[PIPE2D-645] Verify - changing FWHM around 10% and well-sampled Gaussian LSF Created: 22/Oct/20  Updated: 18/Mar/21  Resolved: 18/Mar/21

Status: Done
Project: DRP 2-D Pipeline
Component/s: None
Affects Version/s: None
Fix Version/s: None

Type: Task Priority: Normal
Reporter: ncaplar Assignee: ncaplar
Resolution: Done Votes: 0
Labels: None
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Attachments: PNG File 15xSampling.png     PNG File 1xSampling.png     PNG File 2d_view_of_FWHM.png     PNG File 2d_view_of_residual_FFT.png     PNG File 2d_view_of_residual.png     PNG File 2xSampling.png     PNG File 3xSampling.png     PNG File example_lower_center.png     PNG File example_upper_left.png     PNG File Fabrice_example_fiber255_wv_763.png     PNG File Fabrice_example_fiber2_wv_966.png     PNG File Recreation_Fabrice_plot.png    
Story Points: 2
Sprint: 2DDRP-2021 A, 2DDRP-2021 A 2, 2DDRP-2021 A3

 Description   

As shortly discussed on SPS telecon on October 21, 2020 and on Slack DRP-2D, fmadec reports a) changing FWHM in the data on the level of cca 10% and b) LSF being described with a well-sampled Gaussian c) Are the wings really as strongly suppressed as a Gaussian?

 

The report is at https://sumire.pbworks.com/w/file/fetch/141725733/LAM_Subaru_SM1_LSF_20201013.pdf

 

arnaud.lefur reports that the visits used are:

medRes,argon:25914..25937
medRes,neon:25938..25961
lowRes,argon:26524..26547
lowRes,neon:26548..26571 

 

Verify these claims

 Comments   
Comment by ncaplar [ 09/Jan/21 ]

I have analyzed low res data. I have added a recreation of the plot made by fmadec using a different set of data (but still Subaru). The FWHM has been estimated by fitting a Gaussian to the data and measuring FWHM of the modeled Gaussian. I see the same effect that Fabrice reported. I have also added 2d plots showing how FWHM depends as a function of the position on the detector. Two panels show results that depend along which axis I extract the data - the left panel is extracted along the wavelength dimension (so actual LSF), while the right panel is extracted along the perpendicular dimension.

Obviously, there is quite a bit of change. I will discuss this with Jim.

 

Comment by rhl [ 09/Jan/21 ]

How good is the Gaussian fit?  You could fit a Gaussian to a top-hat and get widths like that, but that wouldn't mean that the data was well-sampled.

 

Comment by ncaplar [ 09/Jan/21 ]

Yes, yes, this was just an interim report. Work ongoing...

Comment by ncaplar [ 18/Mar/21 ]

This was discussed on PFS technical telecom on March 15,2021.

The conclusions is that Gaussian provides semi-good fit, but obviously there are deviations in particular in wings where Gaussian dramatically underestimates flux.
It is not clear that there is much value in quantitative analysis that I did (FFT of data-model). We agreed that it is not worth spending a lot more time on this. Closing now and possibly reopening in a new ticket if further quantitative analysis is wanted. 

Comment by rhl [ 18/Mar/21 ]

I don't think it's worth more of your time.  You've looked more carefully than Fabrice did.

Generated at Sat Feb 10 15:55:47 JST 2024 using Jira 8.3.4#803005-sha1:1f96e09b3c60279a408a2ae47be3c745f571388b.