For the past two weeks I have been going through a load of old images in order to work out the best parameters to use for this season’s imaging. In doing so I realised that it must be odd, to an outsider, why someone should be running a couple of observatories with such disparate kit. In fact, it looks plain stupid. So that got me to thinking how all this lot came together, and although a little less stupid, it is still a bit odd. So this is how it all came about.
The three distinct imaging systems are:
- A Hyperstar 4 on a Celestron Nexstar GPS C11 with UV/IR cut filter and 2600MC Pro OSC CMOS camera.
- Three Sky90s with f#4.5 reducer-correctors, UV/IR cut filters and 2600MC Pro OSC CMOS cameras on a Paramount ME.
- Three Canon EF 200mm prime lenses with UV/IR cut filters and 2600MC Pro OSC CMOS cameras on the same Paramount as 2 above.
The very first imaging system at The New Forest Observatory was the Celestron Nexstar GPS C11 with an original Hyperstar (no rotation or collimation adjusters) and a tiny little H9C OSC CCD. The tiny CCD was 1392 x 1040 pixels (approx. 1.44 megapixels) in a 9 x 6.7 mm active area, and the FOV was tiny. In fact the FOV was so tiny it took a 4-frame mosaic to get the whole of the Pelican nebula. However, when you are taking 100 subs of 60-90 seconds in an imaging session you end up with some pretty good looking very low-noise images – especially when you are lucky enough to have Noel Carboni – possibly the World’s best astro-image processor doing all the image processing for you. If you’re interested in how that came about, you can read it in our book “Star Vistas” published by Springer.
After a few years it became pretty frustrating working with such a small field-of-view, so when Starlight Xpress came out with the M25C an OSC CCD with an APS C size chip, I just had to have one. But the CCD wouldn’t fit on the Hyperstar – what to do? This was the first sign of madness I guess. I bought the utterly superb Sky90 refractor (recommended by Steve Cannistra no less) together with its f#4.5 reducer-corrector. This turned out to be a perfect match for the M25C camera, and I piggy-backed the Sky90 on the C11 in place of the small guide scope that used to sit there. So what do I do for a guide scope? Yes, I’m sorry, I do apologise, but my beautiful C11 was relegated to just being a guide scope. Never mind, I knew this wasn’t going to be forever, but it was for about a year. During this year I was on a Sabbatical from the University which was just as well because I hadn’t realised quite how much my time was going to be clobbered going from an ultra-fast f#1.8 system (the Hyperstar) to an incredibly slow (for me) f#4.5 Sky90. Sub-exposure times went from 60-90 seconds with the Hyperstar to anything up to 20-minutes and not much below 5-minutes with the Sky90. This meant that instead of getting a good-quality image in one night, it now took several clear Moonless nights to get a decent image, and that’s a huge time penalty to pay. So were there any real bonuses that came with the Sky90 beyond a much bigger FOV? Yes there were several, some of them biggies. One biggy was once you had got good collimation on the Sky90, that was it, done and dusted, it just stayed put. The Hyperstar on the C11 on the other hand needed constant attention in that department. Anything else? The Sky90 showed no lens flare, even from the brightest star in the sky – Sirius – and that was truly impressive. Lens flare from bright stars was a pain with the original Hyperstar. There was also an undefinable “something” about the Sky90 star field images that put them above the Hyperstar and I would guess that was better contrast as there was no secondary mirror obstruction in the light-path, but I’m not sure. Anyway, my Sabbatical was about to come to an end and I wouldn’t have anything like the time I had to pursue Sky90 imaging any more – so I was left wondering what to do.
But not for long fortunately. On returning to the University I was sent out to Arizona on business, and you know what Arizona means – it means Starizona and the land of the Hyperstar. I let the Starizona guys know I was going to be over, and where I was staying, in the hope that we might meet up. We didn’t manage to meet, but they did send over to my hotel their new Hyperstar 3 and their microtouch focuser unit for the C11 – thank you guys! Getting back to the NFO, still jet-lagged, I tried to put the new Hyperstar 3 on the C11 and couldn’t do it, the fittings were completely wrong. A quick email to Starizona sorted it. You don’t leave the secondary mirror holder on the end of the Hyperstar when you want to fit it to the C11. Jet-lag can hit you in the most unexpected of ways. First light with the Hyperstar 3 is something I’ll never forget. I was imaging the whole of the North America nebula with the M25C and when Noel had processed the image he was amazed by the “diamond dust” stars – absolutely perfect pinpoints of light – amazing. AND the Hyperstar 3 came along with rotation AND collimation adjusters, so for the first time, with the aid of CCDInspector I could get bang-on collimation with a Hyperstar – another first. Surely this is the ultimate imaging system and I don’t need to move on from here? Yes and no. There was still lens flare from really bright stars, but I managed to sort that by painting the edge of the large glass elements in the HS matt black. It worked! I let Starizona know and I think they probably implemented it in their HS 4 as I see no lens flare from that one – but I’m getting ahead of myself. I did quite a bit of HS3 imaging but for some reason (probably not logical) I still hankered after the Sky90 and its very nice star fields.
So that led onto the next major development at the NFO. The HS3 is f#2 and the Sky90 is f#4.5, so if I am going to get the speed of a HS using Sky90s, then I will need an array of 5 Sky90s working in parallel. I was familiar with the WASP array, so it didn’t seem too stupid to build a MiniWASP array using Sky90s. However, my wallet could not stretch to using 5 Sky90s, especially as the array would be sitting on top of a very expensive Paramount ME (to take the weight). So I downgraded the idea to just 3 Sky90s (and a guide scope) to sit on top of the Paramount. And all this lot of course had to go into a second observatory which the missus was (of course) quite happy about – nuff said. Anyway, over the course of about a year I managed to get all Sky90s imaging in parallel with the SX M26C OSC CCD on the back of each scope – and – after a LOAD of images were taken – I was not very impressed. Oh dear. You see there is a major flaw in the argument that you get an effective f# equal to the f# of one scope divided by the square root of the number of scopes, it’s just that you’re getting down 3x the amount of data that you would with one scope that makes the array “seem” quicker. So what’s the problem with that you say. The problem is, the Hyperstar 3 on the C11 has a focal length of 559mm and the focal length of the Sky90s is 405mm. Again so what? Well, as I found out the hard way, by studying the images over months, there’s more to it than just the speed. The RESOLUTION of the Hyperstar 3 is way better than the resolution of the Sky90s (it doesn’t matter how many you have in parallel of course, it is the resolution of just ONE Sky90 that matters) and that’s what I hadn’t considered and that’s what made the Hyperstar images STILL look superior to the images from the array. Oh dear, what a cock up. And what an expensive cock up. But I still enjoyed getting 3 scopes all tuned in and up and running so I ended up doing a lot more work on the Sky90 array than on the Hyperstar anyway.
Unbelievably I then hankered after even more field of view than I was getting from the Sky90s, so how do I go about that? Well clearly I need a shorter focal length, and at that time I didn’t have the option of telescopes at 200mm but only DSLR lenses. That’s OK I’ll try one of those then. So I bought a Canon EF 200mm prime lens, popped a Canon 5D MkII on the back, and then piggy-backed that on – yes you’ve guessed it – the good old C11! First images down were pretty good and I even made a mosaic of the whole of Cassiopeia using that rig. But I saw the greater potential of a CCD on the back. So I put the old M25C on the back and being well impressed with the results I tranferred to rig over to the top plate of the MiniWASP array in the other observatory. Then, thinking about it for more than 2-seconds (always a bad thing for me to do) I realised that a single 200mm lens at f#3.85 was going to be slower than 3 x Sky90s at f#4.5, and so not to slug the speed of the overall system I would need 2 x 200mm lenses which is what I started off with and have used for a few years now.
So where are we today with all this? Starizona have since provided me with the Hyperstar 4 and the totally amazing 2600MC Pro CMOS camera. The camera is a beaut and on the Hyperstar it is a killer combo. Starizona also sent me 2 x Optolong filters to try out on the Hyperstar. I am sorry to report that the high speed (cone angle) of the Hyperstar means thse filters really don’t work – but who cares? The UV/IR cut filter with the Hyperstar 4 on the C11 and 2600MC Pro OSC CMOS camera STILL remains the unbeaten killer combo at the NFO and I shall be making MUCH MORE use of this rig this season than I have done before.
In the array dome I added a THIRD 200mm lens to the plate of the array (just because I could) and I have only just got this rig all aligned and collimated ready for this season. All 3 Sky90s have now had their M26C OSC CCDs swapped out for the 2600MC Pro OSC CMOS cameras and they are awaiting first light.
So do I really need to be running these 3 rigs and wouldn’t one suffice, say the Hyperstar rig? And the answer is, it really depends on what you’re imaging. I’ll take just one example to make things clear. Let’s say I want the definitive M31 image I can get from this location. What rig do I use? If I use the Hyperstar 4, then this needs a 2-frame (vertical) mosaic and even then it’s a little too tight (side-to-side) to guarantee that I’d get the whole of M31 including the very faint outer regions, in the FOV. The Hyperstar is of course preferable as it has the highest resolution. Well I could go to a 4-frame mosaic, and I might even do that, but 100 3-minute subs which goes up to 4-minutes with dither, for each frame, is committing a LOT of imaging time to just one object. Moving onto the Sky90s, then M31 is a comfortable 2-frame mosaic but the subs go up to 9-minutes, plus 1-minute dither, and I’ll be needing 100 of those for each frame. With 3 imagers this means 5.6 hours per frame, which is probably 2 nights of imaging per frame or 4 nights in total – which is reasonable. Let’s go back to the Hyperstar 4 for a moment. That is only 400-minutes per frame with (a total of) 4-minute subs, which is 6.7 hours per frame – and that’s what you’re stuck with, even for an extremely fast system like the Hyperstar – if you only have one of them! So a 4-frame M31 mosaic on the HS4 means 26.7 hours of total imaging time. I think I’ll take the drop in resolution and use the Sky90s for this one. But hang on – what about the 200mm lenses? M31 fits EASILY in a single frame on that, so isn’t that the answer? Never that easy is it? At 3.88 arcseconds per pixel the resolution is totally unimpressive for a definitive M31 from this site, so for the moment, the Sky90s it is.
So I hope that gives you an overall picture of why I did what I did, and the last example at least also shows you how and why each imaging rig has its own unique place at the New Forest Observatory. What can I possibly say other than “Clear Skies”.
