This is the underlying reason behind Cosimo's issues. We are trying to work out if the Step function (presented in the figures in Smartt et al. A&A submitted) are still there in the current version, and if they are an issue.
We can't see that this step function propagates into the science data. But it looks bad, and we said we would fix it in the paper for DR2.
A reasonable question is - if we do not flat field Gr13 and G11 then do the spectra have any worse S/N ? If not, then is there really any point in applying flat fields ? Hence this was Cosimo's reason for trying to reduce data without the flat fielding. **BUT COSIMO - you don't need to use the pipeline for this test.
So three questions :
can the step function be removed in the normalised flats ? (Stefano)
Do we need to flat field Gr13 and G11 ? (Cosimo)
Even if we decided that flat fielding Gr13 and Gr11 does not improve the data we definitely do need
to flat-field Gr16. In the case of Gr16, I think the counts should be high enough that we do not
need t set the flux at pixel < 400 to 1. I think that IS a bug in the code
This is the underlying reason behind Cosimo's issues. We are trying to work out if the Step function (presented in the figures in Smartt et al. A&A submitted) are still there in the current version, and if they are an issue.
We can't see that this step function propagates into the science data. But it looks bad, and we said we would fix it in the paper for DR2.
A reasonable question is - if we do not flat field Gr13 and G11 then do the spectra have any worse S/N ? If not, then is there really any point in applying flat fields ? Hence this was Cosimo's reason for trying to reduce data without the flat fielding. **BUT COSIMO - you don't need to use the pipeline for this test.
So three questions :
Even if we decided that flat fielding Gr13 and Gr11 does not improve the data we definitely do need to flat-field Gr16. In the case of Gr16, I think the counts should be high enough that we do not need t set the flux at pixel < 400 to 1. I think that IS a bug in the code