There are two basic objects in pysynphot:

A source spectrum can be created from:

A bandpass can be created from:

  • FITS or ASCII table file
  • Wavelength and transmission arrays
  • Pre-defined analytic forms (box)
  • Observation mode string expression (e.g., "acs,wfc1,f555w")

A source spectrum can be:

A bandpass can be:

Flux (for source spectrum) and throughput (for bandpass) values are not actually calculated they are sampled. Evaluation on an as-needed basis enables accurate sampling, especially for asymptotic curves (e.g., power-law) and composite spectra.

Both source spectrum and bandpass can be written out to FITS or ASCII tables (see File I/O).

An observation can be created using a source spectrum and a bandpass. An observation is a special kind of spectrum that enables effective stimulus (including count rate) and wavelength calculations. It has two different datasets:

  • As defined by the native wavelength set, which is constructed when combining the source spectrum and the bandpass.
  • As defined by the binned wavelength set, which uses the optimal binning for the detector (as used by its countrate() method).

While units can be changed, unless explicitly stated otherwise, all calculations are done in pre-defined internal units:

  • Angstrom for wavelength
  • photlam for flux

Below are some items that are commonly mentioned in pysynphot:

  • obsmode - Passband created from a specific HST instrument configuration, a.k.a. observation mode.
  • form - Flux unit of the output data. For example, flam, counts, or obmag.
  • waveset - Wavelength array on which passband and spectrum will be calculated.
  • ref - Reference data parameters, which include graph, component, and thermal tables, telecope collecting area, and default wavelength set. To access them, use setref() or showref(). They are analogous to refdata in IRAF STSDAS SYNPHOT.

Passband and Spectral Computations

The tables below summarize some main functionality of pysynphot. These are only for quick reference. Detailed explanations are available in their respective sections in other parts of this document.

Create a passband:

Command Description
bp = S.FileBandpass(filename) Load from file.
bp = S.ObsBandpass(obsmode) HST observation mode.
bp = S.Box(mu, width) Box centered at mu with given width.

Calculate bandpass parameters:

Command Description
bp.avgwave() Average wavelength of bandpass.
bp.efficiency() Dimensionless efficiency.
bp.equivwidth() Equivalent width of passband.
bp.rectwidth() Rectangular width of passband.
bp.rmswidth() RMS band width as in Koornneef et al. 1986 (page 836).
bp.throughput.max() Peak throughput of passband.
bp.pivot() Pivot wavelength of passband.
bp.unit_response() Unit response; flux (in flam) that produces 1 count/second in the passband.
bp.thermback() Expose the thermal background calculation presently hidden in observationmode. Only bandpasses for which thermal information has been supplied in the graph table supports this method.
bp.photbw() RMS band width compatible with SYNPHOT calculation.

Create a spectrum:

Command Description
sp = S.FileSpectrum(filename) Load from file.
sp = S.BlackBody(Teff) Blackbody spectrum with specified temperature, Teff, in Kelvin. The flux of the spectrum is normalized to a star of solar radius at a distance of 1 kpc.
sp = S.FlatSpectrum(value[, fluxunits]) Flat spectrum with constant flux of given value and optionally flux unit.
sp = S.PowerLaw(refval, expon[, fluxunits]) Power-law spectrum of the form f = (\lambda / refval)^{expon}. The spectrum is normalized to a flux of 1 (in given unit) at refval (in Angstrom).
sp = S.GaussianSource(totflux, mu, fwhm[, fluxunits]) Emission line centered on wavelength, mu , with a Gaussian profile that has given FWHM and total flux in given unit.
sp = S.Icat(catalog, key1, ...) Interpolate a spectrum from given catalog, selected by given search criteria (key1, ...) that could be temperature, surface gravity, or metallicity.
sp = S.ArraySpectrum(wave, flux [, waveunits, fluxunits, name]) Create from given wavelength and flux arrays, in given units and name.

Create an observation:

Command Description
obs = S.Observation(sp, bp) Given spectrum as observed through given bandpasss.

Calculate observational parameters:

Command Description
obs.countrate() Calculate the response of a HST instrument for the given model spectrum and passband.
obs.effstim(fluxunit) Calculate the effective stimulus in given unit.
obs.efflam() Calculate the effective wavelength. This is performed on the binned wavelength set by default.
obs.efflam(binned=False) Calculate the effective wavelength. This is performed on the native wavelength set.

Modify a spectrum:

Command Description
sp2 = sp.renorm(value, fluxunit, bp) Renormalize the spectrum to given flux value in given unit over the given passband. The evaluator computes the integral of the spectrum over the specified passband and rescales it by appropriate factor, forcing the integral to have the requested value.
sp2 = sp.redshift(z) Redshift a spectrum by the amount, z.
sp2 = sp * S.Extinction(ebv, law) Apply an extinction of given E(B-V) using the selected extinction law.

Utility tasks:

Command Description
wv = S.Waveset(minwave, maxwave, dwave) Generate a wavelength set with given min, max, and delta. Alternatively, this can also be done using numpy.
S.showref() Show the current settings for graph, component, and thermal component tables, in addition to wavelength set and telescope collecting area.
S.setref(...) Override the default values by setting any or all of the supported keywords, or reset to software default if no parameters are given.
bp.showfiles() Print all the files that went into generating the passband.
bp.check_overlap(sp) Check whether the wavelength range of sp is defined everywhere of that in bp. The result can be "full", "partial", or "none".

File I/O

Source spectrum and bandpass can be read from FITS or ASCII table via FileSourceSpectrum (also callable as pysynphot.FileSpectrum) and FileSpectralElement (also callable as pysynphot.FileBandpass), respectively.

For FITS table, data is extracted from extension 1, where the first column contains wavelength values, and the second flux (for source spectrum) or throughput (for bandpass). The extension header must contain the following keywords:

In an ASCII table, wavelength and flux/throughput values must be in the first and the second columns, respectively. Wavelength must be in Angstrom. For source spectrum, flux must be in flam. All values will be read in as double-precision floating point. The ASCII file may contain blank or comment lines (defined as any lines starting with "#").

For source spectrum, regardless of file format, flux with negative values will be automatically set to zero, unless keepneg=True is set during initialization.

Both source spectrum and bandpass can be written out to a FITS table with their respective writefits() attribute, which provide options to overwrite existing file with the same name, remove redundant zero flux/throughput rows from both ends, set floating point precision, and add extra information to the primary header in extension 0.


Read a source spectrum from FITS table:

>>> sp = S.FileSpectrum('/some/place/spectrum.fits')

Read a source spectrum from ASCII table. Wavelength and flux values must already be in the units of Angstrom and flam, respectively:

>>> sp = S.FileSpectrum('/some/place/spectrum.dat')

Write a source spectrum to FITS table. Options are set to overwrite any existing file, trim redundant rows with zero flux at both ends, force double precision, and add a new keyword MYKEY1 to primary header:

>>> sp.writefits('/some/place/spectrum.fits', clobber=True, trimzero=True,
...              precision='double', hkeys={'MYKEY1':42})

Read a bandpass from FITS table:

>>> bp = S.FileBandpass('/some/place/bandpass.fits')

Read a bandpass from ASCII table. Wavelength values must already be in the unit of Angstrom:

>>> bp = S.FileBandpass('/some/place/bandpass.dat')

Write a bandpass to FITS table with default options:

>>> bp.writefits('/some/place/bandpass.fits')

Plot a bandpass:

>>> plt.plot(bp.wave, bp.throughput)

Plot a source spectrum:

>>> plt.plot(sp.wave, sp.flux)