I have the 2D data (shape(1125,2058) taken by a long slit spectrograph, which looks as follows: The y-axis corresponds to the distance from the center, where the center is at the most bright green part (y_index about 595). The x-axis corresponds to the wavelengths. The data shows the long-slit-spectra of an elliptical galaxy. And here is where I am getting a bit confused, I expected the red end of the spectrum (i.e. longer wavelengths) to be more bright, since ellipticals have a red continuum, due to their old stellar population. In the plot, the spectrum is clearly more bright on the left, does that mean that the wavelength goes from right to left? Of course I also extracted the wavelength array from my fits file, if I include it into the plot, it runs from left to right. But I guess since my wavelength range does not extend into the red part of the spectrum only until yellow (570 nm) this resolves the problem? After some changes one can see the absortpion lines as well as the rotation of the galaxy.
In the plot, the spectrum is clearly more bright on the left
One possibility is that this was taken with a blue-sensitive CCD, and/or the diffraction grating has a higher efficiency in the blue. In general, the recorded brightness is the product of the object's intrinsic brightness (in photons/sec per wavelength interval) and the spectral response of the system (telescope mirrors/lenses + grating + filters + CCD).
I would be grateful about tips on the colormap settings, so one can see the absorption lines and therefore also the rotation better.
You don't say how you are generating these displays, though I'm guessing it's via the Matplotlib library in Python. So, in case you're not already displaying the data with a logarithmic scaling, try that (i.e., display
np.log10(data) instead of
data. You could try other scalings as well (e.g., square root).
A second point is that you appear to be using the default "viridis" colormap of Matplotlib. This is a terrible colormap if what you want to see are faint variations in your data. I would suggest something like a rainbow color map, maybe the "turbo" color map which is a bit better than the older "jet" colormap.