Why would an orange spark appear blue in a video?

Question

I was testing spark generation on my car's ignition system at night and recorded the sparks on my Lenovo S856 cell phone. However while taking the video I noticed that the spark which appeared orange to my naked eye was showing up as blue on the video (with occasional bursts of orange).

Why would that happen?

Answer 1

I expect this is NOT due to the white balance of the camera adjusting to the ambient light. After all, your eyes adjust to the ambient color too. More likely it is due to two factors:

1. The spark is a broad-spectrum emitter. It emits significant light we can't see. While the camera sensor is balanced to result in close to the same color we perceive for visible light, it will react to invisible light unpredictably. This is why there are UV and IR filters over sensors. These invisible wavelengths will register on some sensors, but result in strange color. What's the right color to display for something we can't see?

   The spark is very bright compared to its surroundings, so some of its energy in the invisible part of the spectrum will make it thru the filters and register on the sensor as unpredictable colors anyway.

2. The spark is very bright, compared to the rest of the scene that the camera has set its exposure to. As a result, the middle of the spark will almost certainly be overexposed. This means that the R, G, and B values will be clipped to 1. As soon as a color channel gets clipped, the color balance is lost. If the red and green channels are more sensitive, which is often the case, then a neutral gray will have a lower blue value that the others. When all the channels have the same value, as when they are all clipped at 1, then it looks like the blue component is brighter than the others.

Answer 2

(with occasional bursts of orange).

This makes me think that the main problem here is overexposure and the phenomena which causes chroma shift is optical dispersion - which is only noticeable in everyday usage when strong light source is present in the frame (the defect is called flare) or contrast details are placed around image border (chromatic aberration).

The sizes of the spark in different wavelengths are different, the size of spark in blue (short) wavelengths is somewhat bigger than in red (long) wavelengths, and that causes the spark to appear blue for camera when red (long) wave image blows up and becomes white.

The color of the spark in this case is also dependent on how far away the objective of camera is focused: somewhat forward or somewhat backward, and where in the frame the spark is positioned: in the center or close to borders.

This is an example of warm light source which looks pink because of dispersion and overexposure.

P.S. It may also be that the larger image is formed not by short wavelengths but IR wavelengths instead which cheap cameras are known to be more prone to. IR light may have cyan appearance.