Affiliation
American Association of Variable Star Observers (AAVSO)
Mon, 01/19/2015 - 21:19

To date, we've made a group purchase of diffraction gratings, and made arrangements with Tom Field to provide a substantial cost savings on his RSpec software.  As a continuation of our support for the SA100/SA200 diffraction gratings, we announce SpecObs, a method of submitting, searching and downloading diffraction grating images.  This is a beta version; we expect to be adding bells and whistles along the way.  Please make use of it and give feedback on this forum.

You can access SpecObs through the Data tab on the homepage.  At the bottom of the menu is a link to "SpecObs(Submit/Search Spectra) Beta."  This puts you into the top level menu of the image database.  As long as you are logged in, you can submit images.  You can also delete any images that you personally have submitted.  The object name, date and time are extracted from your FITS header, so be sure the object is named according to a recognized VSX identifier.  The obscode is picked up from your web profile.  Images can be in any of the following format/names:

.fit, .fts, .fits

native format; or .zip, .bz2 compressed

We strongly request that images be bias and dark subtracted, but not flat-fielded.  We will make this a requirement in later releases.  Note that ANY spectral images so submitted will be public, and can be downloaded by anyone, just as the photometric measures of variable stars can be accessed through the AID.

In the next week or so, we'll add standard stars to VSX so that you can upload spectra of them.  Sometime soon, we'll also give you a list of campaign targets, once I see the limitations of the gratings.

Note that this version of SpecObs is purely for diffraction grating images.  We will be working on Phase II, the inclusion of properly processed slit spectra, at a later time.

Arne

Affiliation
American Association of Variable Star Observers (AAVSO)
test images

I've submitted two diffraction grating images taken with BSM_HQ on 150102 of S Ori, so that will give you something to search for.

Arne

Affiliation
American Association of Variable Star Observers (AAVSO)
Video walk-through

 

I've made a ten minute YouTube video that shows how RSpec can be used to process Arne's first image to SpecObj.

http://www.rspec-astro.com/youtube-calibration.

This video incorporates editorial feedback provided to me by several experts on the forum. Thanks, guys, for helping me keep my foot out of my mouth! 

 

Tom 

EDIT:  Reminder: screen recordings are easier to watch if you use the icons at the bottom of the playback window to select HD resolution (using the gear icon) and the full screen icon (box icon.)

 

 

Affiliation
American Association of Variable Star Observers (AAVSO)
Submitting grating images

I'd like to ask if it would be a good idea to trim the image so that it contains only target spectrum?

Those images you submitted look really good, I made a quick "optimized extraction" to 1-D. I'm pretty surprised, how good the result actually is!

 

Affiliation
American Association of Variable Star Observers (AAVSO)
specobs

Arne,

I'm trying to upload a .fit file to specobs and keep getting the error "Oops! Looks like something went wrong on this page. AAVSO staff has been notified of the problem"

No surprise that the image is of S ORI :)

Affiliation
British Astronomical Association, Variable Star Section (BAA-VSS)
include the dispersion ?

I would suggest adding the approximate (ie linear) dispersion along with the image. For example in this image there is a fairly faint hot star (HD36342 A3iii) with clear balmer lines bottom left  which gives a dispersion of 16.4A/pixel which when applied to S Ori then gives a good fit to the molecular bands in an M reference spectrum

Note the HQA spectrum (blue) has been cropped at ~8000A to avoid contamination from the second order spectrum and has not been corrected for instrument response 

Robin

Edited to add identity of hot star

Affiliation
American Association of Variable Star Observers (AAVSO)
SpecObs

Hello Arne

SpecObs is bouncing my start and end date.  Does it take JD?  What is the approximate date of your image?

I used 2457000  and 2457040.

It looks like 150102 is the date.  Did not catch that.   Does JD work anyway?

It did not take 150101 as start nor 150102 as end.  Is this format YYDDMM or YYMMDD?

Gary

Affiliation
American Association of Variable Star Observers (AAVSO)
dates

Hi Gary,

DATE-OBS gives dates in the format

yyyy-mm-dd

and that is what the search form is looking for in this version.  You can also find all of the S Ori images by leaving the dates blank.

Arne

Affiliation
American Association of Variable Star Observers (AAVSO)
S Ori spectra

Thanks, Robin, for your detective work!  This was an image from BSM_HQ, a 65mm refractor on the roof of the AAVSO HQ, with a 60-second exposure.  While S Ori was V=9 at the time, remember how red it is - it was Ic=5 and so an easy target for a small aperture telescope.

Arne

Affiliation
American Association of Variable Star Observers (AAVSO)
cropping images

Hi Tonis,

I agree that cropping the image so that it just contains the target spectrum would make for faster upload and download times.  However, one of the major advantages of the diffraction grating is its multiplexing ability - getting many spectra, some likely to be good comparisons - simultaneously.  Even Robin found use for one of the stars, obtaining a linear dispersion for the image.  I think that, until we've taken lots of spectra, archiving the entire image will be the most preferred method of handling these images.

Robin, I agree that adding at least linear dispersion information will be useful.  George Silvis is currently researching commonly used FITS keywords for spectra so that we can mold our required set to something close to what the professional community as a whole is wanting.  Once we have that set, I will make sure that dispersion, at least through cubic terms, will be included, and dispersion will be a required header parameter for uploading.

Arne

Affiliation
British Astronomical Association, Variable Star Section (BAA-VSS)
fits header

"George Silvis is currently researching commonly used FITS keywords for spectra so that we can mold our required set to something close to what the professional community as a whole is wanting."

 

Hi Arne,

I do hope it will be compatible with the BeSS standard already in common use by amateurs. Here is the definition of the keywords used there

http://basebe.obspm.fr/basebe/Spec_spectres_BeSS_en.pdf

Robin

Affiliation
British Astronomical Association, Variable Star Section (BAA-VSS)
dispersion equation

 "Once we have that set, I will make sure that dispersion, at least through cubic terms, will be included, and dispersion will be a required header parameter for uploading. "

 

Hi Arne,

The reduced spectra in databases like BeSS, ELODIE archive, UVES bright stars, MILES etc all seem to use spectra resampled to equal bin size so just the start wavelength and bin size are needed in the header. (This is what ISIS produces for example) This is certainly the easiest to use as simple fits table readers can handle them. When working with spectra direct from pro observatories though I have come across more compex headers though like the IRAF multispec format for echelle spectrographs example which includes the dispersion equations. These are problematic if you dont have the apropriate tools though.

Robin

Affiliation
American Association of Variable Star Observers (AAVSO)
R Leo spectra

I've uploaded a few R Leo and NSV 6116 (alpha Com) spectra from BSM_HQ as well, just to have other objects in the database to search for.  If you leave the dates blank, you will see all of the uploaded spectra for a given object.

Arne

Affiliation
American Association of Variable Star Observers (AAVSO)
keywords

Hi Robin,

BeSS is one of the keyword sets that we are considering.  However, my main goal is to have spectral keywords that are recognized by the professional community, so as to increase the liklihood of use of the amateur spectra.  I want to see what is available before committing to any one set.

Arne

Affiliation
American Association of Variable Star Observers (AAVSO)
Arne,
Amateur data based on

Arne,

Amateur data based on the BeSS fits headers is currently widely used in ProAm submissions.

I've never hear of any complain/ issue/ concern from the professionals regards our data.

 

Affiliation
American Association of Variable Star Observers (AAVSO)
RX Tau spectrum, with wavelength and flux calibration

Last night I got a nice spectrum of RX Tau, a Mira variable of type M6e-M7e, currently at Vmag of about 11 (see SpecObs).

The spectrum was taken with the SA200, processed solely in AIP4Win as a dark-subtracted average stack of 10 frames, each with 20s exposure. Telescope LX200, 8" classic, and ST8 camera. The upload to SpecObs worked without any problem.

I played (did I say "played" - I think it actually feels like a lot of work in the beginning...) with RSpec, and calibrated the slightly non-linear dispersion with a cubic fit by using an earlier spectrum taken of Beta Tau (another nice star for the SAxxxs). Beta Tau displays nice Hydrogen and Fraunhofer features (the latter e.g. a deep dip at 759.4nm from Oxygen, a "telluric" line). My dispersion is about 20 Angstrom/pixel.

Finally I attempted a flux calibration utilizing a measured spectrum of Rigel and the reference spectrum in RSpec (the procedure is explained in one of the tutorial videos). The final outcome for RX Tau -  i.e. wavelength and flux calibrated - is shown in the attached picture. I also added a reference spectrum of an M7iii star (line color blue) - the closest I could find compared to the spectral type, and the spectral features line up very nicely!!

I am amazed by this little SA200! I also did not expect to be able to go so "deep". You do not have to look only at bright stars, Vmag=11 was easy.  I included the "synthesized" color spectrum underneath the picture (a nice feature of RSpec and some other software packages); note: for some reason the synthesized spectrum goes only to 780nm. However, more importantly, here you can see very well how red these Miras actually are! This combination of SA200 and RSpec is really a great learning tool.

Feedback greatly appreciated!

Cheers,

Helmar (AHM)

P.S.: I am not sure if I posted this to the best forum topic, maybe the "First night out..." would have been better?

 

 

 

Affiliation
American Association of Variable Star Observers (AAVSO)
Delta and Gamma Cas Uploaded

Hello

I just uploaded a Delta Cas and Gamma Cas SA200 images to the web page.  I have attached the 1D plots as .jpegs here.  Comments welcome.

For SA200 plots, is there any information outside of 3000 to 8000 Angstroms, or can we just crop that out when posting?  Are there special projects that justify going outside these limits?

WGR

Gary

Affiliation
American Association of Variable Star Observers (AAVSO)
SA200 spectral limits

Hi Gary,

The answer is "it depends".  Seriously, the atmosphere will limit anything bluewards of 300nm, and the second order spectrum from the grating starts overlapping the first order about 700nm onwards.  For many stars, the overlap is unimportant, but you need to be aware of it.  Limiting the range from 300-800nm is pretty safe, but my preference is to include the entire silicon response range of about 300nm-1100nm since pixels/storage are cheap.

Arne

Affiliation
American Association of Variable Star Observers (AAVSO)
SA200 spectral limits

Hi Gary (and Arne),

To support Arne's description ("it depends"cheeky), I overlayed your spectrum of Del Cas (an A5V star) with mine of RX Tau (a Mira M6e-M7e star). Here you can see that the Mira star does not show much intensity in the visible, but a lot in the near IR. If you cut off at 800 nm, you would not see that much of that spectrum. The second order (above 800 nm) would be relatively low (lower in intensity than what you see from 400-550 nm in 1st order due to the smaller efficiency in 2nd order), but would "sit on top" of the 1st order intensities form 800-1100nm (in fact what you see on the far tail of RX Tau around 1000-1100nm might be mainly 2nd order (?)).

So, here, for Mira type stars, it might make sense to show more above 800 nm. 

Now, keep in mind that these spectra shown here have not been calibrated in regard to detector sensitivity. This is - I believe - the next step in "mastering" spectroscopy, maybe similar to the "transformations" in photometry. Wavelength calibration is important and sometimes sufficient for certain goals (think of Halpha identification for certain novae), but getting a flux calibration done with the available sample star spectra (e.g. available in RSpec in the "Reference Library", also in VSpec and BASS-Project) is even more exciting! Look at the tutorial video of RSpec "InstrumentResponse" to see how that can be done.

By the way - very nice spectra you took - probably close to the best you can get with the SA200!

Hope this helps!

Cheers,

Helmar (AHM)

 

Affiliation
American Association of Variable Star Observers (AAVSO)
Spectral Limits

Hello Helmar

Thanks for the input.  Your graph really shows that Del Cas is pretty much done by 8000, but RX Tau is just getting interesting.  Thats a great comparison. 

I have been listening to the video on calculating "Instrument Respose", getting ready to jump over that hurdle.  I really love these videos.

BTW:  What aperture and exposures did you use for RX Tau?

Gary

Affiliation
American Association of Variable Star Observers (AAVSO)
Re: Spectral Limits

Hi Gary,

8" (LX200 classic) and 20 second exposures (see also my little earlier forum post "RX Tau spectrum, with wavelength and flux calibration" from Jan. 20).

Best,

Helmar (AHM)

Affiliation
British Astronomical Association, Variable Star Section (BAA-VSS)
How to get reliable data beyond 7500A

Beyond ~7500A there is increasing risk of 2nd order contamination. The light in the 2nd order with the SA100/200 is around 10-20% that of the 1st order but you also have to take into account the relative sensitivity of the instrument which is higher at 4000A say compared with 8000A. and the relative flux in the spectrum. For hot stars for example you would get large errors because the flux at the blue end is so much higher, Less of a problem for cool stars. 

I do not recommend attmepting to correct for  instrument response beyond 7500A though unless an order filter is used as you will run into  problems where the contamination from the 2nd order distorts the instrument response generated using a hot star giving errors when applied to other stars. See my comments in this recent thread (which also has some other tips for producing good instrument response correction

http://stargazerslounge.com/topic/232937-first-go-with-star-analyser/

It is possible in theory to remove the 2nd order contamination during the data reduction if the instrument response in the 1st and 2nd order is known  and I have seen this done for the Star Analyser in the past. More commonly, an order filter is used cutting off the blue end  if accurate data beyond 7500A is specifically wanted eg

http://www.threehillsobservatory.co.uk/astro/Classifying_red_stars_usin…

Robin

 

 

Affiliation
American Association of Variable Star Observers (AAVSO)
Arne,
Is this database now

Arne,

Is this database now "live"??

What are the search criteria??

 

Affiliation
American Association of Variable Star Observers (AAVSO)
Is the spectral database

Is the spectral database still in use?

I can't find any spectra?? What is the search criteria?

Can we list which objects are in the database? Wildcard entries??

 

Affiliation
American Association of Variable Star Observers (AAVSO)
Spectra Obs Beta

The Spectra Obs DB (maybe to be called the AGSD, AAVSO Grating Spectra Database) is still in Beta, below the radar; It is not in the home page Data menu.

But it still exists at http://www.aavso.org/apps/spectra/ and maybe if we demonstrate its usefulness more resource can be given it. I've tried to clarify the FITS header requirements in the help page. ( http://www.aavso.org/specobs-help )

The search page, if you put in nothing for search criteria, will show all the spectra currently included in the db.

George

 

 

Affiliation
American Association of Variable Star Observers (AAVSO)
Diego, I saw that you were

Diego, I saw that you were hitting an error trying to upload your file. Can you email it to me so I can figure out what's going wrong? My email is will@aavso.org