photo:ast_photography_dso
Table of Contents
deep sky astrophotography
see also:
- see advanced astrophotography tutorials:
-
- aim for skyfog to occupy lowest 1/3rd of image histogram with some gap to its left on all 3 RGB channels as using this regime will mean that when N images of T minutes are stacked they will give SNR comparable to a single image of NxT minutes.
- SNR in a 2min integrated image at dark sky sites = 30min integrated image in outer suburbia with 15x brighter sky fog
- a light pollution filter can improve SNR by 2x in outer suburbia ie. you can use half the integration time for same SNR.
- in outer suburbs, sky fog means that you can use a lower ISO and 30-120sec sub-exposures according to histogram for the aperture, but you may need a total of 2-3 hours integration time for the best SNR.
- measuring sky fog:
- use 800ISO at f/4 and expose until peak of sky fog is in middle of histogram then use:
- Mag/sq arc-sec = 13.93+2.5*log10(seconds to mid histogram at ISO 800 and f4)
- web image galleries:
introduction:
- good quality photography of deep sky objects is extremely demanding requiring relatively expensive equipment (at least $A7,000) and an obsessive-compulsive personality that can cope with managing the many fine details that need to be addressed.
- it is one of the most technically demanding forms of photography but has become accessible to amateurs with the advent of CCD cameras and the digital SLR, and particularly so with the advent of APO refractors in the 1990's which have eliminated the chromatic aberration that has plagued refractors and allowed the development of high quality, minimal aberration fast refractors highly suitable for astrophotography.
- a modest six inch aperture telescope equipped with a CCD camera will detect fainter objects than a 36“ telescope with a standard photographic camera!
- examples of results:
- Orion nebula M42:
- Rosette nebula NGC2244:
- SBIG ST2000 with Tak 106NS refractor 24x15min - this is a fantastic image!
- camera + telescope = $US5500, but to get this quality 15min exposures, may have been on a $US12,500 mount and the 3 special narrow band filters would cost an extra $US1500 and of course you need a computer.
- Omega Centaurus globular cluster:
Equipment required:
- a great imaging (non-visual) combination in 2024 is:
- Celestron RASA 8 Schmidt Astrograph Telescope = 400mm f/2.0 for cropped mirrorless cameras or astrocameras and weighs 7.7kg and is 400mm long
- plus ZWO ASI183MC Pro Cooled Color Astronomy 20mp Camera
- plus equatorial mount, etc
- plus live stacking software for almost real time images!
- see http://www.atscope.com.au Australian distributor of high end equipment
- high quality German equatorial mount (GEM):
- a robust mount with minimal vibration, excellent tracking with minimal periodic error and ability to autoguide are essential.
- examples:
- Meade LXD75 for those on a budget
- Losmandy G8 $US1500 for small scopes to 8” SCT or 6“ refractor & max. load 30lb
- Losmandy G11 $US2100 for larger scopes & max. load 60lb (NB. tripod 35lb, head 36lb) 0.5arc-sec;
- Astro Physics AP900
- the heavy weight mounts:
- Losmandy Titan - 4.5arcsec error; can transport in 7 parts;
- Astro Physics AP1200
- Mountain Instruments MI-250 US-made
- Tak NJP Temma 2 - NJP has a very expensive optional hand paddle
- Millennium Mount MMII German-made mount with specs approaching the ME, at $US7000 is a good investment and certainly worthy of consideration - +/- 4arc sec error; weighs 72lb; 120lb max. load;
- SoftwareBisque Paramount ME is the best mount that money can buy even at the $US12,500 plus cost - usually +/- 2 arc sec error; weighs 64lb; max. load 150lb; designed for robotic use;
- if you are only imaging in the outer suburbs in light pollution, you do not need as good a mount as your individual exposure times will be limited to less than 60sec anyway due to light pollution.
- a telescope optical tube with minimal aberrations, fast optical speed (eg. f/5.6-f8) and high contrast
- examples:
- ED or flourite APO refractor 3-6” diameter (60mm or more)
- Canon 200mm f/2.8L II lens for Canon dSLRs
- Canon 400mm f/5.6L vs Televue TV-60is f/6 - see here
- NB. while SCT telescopes are great visual telescopes and for planetary/lunar photography (eg. with a webcam or similar and stacking images), it's focal length is generally regarded as too long for DSO's, even with a focal reducer (and this limits sensor size possible without getting vignetting), and the fork mounts they are usually sold with tend not to be adequate for the precision guiding needed.
- low noise prime focus mounted camera capable of long exposures such as:
-
- for galaxies, star clusters and some nebulae, an un-modified dSLR will do fine
- BUT for emission nebulae, unmodified dSLR's have a IR blocking filter which prevents most of the H-alpha infrared emissions from being recorded, removing this filter allows ~3.4x sensitivity and thus allows ~1/12th the number of sub-exposures that need to be taken to get the similar results as an unmodified camera
- the Canon dSLRs are the most popular and the most commonly modified although Hap does modify the Olympus E410.
- dedicated astrophotography camera:
- examples:
- SBIG CCD camera cooled to -45degC with in-built auto-guider and filter wheel
- SBIG CCD cameras with built-in autoguiders (2004 prices):
- ST-7XE ($US2700) 0.39megapixels x 9micron
- ST-2000ME ($US3500) 2 megapixels x 7.4micron
- ST-8XE ($US6000) 1.6 megapixels x 9 micron
- ST-10MXE ($US7000) 3.2 megapixels x 6.8 micron
- monochrome CCD cameras require a colour filter wheel to allow multiple exposures in each colour to generate a final colour image:
- some use H-alpha (green), OIII (blue) & SII (red) filters
-
- consider a nebula filter:
- high quality H-alpha (green) or OIII (blue) filter to minimise light pollution & maximise emission nebulae
- autoguider:
- if using a webcam as an auto-guider, must attach it to a telescope with at least 70mm aperture to allow guiding on stars down to magnitude 6. An off-axis guider (OAG) will only allow guiding on Sirius unless the webcam is modified to allow longer exposures.
- many CCD cameras have inbuilt autoguiding.
- autoguiding software:
- image manipulation software:
imaging quality
- imaging quality can be measured by graphing the image intensity of a star against the diameter of the star's image, the width of this graph in arcseconds where the intensity is half the maximum intensity for the star is called full width half maximum (FWHM).
- professional ground-based astrophotographers strive for the best imaging quality possible and this usually equates to a FWHM value of 1.5-2.25 arc secs.
- values higher than 2.25 arc secs rapidly result in loss of contrast in the image and thus loss of resolution and detail.
- see Richard Bennion's online video on imaging quality whose experiments show effects of:
- atmospheric seeing:
- great seeing will allow reaching 1.5 arc secs, less good seeing will restrict one to 2.5 arc secs or worse
- a star at zenith may allow 1.9 but on the same night in same conditions, a star at 45deg to zenith will restrict to 3.1 arc secs
- collimation:
- only minor mis-collimation of 10 arc secs, will add 1 arc secs to your FWHM
- check collimation before a session!
- field flatness of optical system:
- Schmidt-Cassegrain scopes may have a 50% curvature in the field which results in stars at the edges being out of focus, this means that these stars may have 1 to 1.5 extra arc secs in FWHM
- so use the central part of the image or use a scope with a flat field to strive for a curvature of < 10% - eg. Tak 106, or RCOS use of a field flattener
- focus:
- focus changes with change in temperature
- a SCT focus may change by 350 microns over a 2deg C temp. drop which is enough to add 4 arc secs to FWHM!
- re-focus every 30min during a session or ensure temperature remains stable such as use of a RC
- optimise focus for a star in the centre of the field - may need to use off-target focusing where software temporarily slews telescope to a suitable star to allow focus then returns to original target.
- tracking:
- polar alignment:
- the greater the error, the more rapidly the star will drift and impact on FWHM, even alignment errors of 30 arc secs is too much - check drift alignment, and aim to be within 0.3 arc secs error
- periodic error due to imperfectly round worm gears:
- failure to employ error correction will cause drift and wider FWHM
- aim for less than 1 arc secs periodic error
- autoguiding:
- aim for 3-5 seconds guided exposure duration to decrease error by 0.2 arc secs in FWHM
- software settings to minimise unnecessary mount corrections and set max. move to 0.5 arc secs to minimise effects of cosmic ray hits
- wind and vibration:
- obviously this will really impact poorly
- use lower pier and stabilise with shock absorbers and increase weight
- flexure and dragging:
- scope flexure needs to be minimised - consider avoiding dovetails
- avoid cables causing dragging
- ensure scope is well balanced to minimise tracking errors
- now you can see why astrophotographers prefer to spend most of their money on the mount and buy an optical system that has a flat field such as a Tak 106 refractor.
Getting ready for a session:
- ensure telescope optics are collimated
- some people “hypertune” their mounts in an attempt to minimise periodic error, for example with a LX55:
- “I have disassembled my mount and performed the following procedures.
1.) Remove old grease and clean all metal parts with degreaser.
2.) Clean all plastic parts with detergent. I wasn't sure what the degreaser would do to them.
3.) Polish all metal on metal surfaces with a Dremel tool, buffing pad and polishing compound.
4.) Relube all weight bearing surfaces and metal on metal surfaces with a light coat of white lithium grease.
5.) Adjust end-play and depth of worm gears for free operation with minimal backlash.
6.) Upgrade Autostar Firmware to Version 32Ea for PEC and 3 start alignment enhancements”
- avoid windy nights that will blur the photos and contribute to poor seeing at high magnification
- avoid poor seeing nights if high magnification is to be used
- go to a dark sky site if possible or early morning when there is less light pollution and better seeing
- hopefully time it to ensure target is near zenith +/- 30 degrees to minimise atmospheric problems
- as long exposures (usually 40sec to 5min) are needed, critical attention to accurate setting up is essential to achieve good tracking:
- ensure telescope is well balanced and mount is level
- accurate polar alignment is important as although it doesn't effect tracking it will cause field rotation if not accurate
- NB. comets should be tracked themselves not adjacent stars if exposures are longer than ~60secs
- a well aligned, level, balanced mount should enable unguided images of stars using 100mm lens for indefinite time, but using a 400mm lens, the periodic error of the drive becomes visible at exposures greater than 5 min.
- train the mount to minimise backlash:
- backlash is when on starting a motor drive correction, the mount actually reverses, this can be compensated for such as on the Meade Autostar by performing the motor calibration and training procedure to teach Autostar how much to compensate for RA and Dec backlash. If you adjust the mechanicals to tighten up backlash, or change OTAs, repeat the procedure.
- ensure light pollution is minimised by selecting an appropriate region & consider using the nebula filter
taking the photos
- focusing can be problematic, ensure it is perfect - see focusing a telescope
- minimise vibrations:
- don't touch telescope or camera for 10sec prior to and for duration of exposure
- use self-timer on camera or remote control shutter release
- consider placing cardboard in front of telescope for 1st 10secs
- usually need to take quite a few photos of the same object in RAW mode to get good results:
- adequate total exposure duration for nebulae:
- the aim is to maximise the signal:noise ratio, and for images that have not become saturated and for which motion blur can be controlled:
- signal:noise ratio improves with the square root of total exposure time
- readout noise is proportional to the square root of the number of exposures taken, thus 4x5min exposures should give better S:N ratio than 20x1minute exposures
- using narrowband filters will increase contrast and visibility of emission nebulae while reducing noise from light pollution, but at the expense of reducing the signal as well and thus a longer exposure will be needed.
- typical total exposure times:
- f/6 with CCD camera 5min to 120min depending on detail required and brightness of nebula
- individual exposure duration:
- this is often trial and error
- minimum individual sub-exposure duration:
- subject cannot be too under-exposed as you will never be able to create sufficient signal:noise ratio for an optimum picture
- increase length of exposure - but do not go past the tracking limits of your system
- use a faster lens eg. f2-5.6
- use a higher ASA rating - if using film, consider gas hypersensitisation
- as long as the subexposure is long enough to register the skyfog at some 30+ times the Read Noise of your camera, then your 30*1min exposures = 1*30min exposure. This is the so-called Skyfog-Statistics-Limited regime.
- in digital cameras, many people aim to get the sky glow histogram midpoint at 10% of maximum if the camera has very low noise (eg. Canon 1 dmk2, 20D, 350XT) whereas those with noisier cameras at ISO 800-1600 (Canon 300D, 10D & Nikon D70) aim for this mid-pt being at 25-50% of maximum.
- suggested sub-exposure durations at ISO 1600 for dark skies:
- 1min at f/2.8;
- due to most people's mount limitations, this is what most aim for, hence they use the EF 200mm f/2.8L lens.
- 2min at f/4;
- 4min at f/5.6;
- maximum individual exposure duration is limited by either:
- image saturation due to either the object (signal), light pollution or noise.
- motion blur due to inaccurate guiding or inadequate mount
- film reciprocity failure - exposures beyond 1hr on film are unlikely to yield significantly more detail
- subject cannot be too over-exposed as all stars will become fully saturated and lose any color, and if there is light pollution, ensure this does not saturate the background which would make gaining a satisfactory signal:noise ratio impossible
- better to use multiple short exposures and combine them (see below)
- consider a filter to minimise light pollution (eg. red H-alpha, OIII, or other red filter) bearing in mind that if filter is not perfectly flat, it may introduce optical aberrations which may require fixing before the image can be stacked with images taken without the filter
- thus a number of photos of the object for stacking purposes
- if using a monochrome camera, will need to take LRGB frames - a number of frames with the various colour filters applied which will be used to recreate a colour image often with exposure ratios of 4:1:1:1
- see http://www.ewellobservatory.com/gallery/ for examples of high quality images and how they were taken
- dark frames (with lens covered) to remove sensor noise
- flat frames (of an evenly lit subject) to remove uneven light density caused by the optical system
image processing
photo/ast_photography_dso.txt · Last modified: 2024/05/03 15:09 by gary1