ASAS SN 14mv

Show your spectra, your results ...

ASAS SN 14mv

Postby Mike Potter » Fri Jan 02, 2015 9:43 pm

New Year's Eve was clear here and as I always avoid going out on "Amateur's Night" it afforded an opportunity to try something different. So I installed the 150 lines/mm grating into the LHiRes and tried getting a spectrum of the recently-discovered transient object ASAS-SN 14mv, thought to be a dwarf nova in outburst. Shown are three spectra - one of ASAS-SN 14mv corrected for response, one with the continuum flattened, and a spectrum of HD 74721, which was the star I used to determine the instrumental response. I show HD74721 so that you can see the "ripples" which also appear in the spectrum of ASAS-SN 14mv - they are artifacts. Despite using a lamp flat, in which the ripples are eveident, they did not divide out. It's likely I made a mistake in processing and will revisit if these data prove useful to anyone. Fairly similar to an A0V spectrum (which is how HD74721 is classified) but without hydrogen absorption lines.

asassn_14mv_norm_2015_01_01.gif
asassn_14mv_norm_2015_01_01.gif (20.79 KiB) Viewed 4441 times

asassn_14mv_resp.gif
asassn_14mv_resp.gif (15.26 KiB) Viewed 4441 times

hd74721_respcor.gif
hd74721_respcor.gif (15.26 KiB) Viewed 4441 times


I should add - magnitude at the time was around 12th. I live just a few miles from downtown Baltimore, MD, so the biggest challenge was to subtract the lovely sodium lines in our night sky.
Mike Potter
 
Posts: 30
Joined: Sat Jan 05, 2013 9:39 am
Location: Baltimore, MD

Re: ASAS SN 14mv

Postby Jacques Montier » Sat Jan 03, 2015 10:08 am

Hello Mike,

I once encountered that ripples problem and the mistake was to have too much smoothed the instrumental response.

example :

reponse.png


There is a good C. Buil's tip to compare your response with the non smoothed one by typing @@ in the comparison window.

With that instrumental response, you get those ripples on the final spectrum.

example :

ondulations.png


May be would it be your problem too ?

Hope that would help you,

Regards
--
Jacques Montier
Societe d'Astronomie de Rennes
IAU code J23
Jacques Montier
 
Posts: 242
Joined: Sun May 13, 2012 2:46 pm

Re: ASAS SN 14mv

Postby Paolo Berardi » Sat Jan 03, 2015 11:18 am

Well done Mike! Although relatively bright, it is not an easy target.
The division by master flat should correct the ripple, it is strange this does not happen. May be an offset problem (causes over/undercorrection)? You should check bias/dark. Also a "wavelenght shift" may leave some ripple but the x-shift should be very large.

Paolo
Paolo Berardi
 
Posts: 524
Joined: Thu Sep 29, 2011 10:51 pm

Re: ASAS SN 14mv

Postby Mike Potter » Sun Jan 04, 2015 4:54 am

Well, the source of the "waves" in the spectrum seen in the third image I posted turned out to be pretty subtle, but in it's essence comes down to:
1) varying sensitivity of the Kodak chip in the ST8 wrt wavelength, and
2) movement in the spectrograph, likely at the camera coupling.

The problem is evident in the flat fields as well.

I took 30 lamp flats using a 500w halogen lamp (with the UV protective glass taken out). The peak counts in each flat was about 25000, well short of non-linear. But near the blue end the level was much less, hence the need to take lots of flats so that I got good signal in the blue end without blowing out the red. There is evident "banding" in each flat image which runs parallel to calibration lamp lines (thus at constant wavelength), but to the eye the flats all look identical. However, when I divide the first of those flats by the 30th you can easily see fringes caused by the "beat frequency" between the distance between consecutive peaks in the actual light distribution and the distance of the shift. If I divide the first flat by the second the result is essentially a null. What I found was that by moving flat #30 by -0.4 pixels I could get it to null out against flat #1. Now, what I'll need to do for the actual spectra I'm not yet sure because they were taken 20 hours earlier and pointing in a very different direction than where I point the 'scope when taking lamp flats. Of course this isn't such a big problem at high resolution because the distance, in pixels, between consecutive peaks in the ST8 sensitivity is much larger (16 times greater for the 2400 line grating) so the effect, while surely there, is much less evident. Just one more strike against the ST8 as a camera for spectroscopy. Too bad since I love the large dynamic range for emission-line objects. Note also that the entire sequence of 30 flats only took about five minutes (only 4.5 second exposures each) so something was moving enough to cause about 4 microns of motion in 5 minutes.

So I have a few ideas how to try to fix this - not because I think these data are worth all the effort, but because I'll have to have some way to deal with this if I'm going to use the ST8 camer + LHiRes for, well, LoRes spectroscopy. Or buy a new camera. Or a dedicated low/mid res spectrograph. Or both. $$$$$$$$!

Attached is one of the flats, which has been divided by a low-order 2-d surface fit to bring out the details a bit. Pretty ugly!
resp_corrected_flat.jpg


Thanks everyone for your suggestions and insights!

Mike
Mike Potter
 
Posts: 30
Joined: Sat Jan 05, 2013 9:39 am
Location: Baltimore, MD

Re: ASAS SN 14mv

Postby Francois Teyssier » Sun Jan 04, 2015 8:55 am

Hi Mike,

Nice spectrum of a cv in outburst. You can send me the fit for your data base, if you want.

About the waves, the problem is similar with ATIK 460EX, but less pronounced.
1. When computing your instrumental/atmo response, choose the lowest value as possible by successive trials (cf. Jacques's answer)
2. And, above all, carefully tune your setup in order that the continuum is ***perfectly*** horizontal.
A soon as the trace of the spectrum crosses the pixels of a line, the waves increase.

François
Francois Teyssier
 
Posts: 1231
Joined: Fri Sep 23, 2011 1:01 pm
Location: Rouen

Re: ASAS SN 14mv

Postby Robin Leadbeater » Sun Jan 04, 2015 11:45 am

Hi Mike,

Last resort?
http://www.astrosurf.com/buil/infrared/flat2.jpg

from this page
http://www.astrosurf.com/buil/infrared/obs.htm

These ripples in Kodak CCD (and in many CMOS sensors) are a real pain for low res spectroscopy, particularly for slitless (eg Star Analyser) applications where a flat cannot help. The Sony CCD are much better in this respect

Robin
LHIRES III #29 ATIK314 ALPY 600/200 ATIK428 Star Analyser 100/200 C11 EQ6
http://www.threehillsobservatory.co.uk
Robin Leadbeater
 
Posts: 1251
Joined: Mon Sep 26, 2011 4:41 pm

Re: ASAS SN 14mv

Postby Paolo Berardi » Sun Jan 04, 2015 12:31 pm

Hi Mike, perhaps ST-8 "waves" are shorter and have a larger amplitude. My ICX694 sensor (SXVR-H694 camera) shows a considerable ripple (~ 8 percent semi-amplitude, measured on flat frame, bias subtracted):

Image

If I divide flat#1 for flat#100 (taken after 13 mins in my sequence) I obtain almost a "null" frame. Shifting a frame manually by 2 pixel I find a semi-amplitude residual of about 1 percent (even small).

Surely the greatest contribution to ripple comes, in your case, from the flat taken in another observing session and in a different telescope position. I don't know if you had refocused the doublet, in case the shift could be enormous. In addition to the advice you have been given, I might suggest (you probably already know) take the flat (with tracking speed off) with optical tube toward the target sky area, starting to operate after some mechanical stabilization for thermoelastic effects (especially if you have just changed the grating).

A possible solution to reuse a master flat only for ripple elimination (not for pixel to pixel efficiency compensation and with a risk for recent modifications through the optical path, i.e. dust) is to associate a neon/star spectrum, then extract a wavelength calibrated and normalized (divided by its continuum) spectral profile of flat lamp, very little smoothed. You have to divide every target profiles (obtained with very smoothed response curve) for it. Pixel to pixel efficiency compensation could be made using a PRNU during main process. But I know this is not a "scientific" procedure... Wait for any more expert people comments!

I apologize for having continued the off-topic, perhaps we should use a dedicated topic...

All the best,

Paolo

PS: I write it during Robin answer...
Paolo Berardi
 
Posts: 524
Joined: Thu Sep 29, 2011 10:51 pm

Re: ASAS SN 14mv

Postby Robin Leadbeater » Sun Jan 04, 2015 1:46 pm

Interesting Paolo,

It seems the new Sony CCD may not be as good as the old ones in this respect. Here is my ICX285 in the ATIK 314L


Robin
Attachments
ICX285AL_full_spectrum_flat_ALPY200_ATK314.png
ICX285AL_full_spectrum_flat_ALPY200_ATK314.png (9.73 KiB) Viewed 4378 times
LHIRES III #29 ATIK314 ALPY 600/200 ATIK428 Star Analyser 100/200 C11 EQ6
http://www.threehillsobservatory.co.uk
Robin Leadbeater
 
Posts: 1251
Joined: Mon Sep 26, 2011 4:41 pm

Re: ASAS SN 14mv

Postby Francois Teyssier » Fri Jan 09, 2015 9:20 pm

Hello Mike,
Here's a message of Taichi Kato on VSNET

ASASSN-14mv spectrum

I noticed this page:

http://www.spectro-aras.com/forum/viewtopic.php?f=6&t=1044

The spectrum apparently Halpha.

The strongest signal appears to be He I 4471.
The other lines are also similar to those of AM CVn.

cf.
http://adsabs.harvard.edu/abs/1975ApJ...200L..23R


I suggest you contact him

Could you send me the fit for aras base ?

Cheers,

François Teyssier
Francois Teyssier
 
Posts: 1231
Joined: Fri Sep 23, 2011 1:01 pm
Location: Rouen

Re: ASAS SN 14mv

Postby Mike Potter » Sat Jan 10, 2015 8:57 am

Hello Francois:

I'd be happy to send you the files. Let me make one last try to reduce the "ripples" and get the proper header keywords installed. Should have it for you tomorrow (Jan 10)

Mike
Mike Potter
 
Posts: 30
Joined: Sat Jan 05, 2013 9:39 am
Location: Baltimore, MD


Return to Spectra, results, information on activity ...

Who is online

Users browsing this forum: No registered users and 3 guests

cron

x