Refractor image of M38 and NGC1907

[timh]

Recently - in the very few hours of clear sky available - I have been attempting to analyse and improve the performance of my old William Optics 110 mm tmb F 7.0 apo refractor in combination with a Flat6A III 0.8X reducer flattener. Most of my experience has been with Newtonian reflectors and I am less experienced with refractors and so less able to judge what good or 'expected' performance should look like (e.g. amount of bloating of blue or bright stars -- field flatness etc). I seem to have resolved some apparent issues (tilt that wasn't tilt) just by tightening a few things up.

So I'm posting this up just interested in any feedback as to where this sits in terms of expectations for a fairly short exposure (45 min) image using an apo refractor at F 5.5 and Bortle7. I was thinking of selling it on and perhaps going for a fast short focal length Newt but having worked with it more now I am warming to the refractor and quite like this image and the wider field possibilities it offers relative to my existing Newts.

WO 110 F 7.0 tmb apochromat + FLAT6AIII flattener/ 0.8X reducer (to F 5.6) on a CEM70 Ioptron mount. ZWO ASI294MC camera 133 x 20s at gain 124 with ZWO UV/IR cut. About the topright 1/3 of the overall field - cropped for better composition. Image is at 1.54 arcsec/ pixel - somewhat undersampled for an FWHM of ~ 2.7.

Tim

[timh]

Given that the above was undersampled (1.55 arc sec pixel relative to seeing which must have been around 2.5 arcsec or less) and also dithered (using SharpCaps non guided dither) the data provided an opportunity to compare Debayering with CFA x1 and CFA x2 drizzling.

It is always nice to see that the theory works -- and it did in this case more or less as you might expect. The CFA X 1 drizzled image resolved detail that was blurred out in the debayered image and the x2 drizzle was very slightly better resolved again. For undersampled images debayering adds significant blurring -- but less so for oversampled images. The estimated FWHM of the X2 drizzled linear image was ~ 2.3+/- 0.3 arcsec and the X1 and debayered image ~ 2.65.

Tim

[admin]

Hi Tim,

well, I must admit that I am usually in the other camp - preferring refractors because they are 'low maintenance' compared to reflectors which need collimation However, I currently have an f/4 reflector in the observatory and an f/12 classical cassegrain, so I am also tinkering with telescopes a bit outside my comfort zone.

To be honest, without pixel-peeping the full size image, it's hard to be 100% sure, but nothing horrible seems to be going on in that image. The stars look pretty good all across the frame, with the possible exception of bluer stars having a bit of spread that isn't visible in the orange ones. That's not unusual, even for triplet/APO refractors the blue end of the spectrum is hardest to keep under control.

I would suspect that all other things being equal, you will spend more time capturing images and less time tinkering with the refractor than you would with an equivalent focal length newtonian, however, the newt might be f/4, so maybe that wins.

cheers,

Robin

[turfpit]

Tim

An interesting comparison of the various methods - thanks. I agree with Robin about the low maintenance aspect of refractors.

Dave

[timh]

Thanks Robin,

Can't argue with your comment about refractors. I think that I just like the diffraction spike effect, colour faithfulness and low cost per photon of Newts.

Thanks for your comment on image quality. It was exactly the blooming that I was mainly seeking feedback on. I like the look of the image myself and am getting the feeling that what I have is about normal for an 'apo' -- or at least that I would have to pay an awful lot more to do much better.

The other issue (and the one I seem to have improved by tightening the tube up and the connection to the reducer/flattener) is the alignment/ overall field flatness. I didn't show the entire field because it didn't provide a good composition of the two star clusters (and this thread is 'the gallery') but I did get it analysed by CCD inspector by a chap "Skipper Billy" who offered to do so online on SGL - it looked to be well aligned and reasonably flat (bearing in mind that the 0.8X reducer/ flattener is unmatched and the backfocus distance is probably close but not not optimised).

Interestingly though I think that the Sharpcap 4.0 collimation tool provides equivalent information to CCDI . FYI here are 3 runs of the collimation tool run on 3 separate SC autostacks (each of 30 frames) run at different times on the same evening versus the same star field. Not perfect circles but nearly centred ellipses with really rather small differences in FWHM marking the contours -- so OK ish at least. I think that the general shape and orientation of the ellipse is real because it was quite consistent - and is a tool potentially useful for fixing any slight tilt / backspacing ajustments.

Tim

Just a PS on the blue bloating. I compared the refractor image of M38 with a Newtonian image of the same taken a couple of years ago and there is perhaps some increase in the relative size and definitely an increase in the blueness of the blue halo with the refractor. However I think that I may have been unreasonable in my choice of UV/IR cut filter in choosing one that cuts off at < ~ 400 nm -- I don't think that most apos are corrected to quite so short a wavelength? -- and I note that the Baader UV/IR cuts off at 420 nm and so this may be a more appropriate choice and maybe fix the issue.

[timh]

Thanks Dave --- I was actually quite pleased to see an improvement with drizzling. I have known about in theory for some time but since with Newts I normally oversample haven't really had a chance to use it but it was nice to see how it did indeed work and resolve small stars from the halos of larger ones. I will certainly use CFA x1 drizzle henceforth. The other useful bit of learning for me was around the slight bloating in blue -- which I have seen before and another reason I think for my bias to Newts. It has made me think about my current UV/IR cut filter and realise that while it might be fine for Newts it really probably isn't for anything other than the most exceptionally well corrected apo refractor? I am going to be interested to see what things look like with the refractor using a (FWHM value) UV cut up at 420 nm (Baader)--which still encompasses most of the visible spectrum above violet..

Tim

[admin]

Hi Tim,

I have thought about using the collimation tool for analysis of tilt etc, but never really got around to doing anything with it, especially since it's a fairly niche thing and there is already software out there that does it. One thing I did notice from your images is that because the range of FWHM is rather small, you are not seeing many contours being drawn. Perhaps I should make the gap between the contours get calculated dynamically so there are always at least 5 to 10 visible?

cheers,

Robin

[celkins]

That would help me, too, with my Hyperstar C11 rig…
So, yes please, Robin

[admin]

Hi,

adding more contours in the elliptical/circular case was easy, so I made that change for 4.1 and it will be in the next update. I also tweaks things so that it shows 'Inverted Ellipse' where the pattern is elliptical but with the highest values (rather than the lowest) in the center.

cheers,

Robin

[timh]

In the odd moments of having clear skies this past month I have carried on tinkering with the WO FLT 110 apo refractor/ WO 0.8x Flat III reducer flattener (unmatched) combination.

Having gone down a bit of a rabbit hole with a Tilt adapter - completely unnecessary as it turned out - the two key improvements were.

a) Switching the ZWO UV/IR cut filter for a slightly narrower (400-700nm window) Baader one --> truer colour on bright blue stars
b) Shortening the sensor/ flattener distance to about 54 mm rather than 55 mm --> better star shape at the field edges -now good across the field

Robin's update of the Collimation tool in the beta 4.1 version was useful to monitor improvements and changes as I decreased the backspace distance and played with the tilt plate.

So for example.. A suboptimal configuration with a maladjusted tilt plate in place and at 55 mm BS ...

Compared with a relatively flat and well centred field in the final set up with no tilt plate and at at 54 mm BS.

The final picture is an image of M101. Just 19x66s frames at F5.6 under poor skies before it clouded over. But I thought not bad considering -- and most pleasingly it is a full 4/3 inch frame (ASI 294 MC) image without any cropping that shows relatively little star distortion in the corners.