Spektrophotometrische Farb-Kalibrierung
Warnung
Die Kalibrierung der Farben durch Photometrie muss unbedingt an einem linearen Bild durchgeführt werden, dessen Histogramm noch nicht gestreckt wurde. Andernfalls wird die photometrische Messung falsche Ergebnisse liefern und es gibt keine Garantie für die Korrektheit der erhaltenen Farben.
Spectrophotometric Color Calibration (Ctrl + Shift + C) is the newest method of color calibration available in Siril. This method uses the extensive spectral data available in the Gaia DR3 online catalogue [GaiaDR3]. Owing to the data volumes, this method is only available when an internet connection is available: there is no offline version of the Gaia DR3 xp_sampled spectrum database that Siril can use.
Warnung
Da SPCC auf die Verbindung zum Gaia-Archiv angewiesen ist, ist zu beachten, dass die SPCC-Funktionen von Siril nicht zur Verfügung stehen, wenn das Archiv wegen Wartungsarbeiten oder einer Störung offline ist. Glücklicherweise ist das Archiv normalerweise sehr zuverlässig, aber im Zweifelsfall können Sie den Status hier überprüfen: https://gaia.esac.esa.int/gaiastatus/.
SpectroPhotometric Color Calibration dialog window.
Tipp
Was ist der Unterschied zwischen SPCC und PCC? Wann soll man die Eine oder die Andere Variante benutzen? SPCC ist eine genauere Version der PCC und macht letztere überflüssig. SPCC berücksichtigt den Sensor und die Filter Ihres Equipments. Dadurch ist die erzeugte Farbe viel näher an der "Realität". Das Beispiel in der Abbildung unten veranschaulicht den Unterschied in den Ergebnissen.
Vergleich zwischen PCC (links) und SPCC (rechts): Zum Vergrößern anklicken. (Mit freundlicher Genehmigung von Ian Cass)
How it Works
SPCC erfordert Kenntnisse über Ihren Sensor und die von Ihnen verwendeten RGB-Filter. Diese werden über ein Online-Repository bereitgestellt, das Siril entweder automatisch beim Start oder bei Bedarf manuell synchronisiert. Sensor- und Filterinformationen werden über dieselbe Synchronisierungsmethode aktualisiert, die auch für das Online-Skript-Repository verwendet wird. (Dies bedeutet, dass Daten zu neuen Filtern oder Sensoren, sobald sie verfügbar sind, dem Repository hinzugefügt werden können, ohne dass eine Aktualisierung der Anwendung erforderlich ist.)
In the GUI you select your sensors and filters from the widgets in the SPCC dialog. Don't worry if there isn't an exact match for your equipment, just pick the closest option, or the appropriate default option. You also need to select a white reference. The default reference is the Average Spiral Galaxy reference which is suitable for a wide range of astrophotographic scenes, however there is an extensive range of galaxy and star types to choose from. The Sun's spectral type is G2(iv) so if you want to balance your image using sunlight as a white reference, you would pick Star, type G2(iv) from the list.
SPCC then uses the stellar spectra in Gaia DR3 and knowledge of your imaging sensor and filters to compute for each star in the catalogue that matches a star detected in the image by Siril the expected flux in each color channel. It then compares this with the actual flux measured in each channel using Siril's photometric capabilities.
Given the sensor and filter knowledge, SPCC computes the expected flux in each channel for the specified white reference. A robust linear fit is obtained to give the best fit of catalogue to image R/G and B/G flux ratios for each star and for the white reference. This fit is used to derive correction coefficients which are applied multiplicatively to each channel, resulting in spectrophotometrically accurate color channels.
Your image must be plate solved for SPCC to work: if it is not already, this should be done with the dedicated tool. It is important to make sure that the plate solving information is correct, as some software is known to add inaccurate WCS data to images.
Graphical Interface
Selection of Sensor In order to select your sensor, ensure that the mono / OSC toggle button is set correctly. You will then see the appropriate dropdown to choose from the available sensors.
Selection of Filters SPCC can operate in two modes.
The default mode is broadband operation. In this mode, the Narrowband mode check box should be unchecked. You can choose either red, green and blue filters (for composited images made with a mono sensor) or OSC filters, for example light pollution filters, for images made with an OSC sensor.
By checking the Narrowband mode check box, you enable narrowband mode. This is intended either for images composited from narrowband filters used with a mono sensor or for images made using an OSC sensor with a dual, tri-band or quad band narrowband filter. In this mode the available controls change, and for each color channel you enter the nominal wavelength and bandwidth of the filter passband. For ultra-narrowband mono filters the passband may be as little as 3nm; for a quadband OSC filter like the Altair QuadBand V2 the passbands may be as much as 35nm. Note that for a HOO composition where two channels are set to the same data, the nominal wavelength and bandwidth should be set equal in the SPCC interface too.
Calibrated HOO image. Image by Cyril Richard.
Tipp
Some manufacturers specify a center wavelength and FWHM. It is fine to use the FWHM as the bandwidth: these filters have very sharp cutoffs.
Warnung
Don't expect to retrieve the Hubble palette for SHO imaging using the wavelengths of the SII, \(\mathrm{H}\alpha\) and OIII filters respectively. The result will be an image with a huge green cast. This is easily explained by the fact that the SII emission line is much fainter than that of hydrogen, and the SPCC gives a representation of real intensities. But this is not the case in the Hubble palette. In fact, manual color calibration will give better results.
SHO image calibrated by SPCC compared to the same, manually calibrated one. The entire nebula was taken as a white reference during manual calibration. Image by Cyril Richard.
Selection of DSLR Low Pass Filter (LPF) DSLRs contain a low-pass filter (sometimes also called a 'hot mirror'. These reduce transmittance at wavelengths of interest to astronomers (Ha at 656nm and S-II at 674nm). If the selected OSC is a DSLR, a dropdown will be provided from which you can the appropriate LPF profile. Options exist for stock LPFs as well as astro-modified LPFs and an ideal Full spectrum filter model for if the LPF has been removed altogether.
Selection of White Reference SPCC requires an absolute white reference spectrum. The default is Average Spiral Galaxy and the source spectra used to create this white reference are taken from the SWIRE templates [SWIRE] in a manner consistent with other astrophotography software providing the same white reference. A wide range of other white references is available, covering the full range of galaxy and star classifications [Stellar]. If you wish to use sunlight as your white reference, you would choose the white reference Star, type G2(iv) as the Sun is a type G2(iv) star.
Graphs showing white reference data from spiral galaxies. At around 350 nm, the Average Spiral Galaxy data become identical to the Sc galaxies, which are also a good representation of the white reference.
NGC 4414 is a great example of a Sc-type galaxy, the type closest of the average spiral galaxy used as white reference by default. Image Credit: NASA, ESA, W. Freedman (U. Chicago) et al, & the Hubble Heritage Team (AURA/STScI), SDSS; Processing: Judy Schmidt.
Tipp
Summary of Stellar Spectral Classifications Stellar classifications have two parts, a Morgan-Keenan type and a Luminosity index.
The first part of the spectral classification (G2 in the case of the Sun) takes one of the following letters: O, B, A, F, G, K, M. O represents extremely hot blue stars, while M represents cool red stars. The sun is roughly in the middle of the spectrum. The number represents intermediate cases, for example a B5 star is halfway between type B and type A.
The second part of the spectral classification is the luminosity index ranging from i to v. Stars with luminosity index i are supergiants, whereas stars with luminosity index v are dwarfs. Main sequence stars such as the sun have a luminosity index of iv.
Graphs showing white reference data for a set of two different star classes, G and K.
Difference in color calibration depending on the choice of white reference. On the left, an M-type star, on the right the average spiral galaxy. Please note that the data are linear, and only an autostretch has been applied to the visualization.
The interface allows you to view details of the selected sensor, filter and white reference using the Details button next to each combo box. From the details information box that this brings up you also have the option to plot the Quantum Efficiency (for sensors) or transmittance (for filters) or relative photon count (for white references) against wavelength. A Plot All button is also available in the main SPCC dialog which allows you to see the responses of all your filters and your sensor and the white reference spectrum all plotted together.
Plotting all the responses of all your filters and your sensor and the white reference spectrum all plotted together
When you are happy, click Apply and SPCC will run. It will cache catalogue data but the first time you apply it to an image it will take a few seconds to perform the online catalogue searches and retrieve the source and spectral data. SPCC will then be applied to the image. Additional plots showing the linear fit of the catalogue Red / Green and Blue / Green to image Red / Green and Blue / Green ratios.
By default, Siril outputs graphs showing the fits used in the process. In this example the magnitude was limited to 17.
Tipp
How do I process L-RGB images? We recommend that you process them as normal RGB images. In fact, the L filter generally has very good transmission in the visible range and therefore has little influence on image colorimetry.
Tipp
Für Bilder, die mit einem OSC-Sensor aufgenommen wurden, empfehlen wir die Verwendung von Bayer Drizzle zur Wiederherstellung der Bildfarben. Dies gewährleistet genauere Farben, wie im folgenden Bild gezeigt wird.
SPCC in identischer Weise auf dasselbe Bild angewendet. Links: konventionelles Demosaicing mit dem VNG-Algorithmus; rechts: Bayer-Drizzle-Technik. Auf dem konventionell debayerten Bild ist ein dominanter Grünton deutlich sichtbar. **Der VNG-Algorithmus wurde für dieses Beispiel gewählt, weil die hier erläuterten Effekte stärker ausgeprägt sind. In Siril ist der Standard-Demosaikierungsalgorithmus jedoch RCD. Klicken Sie zum Vergrößern auf das Bild.
SPCC-Filter- und Sensordatenbank
SPCC basiert auf der Bereitstellung von Daten über verschiedene Filter, Sensoren und Weißreferenzen. Diese Datenbank ist kostenlos, und wir zählen auf die Gemeinschaft, die uns hilft, sie zu erweitern, damit möglichst viele Nutzer davon profitieren können. Die Datenbank und Anleitungen zu ihrer Verbesserung finden Sie unter: https://gitlab.com/free-astro/siril-spcc-database.
Saved Preferences
As most users are likely to do most of their imaging with one setup, or maybe two, it would be tedious to reselect the sensor and filters each time. The user choices are therefore automatically remembered when set and restored next time the tool is used, even if Siril is closed and restarted in between. This works using the preferences system but there is no need to use the preferences dialog to remember the set sensor and filters, it is done automatically.
The chosen white reference is not remembered: the default Average Spiral Galaxy is a suitable choice for most astronomical scenes, and alternative white references would normally be set for a specific image to draw out a particular aspect of the color.
Siril Kommandozeile
spcc [-limitmag=[+-]] [ { -monosensor= [ -rfilter= ] [-gfilter=] [-bfilter=] | -oscsensor= [-oscfilter=] [-osclpf=] } ] [-whiteref=] [ -narrowband [-rwl=] [-gwl=] [-bwl=] [-rbw=] [-gbw=] [-bbw=] ] [-bgtol=lower,upper]
spcc_list { oscsensor | monosensor | redfilter | greenfilter | bluefilter | oscfilter | osclpf | whiteref }
Quellenverzeichnis
Vallenari, A., et al. "Gaia Data Release 3-Summary of the content and survey properties." Astronomy & Astrophysics 674 (2023): A1. 99(613), 191.