Since frequncy is 1/time time is changed writers on relativity confine this to the Transverse shift. That makes little sense the first attached paper shows for both the axial and transverse Doppler everything in a Fourier time series that exists for A will be observed for A/k by the observer (k=shifted frequency/original frequency).
That means the number of observed photons per second (not just each photon's frequency).
Since the axial shift has a cos(B) factor (B=observation angle), there will be a different frequency, intensity, and information transmission rate in each direction. Although the average of cos(B) for all B is zero. The variance is not, Therefore the observed variance of frequncy, wave length, number photons per second, intensity and information transmission rate are changed.
See the attached files for more details.