spectro-aras.com

Hello David,

I am using an SX-H694 camera (which has the same CCD as the Atik 460) binned 1x1 on my LISA with a 23 micron slit on a C11 at f/5.

I focus the LISA objective using the 5852 A line in the calibration lamp spectrum.

I use File Mode calibration in the General tab in ISIS with the following .lst file.
4
1.8
7383.980
7272.936
7147.042
7067.218
6965.431
6871.289
6598.953
6506.528
6402.248
6266.495
6143.063
6096.163
5944.834
5852.488
5764.419
5495.874
4965.079
4879.864
4609.567
4545.052
4510.733
4259.362
4158.590

I give the x-coord of the 5852 A line to locate the spectrum in the Calibration tab and usually get rms values around 0.1.

I hope this is useful.

David

I should also add that this line list assumes you have the newer Ar/Ne lamp in the LISA calibration unit.

David

Hello.

Yes, i am using the Ar/Ne lamp.

I tried your list, changing 1.8 to 3.6 as i am binning 2x2. This is what i got with ISIS.

Wavelength fitting error
Line #1 x = 1253.682 lambda = 7383.980 dlambda = -6.263
Line #2 x = 1223.692 lambda = 7272.936 dlambda = 2.299
Line #3 x = 1189.642 lambda = 7147.042 dlambda = 9.193
Line #4 x = 1168.001 lambda = 7067.218 dlambda = 12.310
Line #5 x = 1140.408 lambda = 6965.431 dlambda = 14.856
Line #6 x = 1130.711 lambda = 6871.289 dlambda = -42.959
Line #7 x = 1030.654 lambda = 6598.953 dlambda = 51.758
Line #8 x = 1030.653 lambda = 6506.528 dlambda = -40.663
Line #9 x = 986.748 lambda = 6402.248 dlambda = 13.084
Line #10 x = 949.618 lambda = 6266.495 dlambda = 10.229
Line #11 x = 915.812 lambda = 6143.063 dlambda = 7.499
Line #12 x = 915.812 lambda = 6096.163 dlambda = -39.401
Line #13 x = 861.406 lambda = 5944.834 dlambda = 3.463
Line #14 x = 836.064 lambda = 5852.488 dlambda = 1.716
Line #15 x = 811.791 lambda = 5764.419 dlambda = 0.586
Line #16 x = 738.128 lambda = 5495.874 dlambda = -2.785
Line #17 x = 591.957 lambda = 4965.079 dlambda = -0.335
Line #18 x = 563.374 lambda = 4879.864 dlambda = 19.647
Line #19 x = 494.007 lambda = 4609.567 dlambda = 5.135
Line #20 x = 476.190 lambda = 4545.052 dlambda = 6.298
Line #21 x = 476.190 lambda = 4510.733 dlambda = -28.021
Line #22 x = 402.236 lambda = 4259.362 dlambda = -7.884
Line #23 x = 369.564 lambda = 4158.590 dlambda = 10.238
--------------------------------------------------------------------------
Wavelength file : c:\users\yo\desktop\demetra(27-agosto-2022)\LISA_Boyd.lst
Coefficient a4 : 1.024735E-09
Coefficient a3 : -2.915287E-06
Coefficient a2 : 2.877742E-03
Coefficient a1 : 2.48436
Coefficient a0 : 2961.594
--------------------------------------------------------------------------
RMS : 23.559606

Using LISA this is what i obtain with the same spectra:

Fit curve coefficients: a0 = 2835.2555, a1 = 3.501, a2 = 0.0003, a3 = -2.5251e-7, a4 = 9.2737e-11, RMS=0.289

Hi David.
In case it can help you...Something similar happened to me.
When I tried to reduce in ISIS my first spectra taken with my DADOS (following the same process that is explained in the manual for processing spectra taken with LISA), I could not obtain a wavelength calibration with an RMS of less than 25-30. In my specific case, the problem was that ISIS did not recognize the lines of the Ne spectrum because the integration time of calibration images was not long enough.
I went from 10 seconds to 30 seconds of integration... and problem solved!

Greetings.

Thank you, Nacho.

My integration time is 60 seconds for the lamp. In former times i was exopsing just 10 seconds.

And Demetra identifies the lines spot on. Why ISIS does not?

I may try much longer anyway and see what happens. I think the lines in Demetra or David Boyd's list will not saturate easily. I will try two minutes next time.

Gracias por la idea.

Hi David,

You might need to use a slightly different linear dispersion value in the file for your specific spectrograph. Try generating a dispersion equation with those lines manually first and then use the a1 coefficient you find in the file or try the 3.501 value you got (from Demetra?)

Cheers
Robin

Hello.

Finally it seems to be a matter of the lines chosen.
Thanks to David Boyd, he came up with this list that works:
4
3.65
7724.623
7635.106
7272.936
7147.042
7067.218
6965.431
6871.289
6506.528
6402.248
6266.495
6143.063
6096.163
5852.488
4764.865
4609.567
4545.052
4200.674

I checked how Demetra works with these new lines and it also gives a good result.

With this list and changing the polynomial degree i obtained the following results, with RMS less than 1:

4º Coefficient a4 : 3.327879E-11
Coefficient a3 : -9.041458E-08
Coefficient a2 : 1.319508E-04
Coefficient a1 : 3.53652
Coefficient a0 : 2845.889
3º Coefficient a3 : 2.587154E-08
Coefficient a2 : -1.320290E-05
Coefficient a1 : 3.61256
Coefficient a0 : 2831.927
2ºCoefficient a2 : 5.588876E-05
Coefficient a1 : 3.55539
Coefficient a0 : 2846.058
1ºCoefficient a1 : 3.65102
Coefficient a0 : 2810.428

I guess Demetra and ISIS use a different algorithm for calibration lines detection.

Greetings.