Hello again.
What I mean is a very subtle, but noticeable change in the trajectory of this particular image.
A golden colored bar can be seen to the right of the image, which appears to be strongly distorted by the fisheye lens. Taking that as a reference (which is not absolutely reliable as I'm assuming it's straight) try correcting the fisheye until you get the bar to look straight.
However, as I already said, the problem still persists.
But what is the problem? I will try to be the most clear as possible
If you are floating still in space, and you toss a marble, it will have a straight path due to the law of inertia. However, if you are in space again, but this time you are slowly rotating to your left and throw a marble: from your rotating frame of reference in which you are 'still', the trajectory of the marble will appear to curve to the right ; But not because it has stopped moving in a straight line, but because your field of vision has rotated with respect to it. So, always using your frame of reference, in the first case: you should be able to describe its path with a straight line. In the second case: you should be able to describe its trajectory (as the marble moves in your field of vision) with a curved line.
We know that stage 2 of the Falcon Heavy is rotating because we can see the dynamic movement of the lighting, and we can see the Sun, Earth and Moon enter and exit the camera thanks to this rotation (in addition to the orbital movement, but in scales very short time periods of the order of seconds like the ones we are using is negligible). It is also confirmed by observing the curved path of many debris. We also know that the cameras are attached to stage 2, and therefore share their rotating frame of reference.
Since the rotation experienced by stage 2 is relatively 'slow', it is expected that particle trajectories that last a very short time on the screen can be described with approximately straight lines, and that those that last longer on screen can be described with curved lines (because they experience the effect of rocket rotation to a greater extent).
The problem with those two attached images is that the seen path does not seem to be fully described with either a straight line or a curved line.
Those are two of my best attempts, but no matter how long or curved the line is, it fails to naturally align with all the points that describe the trajectory.
I think the main cause of this is that the entire path seemed to curve first 'down' and then 'up':
Obviously I did it in a very exaggerated way, but approximately in the green segment is where the trajectory appears to curve slightly 'downwards', and in the celestial segment, where it appears to curve slightly 'upwards'.
My current idea is that the apparent deviation of the green segment comes from an abnormality in the motion of the particle, immediately prior to the tracing of the trajectory:
I originally didn't care about this as it seemed like the particle was just bouncing off the metal tube over the nozzle and changing direction due to this collision. So, to do the trayectory I start exactly after this direction change. But when you look closely at the clip, the movement seems too subtle to be due to a crash. It immediately reminded me of electrostatic repulsion effects.
It could be that both the particle (as mentioned earlier in this thread, it is probably ice) and the metal tube on the nozzle had electrostatic charge, and that as the particle approached, it was repelled, this being the explanation for the 'abnormality 'of the green segment of the path, and the blue segment would simply be the aforementioned effect of rotation.
However, I am not very informed about how the space environment affects the generation of electrostatic charges or about electrostatic charges in general, beyond the basics; I know that metals tend to lose electrons, and that water (although I am not sure if it is water ice, or even if it isn't ice) is also usually positively charged when we talk about electrostatics. So I will be reading about the subject and also thinking in what ways it could be that these objects had been loaded, to draw better conclusions about this possibility. The english is not my lenguage, and I'm sorry if at times I don't express myself well or what I write seems confusing, I hope this time my idea was clearly expressed. Again, thanks for your answer!