart:

The video is encoded at 30fps, so the calculation to get the number of seconds from a frame number is:
NumberOfSeconds = FrameNumber / 30

Notable events:

81050, 00:45:01
There are quite a few "quick flashes" as seen here.

151050, 01:23:50
Another kind of quick flash. These are less frequent, and more subtle to look at. This one happens at the end of a pie factory board.

194550, 01:47:32
The prolonged flash that we first saw in Jace's review video. It's very easy to spot in the chart.

286100, 02:38:56
1 player start pressed

Statistically speaking, each flash (either on or off) usually takes 7, 8, or 9 frames. To generate some counts we draw a line through the center of the plot, and anything over the line is on, and anything below is off.

Name:  1.05M-frames-on-and-off.png
Views: 25
Size:  13.4 KB

You can see just how much of an outlier the prolonged flash is. What's up with that? You can also see that there aren't any other really long flashes in the 1.05M recording.

It might seem strange that we get lots of 1s and 2s, but this is what happens when the line in the plot hovers around the mean brightness for all frames.

So far, pressing '1 player start' is the only thing I can find that affects the flashing.
Likesdatagod liked this post
Join us