# Doppler redshift in special relativity

I came across this exercise in Elementary General Relativity by Alan MacDonald:

A source of light pulses moves with speed v directly away from an observer at rest in an inertial frame. Let $\Delta t_e$ be the time between the emission of pulses, and $\Delta t_o$ be the time between their reception at the observer. Show that $\Delta t_o = \Delta t_e + v\Delta t_e$.

Based on my understanding of special relativity, the space-time interval between two events as measured from two inertial frames of reference should be the same. Therefore, $$\Delta t_e^2 = \Delta t_o^2 - \Delta x^2$$ $$\implies \Delta t_e^2 = \Delta t_o^2 - v^2\Delta t_o^2$$ $$\implies \Delta t_o = (1 - v^2)^{-1/2}\Delta t_e$$

which is not the same relation. What is wrong with my reasoning?

-
If nothing specified shouldn't the pulses move with speed $c$? – Jorge Jun 15 '13 at 14:03
@Nivalth c=1 here – user997712 Jun 15 '13 at 15:33

Your answer is right assuming $\Delta t_e$ is the interval between emission as measured by the emitting source itself. The given answer is right assuming $\Delta t_e$ is the time between emission as measured by the observer. It seems as though this problem is aiming at a lower level than your current understanding of relativity; you put too much thought into it.
But it worked it out in another way: let $E_1$ and $E_2$ be the pulse emission events, $O_1$ and $O_2$ be the observation events. Since c = 1, by definition, the interval between $O_1$ and $E_1$ is zero (they are light-like separated). Same for $O_2$ and $E_2$. This gives $(t_{O_1} - t_{E_1})^2 = x_1 ^2$ and $(t_{O_2} - t_{E_2})^2 = x_2 ^2$. Taking square roots on both sides and subtracting the equations, the expected result appears. – user997712 Jun 15 '13 at 15:29
Yes, because in this case all your times are measured in the observer's frame. The $\Delta t_e$ you get in the above comment is not the same as the $\Delta t_e$ you used in the derivation in your question. Remember, the number used for a time depends not only on the event, but on whose watch is being used to measure the event. – Chris White Jun 15 '13 at 15:35
i have thought about this and i am still confused about the basics: How is $t_{E_1}$, for instance, different as observed by the emitter and the observer? If the setting in the question were to be a real experiment, how would we measure $t_{E_1}$ as measured by the two different reference frames? I am a complete newbie to relativity so any explanations would be helpful. – user997712 Jul 13 '13 at 19:13