0.90
#...
·
·
1
like
|
|---|
|
Still no great options. My combo of RASA 8 and the ASI2600s are a tilt nightmare. ASTAP said what I had was "Extreme" tilt! Oh great. But I eventually figured out that with my Starizona Filter Holder, I had left off the 5mm extension that needed to be on there. I added it, and it helped a lot. I went from "Extreme" on ASTAP down to only "Severe". Sigh. I would love to start working with the octopi. I feel like the price is what is actually "Extreme". Plus, it only works for one camera/scope combo. If I put it on my MC camera, I am still out of luck with my MM. One octopi is overpriced, 2 is ridiculous. So, I try to not look at my corner stars, they're terrible. If I have something only in the center of the frame, I can crop significantly. But this is a poor solution. Someone needs to come up with a more practical tool. |
11.02
#...
·
·
1
like
|
|---|
| The RASA 8 has very little backfocus space available, which makes getting a tilt corrector on it very challenging. I faced a similar challenge getting the Gerd Neumann CTU on the Epsilon given it's 56.2mm spacing from the corrector. The CTU took up too much, so the Octopi was the only feasible option. |
20.43
#...
·
·
1
like
|
|---|
Paul Wilson: Paul, Have you tried a printed tilt shim? It won't give you the precision of octopi but it takes up very little backspacing and you might get lucky in that it improves your frame at least. https://agenaastro.com/astrodymium-rotating-tilt-adjusters-for-select-threads.html |
4.21
#...
·
·
1
like
|
|---|
|
I'm finally getting around to optimizing my tilt, so this article is perfect timing. One complication in my imaging train is I have a Pegasus Falcon rotator. My imaging train consists of a QHY268m, the QHY filter wheel, the rotator, a precise parts spacer, and the Astro-Physics field flattener or focal reducer. When I do a HocusFocus run at say 0 degrees, it indicates the sensor tilt is oriented left to right. If I tell the rotator to rotate the camera 90 degrees, HocusFocus reports the tilt runs top to bottom. If I'm understanding this correctly, this indicates the source of tilt is between the rotator and the scope. If the tilt was due to something like the sensor not being square in the camera, I'd expect the tilt to stay the same regardless of the orientation of the camera. Does this seem like a reasonable interpretation? I'm hesitant ordering something like a CTU (which would require a new precise parts adapter) until I know where to place it. I'm planning on testing this by placing a shim between the precise parts adapter and the rotator based on the hocus focus results. I was going to try placing the end of a feeler gauge strip between the adapter and rotator at the position that needs to move the sensor out. Assuming this helps, I can then try different thickness feeler gauge strips to find the correct amount. Does this sound like a way to start? |
|
0.90
#...
·
|
|---|
Paul Wilson: *** No I haven't tried that. I don't know exactly how I would determine what thickness of shim I would need or where it would need to be placed. And of course I don't own a 3D printer. *** |
|
20.43
#...
·
|
|---|
Paul Wilson: They are purchasable. In the set it looks like there are three sizes. The real limitation for shim thickness will probably be the thread pitch on the adapters you slip them between. M54x0.75 will allow for less shimming than m68x1 for example. |
|
20.43
#...
·
|
|---|
Alex Ranous: I agree that it sounds like you have tilt between the rotator and OTA. Definitely complicates things. In this case you would want to put a tilt device between the rotator and ota. Do you have the spacing allowable to put something in there? Like the Gerd CTU? Hopefully you dont also have tilt between the sensor and the rotator, or at least in a meaningful way. |
22.44
#...
·
·
3
likes
|
|---|
|
Chris, Thanks for such a fantastic report! VERY well done. I've been working on my GTX130 as well and so far, I've only done a first order tilt correction just by centering the FWHM pattern using the FWHMEccentricity script in PI. The bane of my system (and I suspect many others) is the dovetail mounting system. Many dovetails are not properly designed (or implemented) and can introduce significant tilt into the components. In my case, the FLI filter wheel uses only two set screws at 120 degrees to attach the camera. That can work--but ONLY if the interface is properly designed and it wasn't. I could look along the side and see the 0.007" gap being created on one side when the screws were snugged. Tin foil isn't a bad shim but there are better choices. Years ago I ordered this assortment of plastic shim stock from Amazon: https://www.amazon.com/Plastic-Assortment-0-0005-0-030-Thickness-Length/dp/B00065UYKU/ref=sr_1_11?crid=QGW64A6GTUP1&keywords=shim+stock&qid=1659212045&sprefix=shim+sto%2Caps%2C190&sr=8-11 It is basically a lifetime supply and since it's plastic it can be easily cut by scissors without creating a raised edge, which can happen with metal stock. Highly recommend. I shimmed the FW interface and checked it with a feeler gauge and went from the distribution on the left to the one on the right. (These plots are for different colors but I just happen to have them up on my screen so here they are.)  I've still got some work to do to get the system dialed in better so I'll have to try your method to check it more precisely. I also want to check for spacing errors as well. We had a night of truly spectacular seeing last night in Bend (~0.6") and my Ha data looked pinpoint corner to corner. John |
|
11.02
#...
·
·
1
like
|
|---|
John Hayes: The FLI Dovetail system on the filter wheels definitely causes a problem. When I had the ML16200 camera I used some tin foil to solve the same problem you experienced here. On my PL16803 the camera body has two holes that can bolt the camera body to two holes that are on the filter wheel cover. You still have the two set screws in play, but I found that mounting the body to the wheel, and then using the set screws, eliminated the gap issue. I used the GTX + 16803 combo on this image: The Great Orion Nebula and Running Man (rockstarbill) - Full resolution | AstroBin Which came out pretty good, so I did not really play with tilt on it much further. |
|
20.43
#...
·
·
3
likes
|
|---|
|
There is once scenario that I'd like to discuss, where there is more than one "optimal" backspacing. I've presented that optimal backspacing is achieved when the corner curves are perfectly aligned with the center curve, which is true, however this assumption is with the good faith that your corrected field is large enough to accommodate the size of your camera sensor. What happens when your sensor is LARGER than the corrected field? How does this analysis potentially fall short with achieving an optimal result? Ultimately, it really depends on what your goals are. Do you want to optimize for corner star quality by choosing a backspacing that might be a compromise of star quality across the entire field, or do you want to optimize for the corrected field your telescope is able to produce at the expense of the corners, knowing that you can just crop these out? This is one of those scenarios where you cant have your cake and eat it too. If you want the least bad corner stars possible, it will be at a slight detriment to the rest of the field. Conversely, if you want perfect stars as far out as the optics will support, your corner will suffer. You cant beat physics here. You can try focus offsets and spacing offsets to try and cheat the physics, but the optics will only deliver what they can. My AP130GTX with the QTCC Reducer and the IMX455 chip falls into this exact scenario. The corrected field is not large enough to accommodate this chip. So, I decided to see what was the best result I could get if I: 1) Set backspacing to optimize for the corrected field that the telescope with reducer could deliver (smaller area than the chip), or... 2) Set backspacing to optimize for the corners. When you use the NINA Hocus Focus Analysis with default settings, it will direct your backspacing for the first option. It will advise you to adjust backspacing so that all of the curves align, which corresponds to optimizing for the corrected field of the telescope. If this is your goal, you dont need to get fancy with anything, just follow NINA's advice with the understanding that you will end up with excellent stars across the corrected field and just crop out the bad stars in the corners. This is what the NINA graph looks like when you are optimizing for the corrected field of the scope. This is with 104.4mm of backspacing.  All of the curves for corners and center align nicely. There is almost no tilt, and spacing error is negligible. However, you will see that the corner stars look absolutely terrible!  Corners suffer from severe astigmatism, elongation, etc... They just look awful. If you evaluate the full frame however, you will notice that stars look excellent for about 80% of the frame, then they deteriorate quickly from there. If you just crop these out, you get excellent performance for the corrected field of the scope. If you want to find the optimal backspacing that will improve corner stars you will need to do a little trial and error. What I did was reduce spacing by about 0.1mm at a time, until I found the spacing where corner stars were the best. I found this to be at 103.5mm of spacing. Here corner stars looked much better. Note however, that the NINA graph shows backspacing error. Which of course there is! I was intentionally introducing backspacing error to try and improve corner stars.   You might look at this new result and say that it looks much better. And I agree, the extreme corners look much better. However when you evaluate the entire frame you will notice that in order to get the least bad corner stars, you had to give up on star quality further on-axis. In other words, the entire frame suffers from mild field curvature due to this spacing error. It's seen as a flare on the on-axis side of brighter stars. Whats the best choice? That's up to you. Which goal above are you trying to achieve? The best corner stars possible, or the best stars possible across the CORRECTED FIELD OF THE TELESCOPE? With this I think it's helpful to take a look at the full frame test images. I have included them here for download: https://www.dropbox.com/sh/pzj7p3dyki2n55f/AADlNguuisLplPHUkf6j-XRma?dl=0 File names are self explanatory. -------------------------------------------------------------------------------------------------------- A note on the typical backspacing reference that I have seen posted literally hundreds of times over the years. I really DO NOT like this diagram.  I'm going to repeat it, I really DO NOT like this diagram. There are multiple reasons I do not recommend it. 1) Your focus position can obscure what your corner stars actually look like. If you have any focus error, the offset can change the shape of your stars. You can make your stars look like they radiate out from center or around the center simply by changing focus position. 2) If your corrected field is too small for the sensor, this guide will only help you optimize your field for the corners. This might not be your goal! It doesnt help you reduce and eliminate field curvature, it simply helps you to get better corner stars. For me, I would prefer a better overall field and to crop out the corners. 3) If you have any other error such as tilt error, it really confuses things. My recommendation is to ignore this diagram entirely. NINA and ASTAP provide much more robust tools for analysis. By the way, I did a test integration where I optimized my QTCC reducer for the corrected field (at the expense of the corners). In case you want to see it, here it is:  Click for full resolution. Corners look terrible, but for the 80% of the frame that fits within the corrected field... it looks really great! |
|
11.02
#...
·
·
1
like
|
|---|
| I also do not like that diagram. I've never found it to be accurate personally. The tooling, as you mentioned, is much better to use and provides actual results of the system being tested. |
1.81
#...
·
·
1
like
|
|---|
|
yeah - i was going to ask about the scenario you just presented last. i'm pretty sure i'm in that situation - pushing a GSO 6" RC too far with a 0.8x reducer/flattener which was not designed for it. NINA tells me i have some tilt which i need to solve, but my feeling is that no matter the reducer spacing, i'm going to have an improperly corrected field. my assumption is that the field curvature of this small RC is already too much and using the reducer is pulling in "junk" stars from outside the original FOV. maybe i'm wrong but i haven't found too many images online from similar setups. i found a guy with here on astrobin with an RC6 and a (IIRC) CCDT67 but the images are too small to really evaluate the corners. the correction error was somewhat negligible with a 4:3 sensor (8300M) but when i upgraded to the ASI2600MM the corners are pretty awful, to the extent that i'm just cropping them out. probably i should just invest in better telescopes  i suppose the 2600 is sensitive enough that i could just ditch the reducer and run at f/9 with the native FL. might be worth trying to separate out the reducer issues from any curvature issues. |
|
11.02
#...
·
·
1
like
|
|---|
yeah - i was going to ask about the scenario you just presented last. i'm pretty sure i'm in that situation - pushing a GSO 6" RC too far with a 0.8x reducer/flattener which was not designed for it. NINA tells me i have some tilt which i need to solve, but my feeling is that no matter the reducer spacing, i'm going to have an improperly corrected field. my assumption is that the field curvature of this small RC is already too much and using the reducer is pulling in "junk" stars from outside the original FOV. The AP 27TVPH is a little better at producing good fields on RC scopes since it does have a minor flattening effect. You can get one of those and try it out. The spacing can be adjusted to increase or decrease the amount of compression. This should allow you to find a sweet spot for that chip. |
7.67
#...
·
·
1
like
|
|---|
Paul Wilson: The RASA 8 has very little backfocus space available, which makes getting a tilt corrector on it very challenging. I faced a similar challenge getting the Gerd Neumann CTU on the Epsilon given it's 56.2mm spacing from the corrector. The CTU took up too much, so the Octopi was the only feasible option. So, I am a fellow RASA8 owner. I found the best way to adjust it is to just treat it like you would a hyperstar. What you do is find enough extensions to lift your camera above the central obstruction to expose the collimation screws. Make sure that when the camera is raised up its in the orientation you will use it for imaging - within a few degrees will be fine. The first thing is to check collimation in the centre via a bright star like you would for a typical reflector. Adjust if the best you can. If you have a tri-bahtinov mask you can use that. After that, adjust the corners until they look the same. They will all look atrocious because of how much the camera is raised up, but as long as they look the same kind of atrocious then you have done your job! When you take off the extensions and set the required backfocus again it might need some minor adjustments, but at that point you've figured out how to adjust it from the first step and you are good to go. The limitations of this are that its a per-camera and per-orientation thing. If you are afraid of screwing up the collimation, even after using this method, a tri-bahtinov mask showed that things were still fine. CCD Inspector or your favourite tilt diagnostic program will really help you at this stage. What you are essentially doing is avoiding the fun times that I am sure many a RASA8 owner has had to do, which is taking a picture; checking it; unplugging and unscrewing the camera; adjusting either the collimation screws or the tilt plate screws a bit; screw and plug the camera back in; take another picture; scream internally; unplug and unscrew the camera.... Instead of pulling out my hair for weeks trying to adjust the RASA8, I can do it in an hour tops now. I never got to try this on the 2600MC before I sold it, though. What you could do with the 2600 - if you still have the tilt plate on it - is to just use the extensions to raise it up enough to get at the allen screws and do what I said above for the corners. I was going to get the Octopi but the price scared me away. I would've rather sold the RASA than pay for something almost half the cost of my RASA8 just to have good stars. If you have a QHY camera, honestly its just best to pony up for the FCCT. I hear some people have jimmied up a way to use the FCCT I on smaller ZWO cameras.  RASA Redemption at the Elephant Trunk The method in action. |
|
1.81
#...
·
·
2
likes
|
|---|
|
thanks - i'll give that a try - i also own a CCDT67 for my 10", which worked pretty well there (but again with 4:3 so not a challenge) i guess i just want to make sure it's not a fools' errand before embarking. looking at the 27TVPH's graph though it does indeed seem to have a very wide range of allowable spacing/compression so i agree that the reduction could be lowered until the corner stars look OK, if indeed the problem is what i think it is. gonna need some new plumbing tho |
10.80
#...
·
·
1
like
|
|---|
| Excellent work Chris. I too use Nina Hoccus Focus and I found your instruction cleared up a couple things for me. One thing that always confuses me is the orientation of the sensor in the Nina graph though for tilt adjustement. When I look at my camera from the back, is the top left corner actually in the Nina graph actually the top left corner of the camera? |
|
20.43
#...
·
|
|---|
yeah - i was going to ask about the scenario you just presented last. i'm pretty sure i'm in that situation - pushing a GSO 6" RC too far with a 0.8x reducer/flattener which was not designed for it. NINA tells me i have some tilt which i need to solve, but my feeling is that no matter the reducer spacing, i'm going to have an improperly corrected field. my assumption is that the field curvature of this small RC is already too much and using the reducer is pulling in "junk" stars from outside the original FOV. Yep, reducers in general dont perform very well. If you start out with a nice large corrected field than a reducer will probably work with APS-C, but if you already have a limited field the reducer is a recipe for bad tasting stars. Honestly, if you dont need the FOV that the reducer will achieve you'd probably be better off imaging at the native FL, and downsample for a better result but with a smaller FOV. Bills suggestion might be a good one. Honestly I just dont have any experience with RC's. I know with refractors its easier to get larger corrected fields with larger apertures. Maybe that is the same with an RC? I assume 6" is pretty small? In any event, the 3.76um pixels are not very forgiving are they!? |
|
20.43
#...
·
|
|---|
They will all look atrocious because of how much the camera is raised up, but as long as they look the same kind of atrocious then you have done your job! Neat idea raising the camera to get things pretty close. You are correct that a field with no tilt will produce a symmetrical field regardless of spacing error (or even focus error for that matter). Tilt free is tilt free no matter how bad the backspacing is. I'd only be worried that with such a fast scope that the spacers used to raise the camera might vary enough that there might be some residual tilt left over. Sounds like it worked for you though, and got you close enough that only a final tweak was needed. BTW- I just looked at your RASA image of IC1396. Thats probably one of the best corrected RASA images I have seen. Nice work. |
|
20.43
#...
·
·
1
like
|
|---|
Steeve Body: Are you asking about the graph or the table? I assume the table. The top left will correspond to the top left when you are looking at your camera from behind it... as long as your camera isnt upside down. In that case top left would be bottom right. The nice thing about the NINA analysis is that you can test it. Make an adjusment and then run it again. Did it get worse or get better? If it got worse, you went the wrong way. Kind of a cheap answer, but it is true. Also, NINA hocus focus will store the values in another table so as you perform more and more runs it keeps this information in a list so you can easily see if you made an improvement or not. Is this what you are asking? |
|
10.80
#...
·
|
|---|
Steeve Body: Yep exactly what I was asking thanks! Also while I’m at it, would you recommend picking a part of the sky that is near the zenith when making measurement/adjustments to avoid any focuser sag? |
|
20.43
#...
·
·
2
likes
|
|---|
Steeve Body: Sure, but if you have focuser sag there is a good chance this is showing up as tilt in your field. I do think that pointing at the zenith is a good idea. Depending on what tilt device you use, having gravity pull uniformly down on your setup as you tighten and lock it in place is important. For some reason the Octopi on my 268 requires me to do this. With my 600 camera, I can just point at a location in the sky that puts the camera conveniently in front of me when I'm sitting in a camp chair. This is tricky work, stacking the deck in your favor can definitely help! Eliminate sag, make sure your mount is tracking well. The NINA analysis will only be as good as the data you give it. EDIT- Of course, pointing at the zenith would only work for a refractor or reflector where the camera mounts on the rear. For a Newt, you'd probably be better off pointing somewhere else where the camera/focuser seems well balanced. |
|
20.43
#...
·
|
|---|
John Hayes: Thanks John. Theres always something with this hobby isnt there? I'm surprised to hear from you and Bill both that the FLI dovetail wasnt able to retain orthogonality. Thats a big bummer, I'd expect more from FLI. Have you had any problems with the Doveloc? I have found it to compress surfaces accurately after rotating. The only thing I have noticed is that there is a little wiggle so you can get a little displacement of the image train if you tighten one screw down more than the others. I havent seen this result in any sort of tilt issue or anything. I just noticed it by chance when I rotated the image train and my focus star was off center by a bit. I only bring this up, because if you are going to dial your IMX455 chip in on this setup and you have a stubborn corner, consider that it might not be perfectly centered in the optical axis of the scope. I dont know if this tiny bit of error could have a meaningful impact. By the way, I cant imagine what 0.6" seeing must be like! Be sure to post an image! |
|
10.80
#...
·
·
1
like
|
|---|
Steeve Body: Thanks Chris for your help Chris. I just acquired an Askar 107PHQ which is a quad and had no backfocus distance required, just reach focus and you are good to go which is great. I've tried to fix a small tilt issue that I had since Nina showed a 0.2 tilt factor in the aberation inspector... I used the astrodymium shim and got it down to about 0.06 in Nina but it is really fiddly to get right that way... So I have ordered a TS Optics M68 tilt adjuster since that is the only one I could find that would fit in my system and not break the bank. Is that ok to put the tilt adjuster before everything else in my rig...?  One thing that worries my a bit is that the other night even though I got the tilt down to a minimal amount with my sensor at zero angle, when I started to image on the other side of the meridian my tilt went from a 3% in CCD inspector from the test shots I used to adjust tilt with Nina to 20%... Hard to tell right now if the meridian flip is what caused it or the rotator going from 0 to 175 degrees... or both, or something else I have not had time to test again... but trying to figure out what are the possible causes so I can correct the issue. right now I have to have a small 1mm spacer in front of my astrodymium shim so the little lip/plastic handle does not press against the filter wheel making it impossible to use... not ideal to say the least |
|
20.43
#...
·
·
2
likes
|
|---|
Steeve Body:Steeve Body: The 107PHQ is a pretty sweet scope. I've seen some really impressive results from people. Even with the full frame IMX455 chip it does an admirable job. Probably one of the most compelling scopes on the market for price to value. So, that may or may not be a great place to put the tilt device. The problem is, if your tilt source is somewhere between the rotator and the sensor, and you correct for tilt at the OTA, then as soon as you rotate that Falcon, your tilt mitigation is nullified. Ideally you would get it between the rotator and the rest of your image train, and hope that there is no tilt source between the rotator and the OTA. Does this make sense? Curious on your meridian flip test. Did your rotator rotate the image train to get back to 0 degree sky angle, or does your capture software accept 180 degrees as a valid angle? (EDIT- You answered this. Yes, it did rotate. There is no need to rotate after a flip. Your images will register just fine even though your sky angle changes by 180 degrees) How was focus between the pre and post meridian flip shots? ANY FOCUS ERROR WILL AMPLIFY TILT EFFECT. So if you were at perfect focus for the 3% tilt shot, but focus was a little soft in the post-flip image, it could easily measure as 20%. Have you ruled out focuser sag? Simply point somewhere above the eastern horizon and take a pic. Then slew to the western horizon and do the same. If the tilt changes in the frame you have something sagging. As always, seeing data helps. If you can share full frame camera fits through drop box or G-Drive it would be helpful. |
|
1.81
#...
·
|
|---|
Yep, reducers in general dont perform very well. If you start out with a nice large corrected field than a reducer will probably work with APS-C, but if you already have a limited field the reducer is a recipe for bad tasting stars. Honestly, if you dont need the FOV that the reducer will achieve you'd probably be better off imaging at the native FL, and downsample for a better result but with a smaller FOV. yeah - in fact i have been binning the 2600 with this setup because at ~1100mm with the reducer the pixel scale is 0.7app even at bin2. that's way, way oversampled for my skies and of course it would only get worse without the reducer. 6" is pretty small, yeah, and the field curvature gets worse as the mirror gets smaller. not sure if there are smaller RCs but i have never seen one. generally the RC only has field curvature so flatteners take care of that, provided the flattener is matched properly. Teleskop-Service does seem to have flattener compatible for the 6" so maybe i should just stop messing around with the reducer and try that, just to know what is possible. |
