Refractor and Focal Reducer Trials
Submitted: Friday, 25th January 2008 by Paul Russell
While it cannot be denied that the release of the ED80 range of refractors coupled with popular DSLR cameras has been a boon to low cost astroimaging, with an f ratio of f/7.5, the current crop of ED scopes are still somewhat slow for imaging purposes. Along with the moderate f ratio, when using standard chip format DSLRs there is a distortion of stars brought about by field curvature at the outer edges of the field of view. This is particularly obvious in the corners of images.
To improve the f ratio it was suggested to me that possibly the Meade 6.3 focal reducer might be an acceptable option for use with the ED80. This would certainly lower the f ratio from 7.5 to 4.725 (assuming the correct separation of FR to imaging chip) and reduce exposure time for a given object and ISO by 60%. This is a significant decrease in imaging time. As well as the increase in speed, the FOV was increased from 130.1arcmin x 86.7arcmin (2.1 deg x 1.45deg) to 206.5arcmin x 137.7arcmin (3.44deg x 2.3deg). Unfortunately this increase in speed and wider FOV brought with it a much more pronounced distortion at the outer edges. Instead of looking like small ovals the stars now looked like broad triangles.
Through IceInSpace and via correspondence with similarly frustrated amateur astronomers I became aware that possibly the Williams Optics 0.8X Focal Reducer and Field Flattener might be a suitable substitute for the Meade FR. The WO 0.8 was designed for the WO range of 66mm scopes and with their smaller aperture I was a bit doubtful that full chip illumination could be achieved and there would be a significant vignette to deal with. As well, the focal reduction achieved would only be a modest reduction from f/7.5 to f/6, bringing about a reduction of imaging time by 36%. I purchased a WO 0.8X II FR (a version III has been released since aimed at the 80 – 100mm aperture scopes) and while the imaging reduction time is only modest the field flattening abilities of this FR is excellent for a focal reducer that has not been specifically matched to the ED80 optics.
Even though the results were promising, I was also interested in looking for a FR that was relatively inexpensive yet obtained reasonable results, especially for those of us on a limited budget and thought a shoot out between the available contenders would be of benefit to anyone in a similar boat to me, i.e. that wanted a good quality, value for money, focal reducer that works well in combination with an ED80 and a DSLR camera.
The Meade Focal Reducer (and similar) is priced from around $200, while the Williams Optics FR is priced from $250 in Australia. I contacted a number of suppliers of astronomy equipment here in Australia to see what other FRs were available. Apart from 1.25” reducers, only the above mentioned Meade and WO, and the Celestron 6.3 from upwards of $260 (and depending on who you talked to, different to the Meade one with better optics.) were readily available on the Australian retail market. There is also a Lumicon thread in Focal Reducer on the Australian market, but one was not available at the time of this shootout.
I had purchased FRs for other cameras and scopes before from Steve Mogg so on the off chance he might have something I decided to check his site out. Steve had a 0.66x FR for use with refractors with a 2” nose piece to fit into a 2” focuser for about $90. After contacting Steve he was more than happy to allow me to use one of his reducers in a trial of similar products.
I chose NGC104 for the trial object, principally because it was in the right area of the sky at the right time, it would be available from my limited shooting gallery for a couple of months if I was unable to complete the trials over a few nights due to weather, and obtaining clear, sharp images of the density of stars near the center would be a challenge for a DLSR/ED80 imaging setup.
All imaging was carried out through an Orion ED80 mounted on a Losmandy G11 using a Carton 60mm f/7.5 scope with ToUcam 900 for autoguiding via K3CCDTools V3. The Orion has had the original Crayford removed and replaced with a Williams Optics 10:1 focuser. All exposure timings were automated through Rob Beck’s Star-Mate . All focusing was carried out using DSLR Focus with the aim to achieve both a vertical and horizontal FWHM of 2.0 or as close as possible on a star near the center of the field of view but outside the cluster.
For each trial composite image I took 10 light frames @ ISO800 @ 60 sec exposure, along with the same number of darks for calibration. Flats were not used as I did not intend to fully process the images and the lack of flats would also assist in evaluating vignetting effects. The resultant subsets of images were processed initially in Images Plus using the automated processing function, then aligned and rotated using the same stars for all the sets of images to try to ensure alignment and combination uniformity across the samples. A simple average was used to combine each image set. All images were saved as TIFFS, then opened into Photoshop CS2 and had identical curves and levels applied to them. The idea was not to process them as such, but to simply make them more easily visible for evaluation and to gauge roughly how much increase in image brightness was achieved simply by focal reduction alone. There was no attempt of an accurate colour balance (see NB. below). I have tried to keep individual image scales the same when creating checkerboards, however at times differences in outcomes of the processing will create an illusion that NGC104 appears with more stars defined and looks “bigger” in some samples than in others. This is particularly noticeable in image 6 – “Center of Image”.
The following is a summary of each set collection:
1. Control: Orion ED80, no focal reducer.
2. Meade 6.3. FR attached via Orion Prime Focus Adapter. The silver 2” sleeve was unscrewed from the adapter then screwed into the female side of the FR. The male side of the FR was then screwed into the adapter itself. It has been suggested that the Meade 6.3 FR needs to be 105mm from the imaging surface to obtain the stated 0.63x focal reduction. I was unable to use the normal SCT T adapter to reach 105mm, due to insufficient in focus so it is unlikely that the expected focal reduction will be achieved.
3. William Optics 0.8X FR and Flattener. Attached directly to the EOS T Adapter (ring). The WO FR is setup to screw directly into a T ring. No other adapter is needed and the other end slides comfortably into a 2” focuser.
NB. These first three trials were carried out under an 10 day old moon with NGC104 still low to the SE (about 20-25 deg) so there is considerable light pollution from both moonshine and sky glow from the coast and Brisbane. I notice as well when processing these images that there is a definite red cast to each image that wasn’t noticed in later images when NGC104 was much higher in the sky(45 deg plus). Also the images were not at bright as I anticipated. I am assuming that both these problems have been brought about because of various extinction effects of the atmosphere (atmospheric density and pollutants etc).
4. Mogg 0.66 FR. Attached directly to the EOS T Adapter (ring). No other adapters required, the 2” barrel slides into 2” focuser easily. (Very similar in build to the WO, though constructed of strong rigid plastic rather than metal. If you are familiar with Steve’s adapters you will know what I mean.)
5. Celestron 6.3. FR attached via Orion Prime Focus Adapter. The silver 2” sleeve was unscrewed from the adapter then screwed into the female side of the FR. Male side of the FR was then screwed into the adapter. It is a similar situation with the Celestron Reducer as there is with the Meade Reducer insomuch that the required lens/imaging distances cannot be achieved, so it is unlikely that the expected focal reduction will be achieved.
All results here are subjective and no direct measurements on equipment or images have been made in relation to accurate amount of focal reduction, degree of distortion etc.
The 1st image is a view of the composite image for each FR for comparison of vignetting. (NB. Ignore the inequalities of light pollution.) In each set of images the native system (ie no focal reducer) is placed in the center. Images 2-5 compare each corner from a sample image with the corresponding corner from the other images. This was to view the amount of distortion produced or field flattening achieved compared to the native system. Images 6, 7 (fully processed) & 8 are of the center of the FOV, with 8 being a crop of the full size frame.
As can be seen from the composite image, vignetting is fairly equal across the range of reducers with possibly slightly less with the WO reducer and the least with the native setup.
Evaluation of edge of field
All corner images show distortion of stars to a greater or lesser degree, with the Meade and Celestron by far the worst with strongly distorted stars. The WO and Mogg adapters showing the least distortion with quite well corrected star shapes when compared to the native system. Both the WO and Mogg adapter seem to have quite good field flattening abilities though with a greater chromatic aberration with the Mogg adapter.
For viewing purposes, both the Meade and Celestron images would need to be cropped by ½ to 2/3 to achieve suitable stars across the image. Stars from 1/3 of the way out from the center to the edge of the FOV are noticeably distorted for both these reducers. Realistically though, this is not surprising when the difference in curvatures of optical surfaces between refractors and SCTs are considered. The Meade and Celestron reducers are after all designed with Schmidt Cassegrain optics in mind.
Evaluation of Center of FOV
All of the tested FR seemed to soften the image slightly as none were as sharp as the native setup. However none were excessively soft under full view, with the Mogg 0.66 being marginally the softest. The Mogg 0.66 did show some minor star elongation, brought about by chromatic aberration but was only obvious when viewing the full size image under zoom, otherwise it was unobtrusive.
Amounts of Focal Reduction
While I haven’t attempted to measure the exact amount of focal reduction for each reducer mathematically, I have compared each image with the native image in Photoshop and attempted to determine the zoom factor required to achieve a similar spacial relationship. The results are;
As my approximated reduction for the WO and Mogg reducer are very close to the advertised reductions (within 3.6% and 2.6% respectively) I will assume that the results I determined for the Meade and Celestron are relatively accurate as well. If this is the case then the achieved focal reduction in this situation, is very different to the advertised 0.63x reduction for both the Meade and Celestron reducers. As mentioned previously this result was expected due to the inability to correctly space the reducer and imaging chip.
Without doubt the William Optics 0.8x Focal Reducer and Field Flattener performed the best under the conditions at the time, with good field flattening and a moderate focal reduction with a flow on (though not measured) exposure reduction times. It showed a clear reduction of edge of field distortions as well as giving a reasonable increase in field of view. There was moderate vignetting when compared to the native setup, but I would expect that the use of flat fields would alleviate most of this issue.
The Mogg 0.66x reducer has good field flattening abilities and excellent focal reduction with flow on exposure reduction times. It certainly held its own with the WO reducer with only a slight softening of the image and some chromatic aberration. As with the WO there was noticeable vignetting, but again, flat fielding would help to remove this issue. For someone looking for a well priced focal reducer I felt the reducer I had to be a stand out performer when considered next to its more expensive cousins. The WO pipping it in colour correction for chromatic aberration and ease of focus and degree of sharpness achieved. It may be possible to improve the ease of focusing by increasing the length of focusing exposures or using a relatively bright star with a focusing aid, such as a Hartmann mask, in conjunction with computer aided focusing.
The Celestron and Meade reducer’s performances were disappointing in this situation. While the center of the image was comparable to the WO reducer, the edges of field performances were dismal to say the least. This is not to say that these reducers aren’t great reducers, just not in the situation of ED80 refractor/DSLR combination. If all that is needed is the center of a field of view then these reducers would get you out of trouble, but don’t expect to achieve an image that is useful across the full field of view.
NB. It seems to me that these two reducers are very similar in their construction. For an exercise I layered a Meade reduced image over a Celestron one using Photoshop and found that they showed almost identical reduction and distortions. The Celestron image beating the quality of the Meade one possibly because when the images were taken, NGC104 was about 75 deg above the horizon for the Celestron and when NGC104 was only around 30 deg above the horizon for the Meade reducer.
Thanks to Steve Mogg from Mogg Adapters for allowing the use of the Mogg 0.66x reducer for the purposes of this trial.