As of last Friday, the current daily build of TheSkyX with Camera Add-On has a significant update for Canon DSLR users. I'm calling this version 1.0, as the previous releases were really experimental beta's, and now I've settled on how this plug-in is structured and will move forward. I'm already working on version 1.1 too! There are of course release notes for this plugin, but they are buried where most users will not go digging for them, so I thought I'd write something up here explaining how to use it with our Camera Add On, and give a VERY high level overview of how the preprocessing workflow looks for one shot color imagers. First, some general notes about this release.
A DSLR is not a CCD Camera and was not designed for astronomical use. However, many are getting good results with DSLR’s, and we are pleased to offer support for the Canon cameras. There are a few caveats for using a DSLR for astronomical imaging with this plug-in.
- If your camera has a B (bulb) mode on the manual dial, you must always use the Bulb mode setting, even for exposures less than 30 seconds. Setting the dial to M may work on some model for exposures less than 30 seconds, but this is not reliable across all models. The Bulb mode is very fast compared to past beta’s and should be accurate enough for exposures down to ½ a second. If the mode dial is set to anything other than Manual or Bulb, TheSkyX will not be able to connect to the camera. If your model does NOT have a Bulb mode, only then will Manual work for all exposure lengths.
- Mirror lock is not supported at this time. It is currently being looked into for a future release. If the mirror lock feature is enabled, TheSkyX will report an error when you try to connect to the camera, complaining about the camera settings.
- In Camera Dark subtraction does work. Bear in mind if you have this feature of your camera enabled that it will double exposure times. TheSkyX does not currently know how to deal with this, and the countdown will end prematurely. This should not cause an error, but may look like a camera hang. Use this feature with caution and awareness of this situation.
- This plug-in supports ISO settings from 100 up to 3200, in steps that double the image intensity. You should be aware that ISO does NOT increase the sensitivity of your chip, it only provides an on chip amplification of the signal. This amplification typically adds more noise than a linear stretch in most post processing software. ISO settings are most useful for JPG shooting and are intended to emulate chemical film sensitivities. Conversely low ISO’s tend to suppress signal. A good rule of thumb is ISO 400 to ISO 800 for minimal noise and a good tradeoff for photon’s to ADU’s. The ideal ISO varies from camera model to camera model and this may require some experimentation or research on your part. The 5D Mark II with larger pixels actually does the least stretching of data at ISO 1600, so your milage may vary.
- Binning is not supported. A software bin was available in the beta releases, but binning makes no sense for OSC color data. The luminance modes (described below) provide a clean 2x2 pseudo-bin that should be acceptable for image link (plate solves).
There are three acquisition modes in the plug-in currently:
Raw Bayer Data – This mode returns the raw Bayer matrix data from the camera. Raw Bayer data contains four pixels cells that contain one red, one blue, and two green pixels (laid out in RGGB order). The image returned to TheSkyX is in the native resolution of the camera. Image link works quite well on this data despite that it appears to be missing pixels (the blue pixels tend to be very dark). When magnified you will see a characteristic screen door type pattern.
Luminance (true) – This mode works best for automated TPoint calibration and Image Link. In this mode, the Bayer Matrix RGGB pattern is converted using the standard algorithm to convert RGB to a luminance value (the two greens are averaged). Each 2x2 cell of the Bayer matrix is converted to a single pixel (this is typically called a “super pixel”), so be aware the resulting FITS file is ½ the native resolution of the camera.
Luminance (summed) – This simply sums the RGGB values of the Bayer matrix for maximum signal strength. This may improve plate solves some over the previous mode especially for fainter star fields. Canon DSLR’s do not support hardware binning. This software-binning mode simply sums neighboring pixels. It has the desired over all effect, but isn’t quite as effective at read noise reduction. This mode will also be valuable for DSLR’s using a supplementary Ha narrow band filter, as the resulting image should be monochrome anyway, and it will collect the maximum amount of signal available.
Using the summed luminance mode, you could use your DSLR like a monochrome imager, although it's certainly not going to be very sensitive compared to a cooled CCD camera. For one shot color, the work flow has the primary advantage of not having to align images taken with different filters, possibly with focus shift between them. You also ALWAYS get the same number of frames for each color <g>. You shoot in the RAW Bayer mode, and the image when zoomed in looks like it was taken through a screen door. You'll do your darks and flats in the same mode, and do image calibration without converting the image to color. Only AFTER calibration do you convert the Bayer pattern to color. This is called variously, "DeBayer", "DeMosaic", or "Reconstruction" depending on your post processing package. Only AFTER converting to color, do you now do your star alignment for stacking the images. Once you have your stacked RGB image, it's time for post processing with something like Photoshop. PixInsight btw does ALL of this as well.
I shoot with both CCD and DSLR cameras and they both have their places, their advantages, and drawbacks. I'm really pleased that our toolset now supports both as first class citizens in the image acquisition workflow.
02-17-2013 10:33 AM