The Starry Night, 172

:: home ::


              <<  172  173  174   175  176  177  178  179  180  181  182  183  184  185  186  187  188  189  190  191  192  193  SRCH

Thomas M. Back makes a Telescope.

8/27/2016. Right out of the box, the TMB92SS
threatened to make me an observer again. But that's not what I bought it for, and before I can use it for widefield photography, there's the matter of field curvature.

Where it is good, it is very, very good:



Click to see this at 100% scale


That's under 4 minutes of total exposure. Both a 12th and a 15th magnitude galaxy come up in there. The light-grasp is all I could want in such a compact telescope, but the field shown is only 1/12 of the whole.


full field


You don't buy a dozen eggs to cook with just one.

Where it is bad, it is horrible. Here's the center plus all four corners of the same 24x36mm frame compared to one another at the same scale.


four corners TMB

Click to see that at 100% scale.


I've seen enough nicely-corrected photographs made through TMB92's that I didn't panic, but that's the only reason.

OH Fergodssake, look at all these words!

Keep reading if you want or just CLICK HERE FOR THE SUCCESS STORY

While researching options and waiting for parts, it remains the sweetest visual telescope I have ever used. Deep-sky views through Sheldon Cohen's 6-inch F8 A-P (the 'scope that introduced me to modern refractors), and seeing Antares A & B sundered as plainly as Albireo with Ernst Mayer's highly refined 6-inch F8 reflector have left an impression across decades. This little guy has those velvet black skies and those incredibly tiny stars and that SNAP at the point of best focus that makes unsuspecting stargazers into apo-philes. There is beauty beyond 1/4-wave and merely achromatic correction. On an alt-az mount in a farmer's field, away from the lights of a music festival, the little 'scope made believers out of campers by shattering globular clusters and blew fiddlers and singers away with a clear view of Saturn's moons and rings. Every review promises that the TMB92 will deliver that kind of experience with mechanical finesse and no false color at all. The reviews are 100% correct, but if the image surface (it's not a "plane") can't be hammered flat for photography, then what I have here is eleven useless eggs.


8/29/2016. Don't tell anyone, but I just used several feet of black tape, a few protective strips of paper, and a hose clamp to concatenate a 2.7-inch A-P focal reducer / field flattener with the TMB's Feathertouch 3025 focuser. Mine was a sorry MacGyver for such a refined instrument. There are no photos of this kludge, I made damn sure of that, and I removed all evidence of how I did the deed as soon as I had half a dozen test frames. Nothing was aligned, nothing was centered, and the camera drooped visibly. But for all that half-assery, the experiment showed that field curvature was vastly reduced, that the final EFL will be close to 400mm, and that there are more aberrations in a misaligned eff four point something apo than weeds in an unkept garden, and many stars resembled them.

I tried several spacings, none set or recorded with more precision than the messy attachment scheme deserved. Somewhere within the range of easily obtained spacings is almost surely a flat (or flat enough) focal surface. The 6D will come to focus whereever it may be. Those are two very reassuring items.

I centered Arcturus and tested with the camera at the point where severe vertical astigmatism became severe horizontal astigmatism. Was that the very best preparation for measuring field curvature? I doubt it. All frames were exposed for 20 seconds at ISO 800. The first frame was made through the A-P prime focus adapter and its built-in extension. The rest were shot through a wide mouth adapter: first, close coupled, then extended about 5, 10, 15, and 20mm. I am not going to provide before and after images or any CCD Inspector plots of these results. The spacings and the so-called alignments are unreproducible anyway. The best image was the one close-coupled with the wide adapter. Don't make too much of that. Vignetting was not a serious issue in any of them; it will calibrate right out.

I was impressed enough to order from Starlight Instruments an A-P 2.7-inch end-cap for the TMB's 3-inch Feathertouch focuser. The next day, I emailed the factory. I promised not to be any more impatient than absolutely necessary, but I asked if the adapter I had ordered was a stock item or something that needed to be scheduled and machined. "Packed up and ready to go," Wayne replied. When it arrives and I can put the 'scope back on the Mach1 with a camera behind it, I'll get serious about establishing the proper spacing. And then we'll do some show and tell.

Yesterday, I also asked APM in Germany to quote an end-to-end chain of adapters and glass to use their 2.5-inch flattener. They offer to do that in their catalog pages, and I took them up on it. That's for insurance and so I'll know what the cost is for that known solution. Basically: about 85 Euros to connect the flattener to the Feathertouch draw tube, 290 for the flattener itself, plus extension rings, plus whatever simple widget is needed to mate the corrector to an EOS bayonet, minus European-only V.A.T., plus shipping from Germany, plus maybe customs fees. Call it e400 or a little more. Astophotographer Terry Hancock does not seem entirely happy with this train on his TMB92, but I would be entirely thrilled with the results he's posted using it. My mileage obviously varies.

Another alternative would be a TeleVue TRF-2008, made "back in the day" for 35mm cameras and well-reviewed in the July 1999 S&T. Seventeen years later, it's still made, would be a snap to place at the correct distance from the sensor, and can be delivered for somewhere around half the price of the APM solution and much reduced adapter angst. Examples of images made behind TMB92's using that adapter are also beautiful, but the cropping of most leaves room for reservations about the outermost parts of the field.

The good news is that today's experiments surely put such days of reckoning off a little while. I entertain real hope that the A-P focal reducer and flattener (which I already own) mated to the TMB92 using the Starlight Instruments adapter (way cheaper than the APM chain or the 2-inch TeleVue corrector) will put those days off a good long while.


9/1/2016. The adapter fits like it was made for the Feathertouch. Which it was. It came with clouds from tropical storm Hermine. Between this and that, the RV, Happy Valley, book proofs, god alone knows what all, I had time for five quick exposures of the noonday Sun (with the Sun centered and in each corner). Close-coupled, the reduction is only 0.83x which suggests we are well away from the reducer's intended spacing. And now it's raining. With the A-P prime focus adapter (with its built-in extension) the reduction factor is approximately the nominal 0.75x and there is just enough in-travel to reach focus with the TMB92 in its visual configuration. It will probably be Sunday or Monday before I can do star trials.


9/5/2016. Short-exposure, high-ISO tests at Happy Valley were encouraging but ultimately misleading. The reducer had some but nowhere near enough flattening effect. To get the best results, the A-P prime focus adapter with the built-in extension is the way to go. The corrected field approaches APS-C acreage. In-travel (with the TMB OTA extension in place) is just adequate. A 30-second frame of the eastern Veil was instuctive: encouraging light grasp, discouraging correction. It had "high residuals." Visually, the compressor interferes with the visual experience. Stars are not pinpoints wall to wall as they are with the TMB-92 alone. Hold this reducer in reserve for possible use in combination with a non-reducing flattener such as the APM and Orion products.

I remembered seeing an Orion flattener on AstroMart some days ago. When we returned from Happy Valley, I asked whether it was still available. Was. Even new, the Orion flattener is a fraction of the cost of others, and this one was substantially discounted. The bargain solution has enjoyed some good reviews compared to other small, universal flatteners. It seems worth a shot even if it just puts off having to pick and pay for a more robust solution. And yet, hope springs eternal: maybe it will prove better than expected, the corrected field larger and illumination tolerable. Experiments with spacing may be in order (some reviews suggest the Orion needs 10mm of extension beyond its intended 55mm to behave at its best behind short refractors).


9/9/2016. Let us not get too excited too soon, but first looks through the Orion field flattener (at a pine cone atop a tree and at the afternoon Sun) look pretty good. Vignetting is not an obvious issue by day. I'm keeping my enthusiasm in check until the stars come out.

Very nice! Perfect? No. But very nice even on the full 6D frame. The corners are a bit dark, but nothing calibration won't fix. If I have to crop to 90% for vignetting, mirror shadow, or corner correction, that will be OK. If I don't, better. I tried a 21mm extension (because that's the only size I have) and the results were interestingly different. CCD Inspector says the result with the extension is better, but there is room for doubt. Actually, from visual inspection of the image, there is room to think the analytic result completely misses the mark. However, the flattener by itself mounted directly to the T-ring works well enough that there is a chance that using some short extension rings to tweak the spacing will make things DNP (Damn Near Perfect). The shape of the curvature plot suggests that I need to settle on a focusing strategy: focus at the center or 2/3rds out or something and stick with it. Try a Bahtinov mask to make focus consistent.

Oh, and after you're finished with this adjustment, put the A-P focal reducer back in the chain to see if that can be worked out, too. God, I'm a slavedriver. [I got some in-focus test frames which are abyssmal. It's another project entirely if it's a worthwhile project at all. It's probably not. Certainly not a priority.]


9/12/2016. T2-extension tubes (42mm x 0.75) with nominal lengths of 5, 10, 20, and 30mm arrived. Because of the thread depths and lengths, the actual extensions achieved differ a little from their descriptions. I got a micrometer out:

5mm Ring :: 6.5mm actual increase in separation
10mm Ring :: 10mm
5+10mm Rings :: 16.0mm

The Baader Protective T-ring measures 12mm thick, and the Canon EOS flange to sensor distance is 44mm. So when using the Baader ring by itself the flattener sits 56mm (rather than the expected 55mm) from the Canon's sensor. If the separation needs to be smaller rather than larger, then Orion offers a T-adapter (Item 05224) with a thickness of 9.8mm. There may be thinner options out there, and there's always the lathe to take a mm or two off the face.

Flange  +  T-ring + Nominal Extension :: Actual Total Separation

44mm  +  12mm + no rings         :: 56mm
44mm  +  12mm + 5mm              :: 62.5mm
44mm  +  12mm + 10mm            :: 66mm
44mm  +  12mm + 5mm + 10mm :: 72mm

There were too many clouds to try anything systematic immediately. Maybe tomorrow.


9/13/2016. Here's the key. Add the 10mm extension, use the Bahtinov mask I stash with the AT65EDQ, and focus about 3/4ths the way to the edge of the field. Results are not bad at all. I'm getting some tilt, and since the 2.7-inch accessories are not going to be needed anytime soon, a return to the stock Feathertouch 2-inch adapter and its 3-thumbscrew brass compression support addresses some of that. With that, we are in business! Let's shoot some stuff.

ISO 1600 under hazy moonlit skies gives me Mag 13.8 in 20 seconds. That has implications for both deepsky and eclipse photography:

1/20 sec :: 7.3 magnitude
1/8 :: 8.3
0.4 :: 9.3
1.0 :: 10.3
3.0 :: 11.8
8.0 :: 12.8
20 :: 13.8

50 :: 14.8
125 :: 15.8
320 :: 16.8
800 :: 17.8
10x180 :: 18.8
25x180 :: 19.8
64x180 :: 20.8

Want to go deeper than that... use a bigger 'scope. As for the eclipse, how about a new page (or several) for eclipse related maundering?


9/24/2016. Just when you think the stars are sharp enough, somebody offers a said-to-be better flattener at an irresistable price. The Orion does the job. But I think the TeleVue TRF-2008 will do it better. In any event, the TeleVue item is also a reducer. So I decided to audition it. Both correctors may find places in the kit.

Initial tests show that the TRF-2008 is really touchy. The thickish Baader protective T-ring with no extension isn't quite right. But adding 5mm makes things worse (and to confirm that, adding 10mm makes things worse still).

At first glance, that makes it tempting to think a Baader Ultra-Short adapter might be callled for. It would eliminate the space imposed by the T-ring and allow testing in a whole new, shorter regime. A closer look at the initial tests shows something else: with 0mm of exention, star distortions are concentric around a central point, but with the 5mm and longer rings in place, distortions are radial. The ideal spacing would seem to lie somewhere in between no extention and the 5mm extension. Where the star is placed on the focal surface for focusing matters, too. I tried focusing on centered stars and on stars 3/4 out toward the ends of the frame, but I was not sufficiently systematic to make as much of that as I would like. It may be crucial.

Let's do more rigorous testing to decide whether yet another adapter is the answer or if spacing and focusing strategies just need to be refined. Test by unscrewing the T-adapter a mm or so at a time and by focusing in the center and at various points toward the outer field. Somewhere is a sweet spot. Find it! (For encouragement, see images on Astrobin made using this 'scope and this adapter.)

Late last night, I used the lathe to turn down an old EOS T-adpater. It will provide slightly smaller spacing in the unlikely event that that seems the way to go. For testing extensions between 0 and ~3mm, use an M48 parfocalizing ring, and keep those Delrin spacers from Agena Astro in mind if some small spacing needs to be standardized.


9/25/2016. This evening I was racing incoming weather to get tests done before clouds shut down the sky for a few nights. I followed a script to organize and test 15 configurations. Exactly one looks good, and it looks very good. Surprise! The best results are from shortening the separation after all.

I tried the Baader T-ring and no spacers again, focusing on Vega when centered, halfway to the end of the frame, and about 80% of the way to edge. Yep: that still sucks. Then I inserted a 3mm-thick parfocalizing ring and tried all three focusing strategies again. That's worse. Then I exchanged the Baader T-ring for the one I cut down on the lathe last night. The test frame made after focusing on a star at the center of the field looks good. Then I added the 5mm extension tube. Terrible. Then the turned-down adapter and the parfocalizing ring. Also pretty awful.

The turned T-ring has a thickness of 0.383-inches (= 9.72mm) or just a whisker thinner than what Orion claims for their EOS t-ring (9.8mm). As long as this one remains solid, I'm fine with it. But if I need a spare, Orion has them. Another option would be Baader's "Ultra-thin" T2 to EOS adapter which has somewhere between 0 and 2mm of optical length, but that would mean adding a 7.5mm ring (which I do not have) and adjusting again using spacers.

I added more Superglue to make sure the inner thread ring doesn't move and then polished the face of the shortened adapter using medium grit sandpaper on a firm surface to insure that the face of the adapter is smooth, flat, and square. I'll take another look for hints that a sub-millimeter adjustment via thin spacers would help, but for now, let's leave it alone.

After a month of hammering, I have these two useful configurations:

  1. EFL=510mm, F5.5. 2.65"/pixel on the 6D. Use the Baader T-adapter and 10mm ring with the Orion flattener. Focus ~70% out to the end of the field.

  2. EFL=402mm, F4.4. 3.36"/pixel on the 6D. Use the shortened T-ring and no spacing with the TeleVue flattener/reducer. Focus at the center of the field.

ccdi plots


9/28/2016. First-light (photographically) for the TMB92. I'm trying full-frame, white-light shots of NGC 7000, the North America Nebula, composed to include Deneb to showcase the very wide field offered by the F4.4 configuration. The very first frames were through a pine tree, and it turns out that I missed the field (I aimed too far west). Let's try again.

I'm doing three minute exposures at iso 1600, guided with PhD using the Meade DSI behind the 50mm guidescope. I'm powering the 6D from the Kendrick battery using a 12v adapter from Orion.

How did I get along for so long without a headlamp? It's great for night-time telescopy and presumably around camp. It's great to have both hands free.

The sky closed down not long after midnight. Haze, fog, dew. I have 93 minutes of useful exposes, including one hour's worth of reasonably fog and haze-free frames. I brought the camera and 'scope inside and left it making darks since the temperatures inside and outside are not very different. Consider this a rundown test for its internal battery. It had 78% capacity when I started shooting 180s exposures. Got 61 before the battery gave up.


ngc 7000

NGC 7000, Deneb in upper right
21x180s (1h 03m)
Canon 6D @ ISO 1600
TMB92SS @ F4.4
TeleVue TRF-2008 Flattener
Click the image for a better view


You take a 3.6-inch apochromatic refractor at F4.4, put it in front of a Canon 6D, aim it into the Milky Way and open the shutter for an hour while guiding the whole shebang to an arc second or better... That captures a lot of stars.

I've cropped the frame by 15-20%. The corners were not "handled" by the afternoon-sky flat. They remained dark. Correction is not perfect into the corners at full-resolution, but perfection really isn't a requirement for any plausible web presentation. Point spread functions in the outer area of the frame, including some shown here, are "complicated." If PSFs can be improved with spacing, and illumination with better calibration, do it, otherwise it's fine. Really. That's one hour on a hazy night in the Carolina foothills. It'll do!

There are plenty more examples of deep-sky photos made using the corrected refractor on the next pages.



My deep-sky photos are made with a variety of sensors and optics. Deepest images usually come from a SBIG ST2000XM CCD behind a 10-inch Astro-Tech Ritchey-Chretien carried on an Astro-Physics Mach1GTO. The CCD is equipped with a CFW-10 loaded with Baader wide- and narrow-band filters. Camera control and guiding are handled by Maxim DL 5.12. A Canon 6D and a modded 50D find themselves mounted behind an Orion 10" F4 Newtonian or a 92mm Thomas Back refractor or a tiny but mighty AT65EDQ refractor, sometimes with Backyard EOS in control and PHD Guide keeping things on target. Really widefield photos are often made using the 6D and various camera lenses and an iOptron Skytracker mount. PixInsight does most of the heavy lifting in post-processing — alignment, stacking, gradient removal, noise-reduction, transfer function modification, color calibration, and deconvolution. Photoshop along with Focus Magic and a handful of other plugins get their licks in, too, especially when polishing for the web.


:: top ::



                   © 2016, David Cortner