Ikonnikov

No worries! I've had the secondary dew up a couple of times after several hours when it's been particularly damp and now use a large dew strap placed around the outside of the tube just above the primary (as recommended by folks on Cloudy Nights) on a low heat setting which works well without noticeably affecting the image quality. This isn't an option for the truss tube version if you're interested in that but you can get a dedicated secondary heater which fits under the mirror with the power cables going across one of the spider veins (to prevent them from creating unwanted diffraction effects).

Once you've done the fine tuning and are happy with it, it should hold collimation pretty well for months/years; I dismantle my setup after every imaging session and I've only had to redo it once after accidentally bashing the end of the scope on a door bringing it in the house.

I have the old Altair Astro solid tube version of this scope and like it a lot, although it took quite a bit of work to get it up to scratch. I strongly recommend getting a version that has the primary mirror cell decoupled from the focuser (I think this is the case for the newer truss tube versions) otherwise collimation is very tough. Once you get the collimation close with tools (e.g. a cheshire) you can fine tune it with your camera attached by inspecting star shapes (see the DSI method on the web) to get it pretty accurate. Then it seems to hold pretty well. In answer to your questions, I think APS-C is generally accepted as the largest chip size that you can use with the GSO-type RC10 scopes (using mirrors only) without encountering significant field curvature and degeneration of star shapes. I find using my KAF16200 CCD (APS-H 6 micron pixels) that stars are noticeably distorted in the corners without a corrector, but using a 2" RC flattener from Teleskop Service, (rated up to 44mm image circle) the whole field is pretty well corrected. I image at bin 1x1 or 2x2 depending on seeing although 1x1 is a bit on the over-sampled side 95% of the time. The basic GSO 0.75x reducer has no field flattening capability so it reduces the usable image circle to down to 15mm or so; reducer/flatteners that work well on RCs (e.g. the TS 0.8x RC reducer which I am tempted to try) are unfortunately a lot more expensive but supposedly give well corrected 44mm or bigger fields. My main complaint with the scope is that I don't get many nights where conditions are good enough to make the most of its resolution. This is a recent image with the RC10 plus TS 2" RC flattener using 2x2 subs (~1.2" per pixel) and very rough and ready processing (without preserving stars properly).

If the reference scope is much larger in aperture than the test scope wouldn't this help to reduce the apparent error (assuming the lager scope was reasonably well collimated)? I seem to remember people on CN using big dobs for this purpose.

Yes, its a reducer/flattener; as far as I know all Takahashi reducers are. You can see the optical diagrams here (under documentation): https://www.telescopes-et-accessoires.fr/fc-76dcu-ota-tube-seul-au-coulant-3175-pare-buee-non-retractable-c2x30343677

I also had a go at a drizzle experiment recently, and came to a similar conclusion: Capturing subs on my E130D with a ASI290MM (giving 1.39 arcsec/pix image scale bin 1x1) with medium SGPro/PHD2 dither between each sub, using PI for (pre)processing. Capture 1: 64x subs bin 1x1 15sec exposure, registered with auto settings and integrated with default linear clipping rejection settings. Average FWHM about 2.8 arcsec by PI FWHM eccentricity script (Moffat 2.5), similar values by dynamic PSF of a selection of stars. Capture 2: (back to back with capture 1, no obvious change in sky conditions) 80x subs bin 2x2 5 sec exposure registered with bicubic spline setting (to avoid horrible dark ringing resulting from auto/lanczos interpolation) and create drizzle files, integrated with linear fit clipping rejection or additionally drizzle integrated with scale set to 2. In short the drizzled 2x2 data never really got close to the original 1x1 image in terms of star FWHM . I kept reducing the dropshrink setting which did lower the FWHM slightly but 0.5 was the lowest drop shrink value I could use without major artifacts (caused by lack of coverage) and this still gave stars with ~30% higher FWHMs than the integration of 1x1 subs. Also the FWHM in the drizzled image (drop shrink 0.5) is barely lower than that of the undrizzled 2x2 integration and using the default drop shrink of 0.9 it was actually higher! I appreciate that at 2x2 binning here the FWHM measurement accuracy might be compromised due to undersampling but I can't say my results fill me with confidence as to drizzle being worth the bother either! Maybe a different selection of pixel interpolation in image registration might help? Paul Integrated masters attached drizzle_integration_BicubicSpline_2x2_DropShrink_0p5.fit integration_bicubic_spline_2x2.fit integration1x1.fit

These images have stars with low snr (especially Ha one) and a lot of hot pixels which are being picked up as stars by the PixInsight script and very significantly lowering the reported FWHM. Manually measuring each of the stars with dynamic PSF gives an average of about 5pixels (2.6arcsec) FWHM for Ha and 3.2 pixels (3.3 arcsec) for OIII, so I'm in agreement that by most measures you're considerably oversampling here. From my understanding, optimal sampling rate for DSO imaging (i.e. to record maximum detail without oversampling) is a continued source of debate on this and other astro forums; Vlaiv has made a detailed argument for a relatively low rate of about 1.6x smaller than the real-world FWHM of your system whereas folks on Cloudy Nights generally seem to go in the other direction with rates 2-3.5 x smaller than seeing with their own technical rationalisation (e.g. see https://www.cloudynights.com/topic/650493-understanding-criteria-for-what-is-proper-sampling-of-imaging-system/?p=9150558 ). Personally I guess I'm of the opinion that for long FL DSO imaging it's better to slightly oversample than to undersample (which can easily result from increasing binning level) as you're never going to recover lost resolution fully from the undersampled subs (even with drizzle which also is a controversial topic here!) but you can (eventually) get the snr up from oversampled subs by taking more of them. Paul

What you don't show/mention is the quality of the in-focus image; if that's good after adjusting with the star test then why worry about the Cheshire? Personally I would use mechanical alignment as a starting point and always fine tune collimation with star testing when possible; with an RC scope you can follow the procedure below for adjusting your primary and secondary with your imaging camera attached to get very accurate collimation. Since it also relies on images of (slightly) de-focused stars you don't need amazing seeing to get a good result. https://www.deepskyinstruments.com/truerc/docs/DSI_Collimation_Procedure_Ver_1.0.pdf Paul

Great first light with your new camera, agree that the OIII filter is very likely the cause of the halos. I wonder if you might be able to get even better signal to noise using longer sub exposures since the KAF16200 sensor has quite high read-noise; certainly I find that 30 mins per sub helps with the Moravian G3 camera and 3nm filters even on an Epsilon at F3.3 (although I've not gone beyond this so far for reasons of pixel rejection, imaging time wasted by spoiled subs etc). Paul

Using various online calculators (e.g. the Wilmslow Astro website) the airy disk size for green light (510nm) at 80mm aperture comes out at 3.2 arcsec and for red (650nm) as 4.1 arcsec i.e much larger than typical guiding error and larger than typical uk seeing. This being the case then wavelength would have more of an effect on the final image fwhm between oiii and Ha filters than you suggest above. I still typically see smaller star sizes with shorter wavelength filters when imaging with a mirrors only 250mm RC scope (where the airy disc size is in the order of 1 arcsecond for green and 1.3 for red) suggesting some effect of wavelength on fwhm even at larger apertures. Paul

As you mention above, shorter wavelengths will intrinsically give higher resolution images (with a smaller airy disk and therefore star size) than longer wavelengths. The theoretical airy disk size for Ha emission wavelength is 31% larger than for for OIII primary emission wavelength so could this not account for the differences in star sizes? I appreciate that bad seeing affects longer wavelengths less but i don't know at what point this would outweigh the intrinsically larger airy disk size. From experience I've consistently seen smaller star fwhm measurements for OIII over Ha with reflectors and refractors. Paul

Hi Richard, this article might be worth looking at as it highlights some of the potential problems with star shapes that can arise in the esprit scopes: http://interferometrie.blogspot.com/2014/08/esprit-tuning-how-we-finetune-esprit80.html I was going to hold off on my own Esprit experiences for now but since it seems relevant to this thread I'll continue... TLDR is, (especially if imaging in colder weather) it can be a delicate balance with these scopes between preserving element centering/collimation and preventing some pinching of the optics. I bought a new Esprit 120ED last autumn and although perfectly collimated I noticed some funny star shapes & (asymmetric) diffraction spikes in images appearing only when the weather is colder (below about 4 or 5 C) suggesting mild pinching by one or more of the element centering screws (e.g. below). p Seeking perfection (foolishly perhaps) I decided to very slightly loosen the centering screws for the front element (WARNING this is a high risk procedure!!!) The good news is that I managed to isolate the screw causing the problem and stop the uneven spikes regardless of temperature. The bad news is that I managed to very slightly de-centre the element which has made the star halo slightly asymmetric and resulted in a slight lateral colour shift. I can get good star shapes now if I combine subs from each side of the meridian flip as the distortions essentially cancel each other out (see below) but this is hardly ideal. Also aligning RGB channels is problematic. I figured I had two options from here, admit defeat and fork out to have the scope sent back to Es Reid or attempt more advanced adjustment techniques. Since there was no guarantee that pinching wouldn't result again in cold weather after the scope was recollimated I figured it would be better (and more fun!) if I could attempt this myself at home as required. At a not inconsiderable expense I've managed to put together a double pass autocollimation setup with a beam-splitter, focuser-insertable artificial star and optical flat as described in the Teleskop Austria article above and the test setup seems to check out on my (well collimated non adjustable) Tak FC76. Next step is to try it out on the Esprit... Paul

For fine tuning collimation under the stars with imaging camera attached the procedure below works beautifully. You need to start with the collimation reasonably accurate though, e.g. set using a cheshire, collimating telescope etc. https://www.google.co.uk/url?sa=t&source=web&rct=j&url=http://www.deepskyinstruments.com/truerc/docs/DSI_Collimation_Procedure_Ver_1.0.pdf&ved=2ahUKEwjou_G9h43eAhXlJcAKHdUvDtoQFjAAegQIAhAB&usg=AOvVaw1TQ8YAhohC9_gr_IRXBo3x

For my 10 inch rc I use a takahashi collimating scope then fine tune using the DSI method (google DSI collimation). In essence this uses your camera to monitor star shapes on and off axis to guide adjustments of primary and secondary mirror collimation. It works very well. You could probably just use the Dsi method from your current collimation state without needing the Tak scope. Regarding your focuser i suspect that the large bolt is a pull screw and the two smaller ones either side are push screws. In which case before dsi collimation procedure I would reset the focuser tilt by fully tightening the three pull screws making sure you've sufficiently loosened the push screws to allow this. You may still have some tilt in the system but at F8 this is unlikely to be a major contributor to the misshapen stars. Once the collimation is as good as you can get it by mirror adjustments, if there is still a focus gradient you can then try to alter focuser tilt and monitor its effect with a programme like pixinsight or ccd inspector. The circular defects you describe are residual dust bunnies not fully removed by your flats. This could be due to a differing focus position between flats and lights or by a difference in calibration between the two e.g. no dark/bias subtraction from on or the other or different readout speed etc. Hope this helps. Paul

Had a look at this a while back (before I sold my Canon EF200L) and Moravian and QSI definitely do dedicated adapters for canon EOS lenses that fit their cameras with integrated filter wheels. I think the only thing was you couldn't use it on the versions with integrated guide ports as well since they added too much length to the optical path. For shorter back focus CCDs without integrated filterwheels TS do a 15mm thickness filter drawer which is the thinnest I know of. Paul