Part II – Generating a profile using Visual Spec


Updated 2017 January 14




Spectroscopy, Part I – Using the SSON Transmission Grating Spectrograph


Spectroscopy, Part III – Analysing spectra using Visual Spec


Spectroscopy, Part IV - Resources


Spectroscopy, Part V – The Miles catalogue of spectra




Visual Spec software can be downloaded from http://www.astrosurf.com/vdesnoux/  This website also provides access to tutorials on all aspects of using Visual Spec. A User Manual is available via the ‘?’ on the Menu tool bar.


For the purposes of this exercise we will be concentrating on the Balmer series of lines due to hydrogen listed in Table 1. You may of course choose your own file names but do;

- choose ones which easily relate to the operation you have just carried out

- remember to save the various formats in case you have to redo part of the exercise


Balmer Hydrogen line

Wavelength (Angstroms)









Table 1. Hydrogen lines


The objectives of this tutorial are to demonstrate;

- setting user preferences

- opening an image

- generating an intensity profile (binning)

- calibration

- correcting for camera response

- flattening profile

- measuring Signal to Noise Ratio (SNR)

- stacking profiles to improve SNR

- formatting the profile


Once you are familiar with the process the summary table at the end of this tutorial may suffice.


Setting user Preferences


These are accessed from the Menu tool bar via Options/Preferences, Figure 1. Note that new folders cannot be created here – that must be done using, for example, Windows Explorer. I selected ‘SSON images’ as my Working Directory by double clicking on it in the Preference window, set Image to fits and Profil(e) to spc and saved these in the usual manner by clicking on OK.



Figure 1. Preferences


Opening an image


Vspec buttons 1

Figure 2. Vspec buttons


The buttons for this section are shown in Figure 2. Vspec works with calibrated spectra (application of dark frames and flat fields – not to be confused with spectral calibration) so that process must therefore be done using other software. SSON provides calibrated images so this is not necessary. Images are opened using File/Open or the icon in the Image toolbar immediately below. For this tutorial I used image 1_330900 of Theta Aur (available here) obtained on 2014 November 5, Figure 3. As you can see the image is not quite horizontal so needs to be adjusted using the Image Rotation tool. The intensity of any part of the spectrum can be determined by moving the cursor along the image and noting the reading in the I box in the Menu tool bar. In this case the maximum intensity was just under 33000. Somewhere between half and two thirds maximum is satisfactory. If necessary the high and low image thresholds can be adjusted to better display the spectrum. Visual Spec requires the blue end of the spectrum to be to the left of the image and red to the right – if necessary use the Mirror image button to reverse the image.


Figure 3. Image of the spectrum of Theta Aur


Generating an intensity profile (binning)


There are probably several spectra in the image so it is necessary to select the correct one. The correct spectral image is selected using the Display reference binning zone binning button and dragging the box from the top of the image to position it over the spectral image. Adjust the height of the box by dragging the top line to the required position, Figure 4.


Figure 4. Spectral image delineated by the box


To obtain the spectral profile click on the Object binning button. The resulting profile is shown in Figure 5 - the Y axis has been expanded to better display the profile. To do this note the high and low intensity levels, click on the Scale Y button, Figure 6, enter the values in the Hi and Lo threshold boxes, click on Apply, and then OK Save the profile as 1_330900.spc .


Figure 5. Intensity profile


Vspec buttons 2

Figure 6. VSpec Scale Y button




Because the spectrum was obtained using a transmission grating spectrometer the dispersion is non-linear, Figure 7, and therefore multiline calibration must be used for greatest accuracy. The lines listed in table 1 plus the zero order will be used. The Balmer Hydrogen lines can be identified in Figure 5 and in the SSON example, Figure 8.


Curved focal plane

Figure 7. Curved focal plane. Credit David Boyd.




Figure 8. Spectrum of That Aur indicating the Hydrogen lines


- select Spectrometry/Calibration multiple lines on the Menu tool bar

- select the first line (zero order) by dragging the vertical line across it with the left mouse button depressed

- enter 0 is in the text field to the right of the selected line and press return

- repeat for the other lines

- in the Non linear calibration window select Degre 2 and then Calcul

- the dispersion is shown together with the equation and deviation (d_lambda) – Figure 9

- save the profile as 1_330900.calibrated_multiline.spc


Figure 9. Spectrum calibration


As the cursor is moved across the profile both the pixel coordinate and wavelength are displayed in the x;l box on the Menu tool bar.


Theta Aur is a double star of spectral types, A0pSi and F2-5V. For the purposes of this exercise I selected (Tools/Library) the nearest type, a0i, from the library and dragged the file onto the spectrum. Figure 11 shows the reference library spectrum overlaying that for Theta Aur. The various hydrogen lines can be seen to match.


 Figure 11. Library spectrum overlaying that for Theta Aur


Correcting for camera response


In Figure 11 it can be seen that the shape of the observed spectrum does not match that of the reference star. This is due to the varying response of the camera over the range of wavelengths. To correct for the camera response first select a range of wavelengths from around 3000 to 10000 Angstroms;

- position the cursor between the zero image and the rising response curve

- click the left mouse button and a vertical dotted line will appear

- hold the left mouse button down and drag the cursor to the right of the response curve

- select Edit/Crop. The result is shown in Figure 12


Figure 12. Cropped response curve overlaid with library spectrum


The next step is to divide the observed profile by the reference profile;

- select ‘Intensity’ as the active profile in the box at top left of the screen

- select Operations/Divide profile by a profile

- select a0i from the Selection window, Figure 13

- click on OK and a corrected profile will be overlaid, Figure 14


Figure 13. Selection window


Figure 14. Profile corrected for camera response (plus calibrated and a0i profile from library)


The corrected profile show ‘emission’ lines where the original showed hydrogen absorption lines. To tidy up the plot select this profile;

- click on the Erase graphic icon

- select Division as the active profile from the list in the box at top left of the screen


Vspec only works with Intensity profiles so the Division profile needs to be converted to that format.

- select Edit/Replace

- in the drop down box select Intensitie

- click on OK


Vspec buttons 3

Figure 15. Buttons for next section


Now we need to convert this profile to a smooth useable camera response curve and remove the ‘emission’ lines;

- Select Radiometry/Compute continuum

- click on the Point/Zone icon

- move the cursor along the profile selecting points but missing the lines, Figure 16

- select Execute and a Continuum window will appear, Figure 17


Figure 16. Selecting the continuum


Figure 17. Continuum window


Varying the value in the box below the slider and the position of the slider should vary the smoothness of the curve.

- click on OK in the Continuum window

- delete contents of display window and bring up just the Fit.intensity profile

- convert Fit.Intensity profile to an Intensity profile by selecting Edit/Replace, Figure 18

- save as 1_330900_response_multiline.spc

- close the response profile (if you leave it open Vspec does not appear to work correctly)


Figure 18. Replace window


- open the calibrated and camera response profiles, Figure 19


Figure 19.Original and continuum profiles


- copy the response profile and paste it on to the calibrated profile, Figure 20

- crop from approx 3000 to 10000 Angstroms


Figure 20. Continuum pasted onto original profile


- select intensity as the active profile

- select Operations/Divide profile by a profile

- select 1_330900_response.spc  (the response profile appears in the window), Figure 21 and then OK


Figure 21, Division Selection window


The spectrum corrected for camera response is displayed (together with the original and camera response profiles, Figure 22. Note that the Y axis values may need changing to display the corrected spectrum.


Figure 22. Original, camera response and corrected profiles.


- delete contents of display window and bring up just the Division profile

- convert the Division profile to an Intensity profile by selecting Edit/Replace, Figure 18

- save profile as 1_330900_corrected.spc


Flattening the profile


The profile can be flattened by removing the continuum.

- open corrected profile

- select Radiometry/Continuum division

- click on the Point/zone icon

- move the cursor along the profile selecting points but missing the lines

- execute

- smooth curve

- adjust the Y axis to a Hi threshold of 1.05 and a Lo threshold of 0.85

- save profile, Figure 23, as 1_330900_final.spc


Figure 23. Final profile


Figure 24. Theta Aur profile with Hydrogen lines indicated


In Figure 24 the Balmer Hydrogen lines are labelled and their measured wavelengths compared with the actual wavelengths in Table 2. Considering that the dispersion is of the order of 9Angstroms/Pixel this is a very good agreement.


Balmer Hydrogen line

Actual Wavelength (Angstroms)

Measured Wavelength (Angstroms)













Table 2. Actual and computed wavelength comparison.


Figure 24 can be compared with Figure 25 taken from the SSON TGS User’s Guide . The original image was taken using the same exposure time, four seconds, and offset, 500.























Figure 25. Example profile from SSON TGS User’s Guide


Measuring Signal to Noise Ratio (SNR)


In his book ‘Astronomical Spectroscopy for Amateurs’ Ken M. Harrison states that a Signal to Noise Ratio (SNR) of 50 is good but for Pro-Am collaboration >100 would be acceptable. SNR can be measured using Visual Spec;

- open a profile corrected for camera response

- hold down the left mouse button and drag the mouse to select a small section of the profile devoid of lines, Figure 26


Figure 26. Profile selection for SNR determination


- select Spectrometry/Computational preferences, then SNR and OK


The SNR value is displayed and the process can be repeated a number of times to obtain an average value, Figure 27.


Figure 27. SNR values using profile corrected for camera response


Stacking profiles to improve SNR


To improve the SNR several intensity profiles can be stacked;

- open the profiles, calibrated and corrected for camera response, to be stacked

- select Operations/Compose

- the other open profiles are added to the active profile

- save this profile under a new name


For the purposes of this exercise the profile was saved under four new names, 1_330900_corrected_a, b,c and d. These were opened and summed, d being the active profile.


The stacked profile, overlaid with one of the camera response corrected profiles is shown in Figure 28. The SNRs for a single profile are >50 but those for the stacked profile are little changed on average. 


Figure 28. SNR values using profile corrected for camera response


SNR was also measured using profiles corrected for camera response and flattened to remove the continuum, Figure 29. The first three values are for a single image and the second three for a stacked image. The second set, on average, are much higher than the first.


Figure 29. SNR values using profile with continuum removed.


Formatting the profile


The profile can be formatted in various ways. As previously demonstrated both the X and Y axis can be varied using the threshold and crop features. You can also;

- add axis labels

- add a title

- change the font

- change the background colour


Remember to save the final result as a bmp file, Figure 30, or all your efforts will be lost on closing the spc file. All the features mentioned here, and others, are explained in detail in the Visual Spec tutorial at http://www.astrosurf.com/vdesnoux/tutorial4.html



Figure30. Formatted profile


Summary table


Generating a profile using Visual Spec






Image = image name


Profile = profile name







Generate profile




Rotate to horizontal


Zero image to left


Adjust thresholds


Position Binning Zone


Object Binning


Adjust Y axis





Calibrate profile


Calibrate (multiple lines)


- select 1st line, Enter


- select 2nd line, Enter


- select additional lines





Correct for camera response

- generate Response profile


Drag library profile onto calibrated profile


Crop, 3k to 10k Å approx


Select Intensity as active profile


Divide calibrated profile by library profile


Erase plots


Select Division as active profile


Edit/Replace as intensity profile



- smooth Respone profile


Select Radiometry/Compute continuum


Click on Point/Zone icon


Select points on profile (missing lines)


Select Execute


Smooth curve


Erase plots


Select Fit.Intensity profile


Edit/Replace as intensity profile




Close Response profile



- correct profile for camera response






Copy response profile on to calibrated profile




Select Intensity as the active profile


Select Operations/Divide profle by a profile


Select response intensity profile as divisor, OK


Adjust Y axis


Erase profiles


Select Division profile


Edit/Replace as Intensity profile





Flatten profile


- generate continuum profile




Select Radiometry/Continuum division


Click on Point/Zone icon


Select points on profile (missing lines)


Select Execute


Smooth curve


Adjust Y axis