Table of Contents

  1. Introduction
  2. GUI - Main Window
  3. Create SSAP VOTables
  4. Analysis Tools
  5. Fitting Utilities
  6. SED Spectra
  7. Photometric services
  8. Simple Line Access
  9. Theoretical Access
  10. Interoperability
  11. VOSpec Guys
  12. References
  13. User Feedback


Analysis Functions


Please click on tabs to get details or keep reading

Bisector

- This tool deduces the bisector by evaluating the midpoint between the two halves of the spectral line at an input number of flux levels. - Since outliers can produce bothering effects, for instance while calculating the peak position of the line, they must be removed.- In order to solve this issue, you can select between two options: - automatic bisector evaluation: VOSpec works with a suitable smoothed version of the input spectral line. The smoothing method depends on the input line with the aim to modify it at less as possible. To select this option click on the “Previous automated line smoothing” checkbox.- if the automated line smoothing checkbox is not selected, the bisector algorithm expects a suitable input spectral line. This option allows the user to perform by him/herself any previous smoothing, de-noising, … method.

Bisector parameters

        The distortion of the spectral lines can be characterized by two conventionally used parameters:

-         Velocity Span (or amplitude): measures the velocity difference in the bisector at two flux levels. The considered levels are the continuum and the line peak.

-         Bisector Curvature: is defined as the difference between bisector spans computed for the top and bottom halves of a line. Three bisector points are considered to calculate the velocity span values:

 


Equivalent Width

VOSpec allows the interactive calculus of the equivalent width of a given spectral line.

The formula used to evaluate the value of the equivalent width is:

 

where:

is the continuum flux. For the time being, a linear continuum is considered.

is the spectrum flux.

 

    

 

Statistics

The Statistics Window allows you to calculate some statistical values of your spectra, wavelet transforms, fits …

These are the formulas used by VOSpec to calculate each statistical value:

 Measures of central tendency

 Give us an idea of the "middle" or "typical" value of the data.

-Mean:

 Note that the mean can be drastically affected by outliers.

 

- Median:

The median is determined sorting the data set from lowest to highest values and taking the middle one. In the case of an even number of values, the median is the mean of the two middle numbers.

The median is much less affected by outliers than the mean. For this reason, the median often is used when there are a few extreme values that could greatly influence the mean and distort what might be considered typical.

 

 Measures of dispersion

 

 Characterize the spread or variability of a data set.

- Variance:

                               

 

- Standard Deviation:

It is just the square root of the variance.

Restores the units of the spread to the data units.

                               

 

- Median absolute deviation:

- Very robust estimator

- Highly insensitive to outliers

               

 

- Range: is the difference between the largest and the smallest flux values.

Tuning 

In the tuning window you can find some useful tools to perform simple modifications in your spectra.

The currently available tuning utilities are:

Multivalued Spectra Averaging

Replaces all flux values at the same wavelength position by their mean value.

 

   

 

 


Re-binning

- Increase or decrease spectra resolution

- The output spectrum is evaluated at an input number of evenly spaced wavelength positions selected by the user.

 

               

 

 

Reject zero and negative values

This utility allows you to remove zero and negative flux values from your spectra

 

           

 

Mirroring

A simple way to check the asymmetry of a spectrum section is the mirroring method. It works by evaluating the mirrored image of a given spectrum line or region with respect to an axis.

Currently there are the following available options:

 

Line Mirroring

 Given a spectral line, this method evaluates automatically its axis (peak position) and returns the mirrored line respect to this axis.

                       


Spectrum Mirroring

 

 Given a spectrum (or a spectrum selection), this mirroring option evaluates the mirrored image of the input spectrum (or the spectrum selection) with respect to the axis introduced by the user.

If no axis is introduced, the tool considers the midpoint of the wavelength range where the input spectrum is defined.

 


Flux

The Flux window gives access to Line Flux and Integrated Flux utilities. They both work by selecting in the main VOSpec display window the region in which we are interested in  (drag right button of mouse over the spectrum region on which the fitting algorithm shall be run) and then clicking in the "Calculate" button.


Line Flux

Measures the flux between a spectral line and a linear continuum.

The spectral line must be selected in the main VOSpec display window before clicking the “Calculate” button.


Integrated Flux

Estimates the area under a region of a spectrum using simple trapezoidal integration.

The region of the spectrum we are interested in must be selected in the main VOSpec display window before clicking the “Calculate” button.

 

 

                         

 

****************************************************************************************************

 

Filters


 

 

VOSpec offers many types of filtering or de-noising methods:

- Averaging FiltersAveraging_Filters

- Mean     

- Median

 

- Kernel Filtersernel_Filters

- Constant width: Kernel FiltersKernel_Filters

- Adaptive width: Adaptive FiltersAdaptive_Filters

 

- Wavelet FiltersWavelet_Filters

Averaging Filters

 

The averaging filters offered by VOSpec are:

- Mean filter

- Median filter

For both types of filtering methods there are two different options:

Defined by input number of adjacent pointsDefined_by_input_number_of_adjacent_points

Defined by input number of subintervals*Defined_by_input_number_of_subintervals

Defined by input number of adjacent points

-These averaging filters replace every flux value of the input spectrum by the mean/median value of the set composed by such input flux value plus its N backward and N forward adjacent flux values. N is the input number of adjacent points.

-The output spectrum is defined at the same wavelength positions than the input spectrum.


Defined by input number of sub-intervals

 

This option returns an output spectrum that is defined at N+1 evenly spaced wavelength positions:

 

 

where  N is the input number of sub-intervals.

 

The flux values of the output spectrum at these wavelength positions are the mean/median value of all input flux values lying within the following interval:

 


Kernel Filters

 

These filters replace each flux value by a weighted average of its adjacent flux values:

       

                                   

 

is the input spectrum evaluated at the kth position

is the weighting function (kernel) evaluated at the jth position

 

Available weighting functions (kernels) are:

-Gaussian

-Lorentzian

-Voight

-any spectrum: you can use any spectrum as kernel filter. You just have to select the spectrum that you want to use as kernel from the spectra list.

 

The width of the Gausian, Lorentzian and Voight functions:

-is constant at every position

-can be introduced as input by two different ways:

-real width

The algorithm uses the width value introduced by the user without modifications.

-fitted to range width

If the user doesn't know a suitable value for the width, the input value introduced by him/her is fitted to the range where the spectra are defined.

The value of the fitted width is proportional to the input value.

 


Adaptive IDS Filters

 

These are convolution filters where the width of the weighting functions at each wavelength position is adapted to the shape of the spectrum.

The width is narrower close to the peaks of the spectrum and wider close to the continuum.

The aim of these filters is to preserve all the spectral lines, even the smoothest ones.

 

 


Wavelet Filters

 

These filters use the properties of the wavelet transform to smooth the spectra.   

The technique works in the following way: when a data set is decomposed using wavelets, some of the resulting wavelet coefficients correspond to details (high freq.) in the data set. If those details are sufficiently small (smaller than a given threshold) , they might be omitted without substantially affecting the main features of the data set.

 

The result is a cleaned-up spectrum that still preserves the important details.

 

Available wavelet functions are:

 

-Daubechies

-Coiflets

-Symlets

 

The threshold used is the Donoho's threshold.

 

Allowed hard and soft thresholding methods.

 

Notes:

- In order to not extrapolate and to not decrease resolution, the operations are performed at all wavelength positions of every involved spectra lying within the ranges where such spectra are defined.

- Spectra re-binning is done using simple linear interpolation when necessary.

- If the spectra are defined in disjoint intervals the arithmetic operations are not performed.

- Multi-valued flux values (more than one flux value at the same wavelength position) are replaced by their mean value.

 

 

 

Normalization
Normalization can be applied to both Theoretical and not calibrated spectra in order to be adapted to real and calibrated flux.

De_Reddening
Interstellar reddening is a phenomenon associated with interstellar extinction where the spectrum of electromagnetic radiation from a radiation source changes characteristics from that which was emitted.
Reddening occurs due to the light scattering off dust and other matter in the interstellar medium. It preferentially removes shorter wavelength photons from a radiated spectrum while leaving behind the longer wavelength photons (in the optical, light that is redder), leaving the spectroscopic lines unchanged

VOSpec provides support to this phenomenon allowing to apply the Calzetti, Cardelli O’Donnel or LMC formula, depending by the case.
Download the Video

Please see the video of VOSpec De_Reddening (presented in Moscow Sept. 2006)

Luminosity Viewer
Tool to calculate the luminosity difference

Red-Shift
Select the checkbox and apply the proper redhift


VOSpec Calculator 

  

VOSpec Calculator is the tool of VOSpec to perform arithmetic operations among spectra or between a spectrum and a constant.

 

To perform an arithmetic operation you just need to:

 

- Select first operand: drag a spectrum from the Spectra List and drop it on the operational area of the VOSpec Calculator

 

- Select the operation you want to perform

 

- Select second operand:

- drag a spectrum from the Spectra List and drop it on the operational area of the VOSpec Calculator

or

- insert a constant in the constant text field and click the “Add” button. The input constant should appear now in the Operational Area.

 

- Click on the “Equals” button

 

- The resulting spectra will appear in the Spectra List of VOSpec and in the History track of the VOSpec Calculator

 

- Click the “Reset” button before performing a new arithmetic operation

 


********************************************************************************************

 



<-GO to TOP->