An update on how the binary model is generated:

The binary model is now generated using martins rv tables and shifting the stars in the modeled binary by the primary and secondary rv values. I also use martins mass ratios in order to adjust the how much the secondary component of the binary is involved.

This gives an accurate (initial) representation of the binary in the rest frame.

To test this, I generated the following plot. The the unshifted spectrum is the resulting spectrum from apStar without any shift. The shifted spectrum is the observed spectrum from apStar but shifted by the visits VHELIO value in the apStars header.

As can be seen, the shifted spectrum is aligned with the model spectrum instead. Showing that to put the observed spectrum into the rest frame we need to shift the observed spectrum by VHELIO to bring it back to the rest frame.

So far I have been able to plot the CCF between the model spectrum and the observed spectrum with two distinct peaks showing up with the code being found here. But the differences between the official CCFs by APOGEE and mine are very clear. Their CCFs are smoother and show more characteristics than mine as shown between Figure 1 (mine), and Figure 2 (APOGEE’s).

Figure 1: APOGEE Official CCF. Shows the first visits CCF against the model grid

Figure 2: My CCF. Shows the first visits CCF against the model grid.

One thing that could give us a hint of what is making the difference between the two methods is how Figures 3 and 4 below. My plots have two distinct peaks while APOGEES only have one. I think we’ll need to figure out a way to correct for this.

Figure 3: My CCF. Still show two distinct peaks.

Here is a simple way to show all the visits of target 2M05350392-0529033. All files for several targets and their visits can be found here.

Figure 5: Shows how the graph changes per visit.

Figure 6: Shows how the Official APOGEE CCF plots change over visits.