Hi
Tried a couple of WR type stars in Cepheus tonight as observing due south hopeless with the near full Moon.
Hopefully my response is improving! Not sure of the image size for posting as the bottom of the plots seem to cut off?
Clicking on the image brings the full size image up
Your instrument response correction does not seem to be working right. These WR stars have a blue continuum. Compare for example your results with Christian Buil's here http://www.astrosurf.com/buil/survey/wr ... stars.html
(CQ Cep = WR155, GP Cep =WR153)
As a first step, I suggest checking your calculated instrument response by reducing your reference star spectrum using it. The result should of course be very close to the library spectrum for that star.
Also there are several emission lines which appear in your spectra which are absent in Christian's and are common to both of your spectra , notably one which looks like Na D lines. I suspect these may be from the sky background. Did you do a background subtraction?
Hi Robin
Thanks again, I will get my spectra spot on one day .
Regarding the sky subtraction with ISIS I wonder if you could advise me on that please?. With my Alpy 600 the nearest I can get to horizontal is within 0.7 degrees so all my spectra fits including the Argon/Neon calibration require a 0.7 degree rotation which ISIS seems to cope with ok except the sky subtraction lines are near but not quite parallel to the spectra and could possibly include some light pollution or nearby star in the processing. Say I have a batch of 20 spectra xxxx-1 to xxxx-20, I wondered if I should rotate xxxx-1 horizontal (ie give it a 0.7 degree tweak) and therefore get the sky subtraction as tight as possible before processing the rest of the spectra or would ISIS then give the rest of the spectra a 2x0.7 degrees rotation ?? Hope you can follow that?
Regards
Steve
I am not exactly sure how the precise location of the binning and sky background zones are affected by any tilt in the spectrum but if you are using the ALPY in slit mode there is little risk of interference from other stars as only stars in the slit will appear in the spectrum. You do still need to check though that the sky background zones are well clear of anything like this just in case. The zones are adjustable though so this is normally straightforward even if a potential problem arises (This is not always the case with faint objects and slitless systems however and you need to take extra care in this case. eg as here http://www.threehillsobservatory.co.uk/ ... _zones.png
You want the light pollution to appear in the sky background zones of course so it can be subtracted from the spectrum. You should set the inner edges of these zones away from the area containing the spectrum to make sure you do not include any of the spectrum in the background. similarly you should not set the spectrum binning zone too tight. It is important to include all the rows with the spectrum in this zone. It does not matter if some background is included with the spectrum in this zone. This will be removed based on the sky background measurement. (If your spectrum is well isolated you can even set them quite a bit wider than the spectrum and use the optimised binning mode which will automatically select the specific regions within the binning zone which have the spectrum content, to maximise the SNR ).
As well as correcting any tilt in the spectrum, to get accurate removal of sky background lines it is important that the lines are exactly straight and vertical. This is done by measuring and setting the slant and smile. (ISIS does these geometric corrections independently rather than just rotating the image) If you are following the ISIS tutorials it shows how this is done automatically by selecting horizontal and vertical areas and clicking "tilt" and "smile" (Note the "smile" correction corrects both slant and smile). As a check that all is well, I usually look at the final fully processed fits image, stretching it to bring out any background. You can then see how the background including any sky lines is precisely cancelled in the region around the spectrum.
Hi again Robin
Not sure I quite described it correctly but I was wondering if say this spectrum of Mars I did the other evening (for some reason it comes out in colour with ISIS?)
was entered as the reference image as you can see it is tilted by about 0.7 degs and the sky subtraction lines can be opened to include the whole spectrum, now should I be correcting that tilt on this image and getting the subtraction lines parallel before I process with ISIS or leave it as it is?. I assume ISIS will rotate the remaining images by 0.7 degs! Leaving the image as is I seem to end up with an RMS of 0.1 to 0.2 which seems ok
best
Steve
I cannot see the binning zone in your image (there should be blue limit lines either side of the red dotted centre line). Is the inner sky zone the same as the binning limits by any chance or perhaps the binning zone is set very wide? ) Unlike RSpec, in ISIS you can set all the zones independently which stops the risk of cross contamination between the two. Attached is an example of how I would set up the limits in these circumstances. I stretch the image until I can see clearly the full extent of the spectrum and set the centre line and binning zone to just include all signs of the spectrum, including any tilt
(This is probably wider than needed as I expect ISIS does the geometric correction then applies the binning limits but unfortunately you cannot visualise this easily in ISIS. It would be better if ISIS did the geometric corrections and summed the images first and then allowed you to set the binning limits looking at the corrected and summed image - but I digress...)
I then set the sky background regions (yellow dotted lines) slightly away from the binning region to make absolutely sure of no cross contamination of the background by the spectrum.
I then measure the tilt on the spectrum and the smile/slant on a lamp or sky background line. ISIS picks these figures up and does all the necessary corrections when you press "go" producing a fully corrected image and a binned spectrum profile.
None of this will affect the wavelength calibration (provided the same geometric corrections apply to both reference and target) but incorrect setting of binning and sky zones can affect the shape of the spectrum and hence the instrument response correction.
Ah! my bad !!! I mistook the yellow dotted lines for the binning lines Doh!! So on that basis I haven`t done any sky subtraction yet
I don`t think I always pick up on the French to English translation in the tuts. As Ken says "Onwards and upwards" .
cheers
Steve
.
