Curvature of space/time question/problem

[curtmorehouse]

While watching a video on youtube about space/time, it explained space/time like a fabric with a ball on it. Rolling another ball past this first ball caused the second ball to curve. I get that part. then they said that Arthur Eddington went to test general relativity by photographing a solar eclipse and proved general relativity by photographing stars that were behind the sun.
this video is here

I understand the curving of light on it's way past an object (sun in this case) but how or why does the light headed straight for the object suddenly curve away from the object before it curves back towards the object to make Eddington's photos real?

In other words, if Gravity makes the object curve towards something as it passes it, why wouldn't the object (or light from it) crash right into the object if it was on a perpendicular path towards the object?

watch the video and tell me what I'm missing please.. the part I am talking about starts at about 3:15

 

[WannabeNewton]

In the case of something like the sun or a star, to a very good approximation, the geometry of the space - time curvature is a 4 - dimensional sphere. The photons that make up the ray of light will move according to this geometry (so along the 4 - sphere in their path to put it loosely) so they will move in the fashion shown in the video until they are farther away from the sun where space - time is pretty much flat again.

 

[curtmorehouse]

I guess I don't understand what you are saying, because that doesn't explain why it moves AWAY from the sun before it curves back towards it. Also why doesn't a ball rolling towards (but not in a intersecting path) an object move away before it curves back towards it? Are the animations in the video not correct? OR just the one showing the photographing of starts behind the sun?

 

[WannabeNewton]

Yeah it moves along the geodesics of the 4 - sphere and speaking in terms of space alone, which is what the video shows, a sphere bulges out at the sides...so it goes out then in and keeps going off to infinity.

 

[cosmik debris]

I see what you are saying curtmorehouse. It is a bad graphic, yes photons going straight into the star will just go straight in, it is the photons that are glancing the star that get bent. Don't forget that photons are going out spherically from the light source behind.

 

[Daverz]

That video has some of the worst explanations of those analogies I've ever seen, analogies that are already sources of confusion for the unwary. Not surprised it's from the history channel.

BTW, the concept of spacetime comes from Minkowski's paper of 1909, and is not in Einstein's paper of 1905.