With unique features…

Astro Pixel Processor has several features which are unique and innovative for Deep Sky astrophotography:

  • advanced image registration using true optical distortion correction
  • the ability to create huge mosaics automatically by solving the registration problem of all mosaic panels as one complete problem
30 panel RGB mosaic by Scott Rosen using APP’s mosaic capability. The illumination differences between the panels aren’t corrected purposely in this case, just to illustrate the rather complex mosaic calculation. 12 panels were shot with a Canon 6D and the other 18 panels were shot with a Canon 450D, both using a 50mm f/1.8 Olympus Zuiko lens. In the center you can find Polaris, at the top the Milky Way. And at the bottom, you can find Ursa Major. Most of the 30 panels have overlap with more than 10 other frames, making this an extremely difficult situation to get perfect registration for all frames.
  • advanced image normalisation using only the data areas that exactly overlap between the reference frame and the frame that needs to be normalised.
  • a new innovative technique called Local Normalization Correction or LNC. This will greatly improve data normalization in the complete stack before actual data integration.
  • Multi-Band Blending will remove stack artefacts in regular stacks or seams in a mosaic integration.
Cygnus, From North America Nebula to the Crescent Nebula. Data courtesy of Yves van den Broek. 18 panel HaRGB mosaic shot with a WO Star 71 apo, a Nikon D810a and a cooled monochrome Nikon D600. 38 hours of exposure in 9 panels for RGB and 9 panels for H-alpha. Completely processed in Astro Pixel Processor. In detail: data calibration using bias, artificial flats (Kang Weiss modelling) and Bad Pixel Mapping, mosaic registration with dynamic distortion correction and camera calibration, data normalization, mosaic integration with multiband blending and 8th degree local normalization corrections to remove the seams of the mosaic. Light pollution correction, background calibration, HaRGB composite. Stretched with APP’s preview filter including contrast and saturation.
  • a special and unique demosaic algorithm called: Adaptive Airy Disc that will reduce green/magenta cast in your RGB data, will improve sharpness and will make your stars rounder when compared to the well-known AHD and VNG algorithms.
A comparison of debayer algorithms. From left to right: BLN, bilinear with lots of artefacts and chromatic noise. AHD,Adaptive Homogeneity-Directed showing artefacts in the background and bad star shapes. VNG, Threshold-Based Variable Number of Gradients, showing less artefacts in the background but really bad artefacts at the star borders. AAD, Adaptive Airy Disc, showing the least artefacts in the background, no artefacts at the star borders, best resolution and least chromatic noise in the background.
  • demosaic algorithms for direct processing of monochrome narrow-band data (H-alpha, SII, OIII,…) acquired with One Shot Color (color CCD/CMOS, DSLR) camera’s.
  • chromatic aberration correction as part of RGB data calibration using a special registration model that will correct all forms of chromatic aberration to a very high degree.
  • a tool to correct vignetting in your lights/stacks using Kang-Weiss models with or without a geometric factor. The geometric factor is needed for optics consisting of several optical elements.
Rosette Nebula by Ivring Pieters, Michael van Doorn, André van der Hoeven, Rob Musquetier & Ruud de Vries. 18 hours of combined Hydrogen-Alpha exposure time from 5 different telescopes and cameras.. Top left: integration result from conventional normalization of the frames. Top middle: integration result using APP’s advanced normalization mode with Local Normalization Correction enabled. Top right: integration result using APP’s advanced normalization mode with Local Normalization Correction enabled and Multi-Band Blending enabled. Bottom: for each integration result, the normalization map is shown. The normalization maps shows the dispersion of all pixel stacks in the integration. The lower the value, the better the integration. Please note that MBB removes the stack artefacts and seams and that the advanced normalization plus the LNC algorithm gives much better data normalization.