I want to start by explaning a few things about how my thought process works, I need to do this first so it puts the rest of my research in context. I had a stroke several years ago, brought on by a cardiac arrest episode I had back in 2009, I have been implanted with a pacemaker/difibrillator. I am very fortunate that I am a Veteran and get all my medical care at the VA hospital, well the stroke didn't help but I recovered from that also, the only long term side effect from that is, I have a short term memory problem that I have to work at diligently every day on, so I have to keep a log book of my day's events that I feel are important so I do not forget.

This includes much of my research, that is why I have log books especially designated just for that. I feel that its important for me to divulge this kind of very personal information because it not only helps me mentally, but hopefully will bring a little clarity to some of my remarks. So engaging in this type of work to me is very much like an athlete rebuilding himself after a severe accident, I would just ask for a little forgiveness and patience.

So I am having trouble understanding how to apply gain correction as it pertains to data processing of spectra? I have re-read all of Dave @Stoft work on the subject and I understand the concept of it but not how to apply it when I am processing data in spekwin. The help files in spekwin32 are very limited in their scope of explanation and Dr.Menges is not always available to ask these kind of questions.

So I downloaded the actual solar spectrum from the National Weather Center and I will post the link also, I produced a plot with my Solux 50W lamp, the solar spectrum and my latest CFL calibration, the plot is elow and this is were I do not know how to proceed from there?


[![ALL_3_spectrums_solar_solux_and_cfl_june_27.png](//i.publiclab.org/system/images/photos/000/016/811/large/ALL_3_spectrums_solar_solux_and_cfl_june_27.png)](//i.publiclab.org/system/images/photos/000/016/811/original/ALL_3_spectrums_solar_solux_and_cfl_june_27.png)

Here is the link, scroll to the bottom of the page to download the xls file:

http://www.pveducation.org/pvcdrom/appendices/standard-solar-spectra

**Comment #1**

Step 1: Obtain a 4700K Solux which 1) has a much more uniform spectral response (flatter, smoother, less dynamic range over 400-800nm) than the 3500K and 2) has measured spectral data for use in performing gain correction. Then .......

Step 2: Measure that 4700K Solux with the spectrometer while keeping the peak RGB just shy of clipping for the best SNR.

Step 3: The idealized relative linear gain correction is then (PLab Measured data (from Step 2) minus the Solux 4700K data from Step 1). This could be computed just once in a spreadsheet.

This obtains the theoretical gain correction which then requires some limitations at the spectral ends where the correction value "blows up" because the errors get large due to camera sensitivity falling to near zero.

-@stoft

hey Dave, ok, now from step 1 aside from getting the 4700k lamp itself, when your talking about that it has measured spectral data are you saying that it comes with the lamp or is on their website and after capturing the data from the 4700k solux, that then I subtract that spectrum from the data from step1?

Did that make sense? @stoft