20.43
#...
·
·
3
likes
|
|---|
Jason Coon: To be honest, I'm not sure how helpful the Sensor Model output is for guiding tilt adjustments. I believe it forces the assumption the corrected field is infinite and helps you to determine the orientation of tilt. Notice that the values are exactly inverse from opposing corners rather than a unique delta from the average optimal focus position of the sensor. Again, I'm not sure how useful this part of the analysis is when it comes to telling you which screw to turn. I'd leave the sensor model feature disabled and just use the raw analysis per corner. Additionally, my post above which explains the problems with HFR analysis for the purpose of tilt, explains why the sensor model output is suspect. The sensor model assumes that tilt is perfectly represented by simple defocus... which in reality, it is not, with most optical systems. |
2.39
#...
·
|
|---|
Jason Coon: Thank you! So then, is there value in using the Aberration Inspector in NINA at all at this point? I used ASTAP and an Excel Spreadsheet last year to get my OctoPi dialed in on my RASA 8. I was hoping the Aberration Inspector would short-cut some of that tedious work. |
|
20.43
#...
·
·
3
likes
|
|---|
Jason Coon: ABSOLUTELY! Both programs are excellent, and suffer the same limitation. HFR and HFD are not perfect proxies for tilt analysis. I have found that ASTAP is slightly more accurate because the corner values are the delta between the center and the respective corner and that with HocusFocus it is the delta between the corner and an AVERAGE of all of the corners. This is significant, however only when you get close to a flat field. So both will get you close to a flat field. If you want to take it to the next level then that is when you switch to visual inspection. To succeed with this last step you will need an understanding of how your adjustments impact the image beyond what a piece of software tells you to do for an adjustment. In practice, I much prefer Hocus Focus for the simplicity of running the program and getting a recommendation. It's so easy! I've also gotten really adept at visual correction once I get close, so I don't agonize with Hocus Focus. I run it a few times, get close, then switch to visual. ASTAP is still easy to use though... so please dont interpret my comment as an endorsement of one over the other. Both are good tools . |
11.02
#...
·
·
2
likes
|
|---|
Jason Coon:Jason Coon: I used Excel and ASTAP to get mine close then did visual inspection to do the final tweaks. What's going to matter more than the tool is the sensitivity of the field (which is based on the focal ratio, size of the sensor, and the quality of field correction) as well as the optical system in general - i.e a field in a scope like a CDK in general will be easier than say an Epsilon 160ed because the CDK has no issue from a backfocus perspective. If you're in focus, you're good, while the Epsilon can have issues especially at the field Edge that may appear to be tilt related but could also be influenced by backfocus issues less than 0.1mm. Likewise operating on 70mm corrected field vs a 44mm corrected field leaves a bit of room for some error (i.e sensor isn't perfectly centered). That's why the fine threads on devices like the Octopi and Photon Cage are important. You want ultra fine control of tilt when you go to make the adjustments. |
6.15
#...
·
·
2
likes
|
|---|
Jason Coon:Jason Coon: I think before taking a decision or trying to decide from the conversations here, you would be better served by having a discussion with the developer. The NINA Discord channel is very active and has a section dedicated to Hocus Focus. The developer does have a full time job outside of this, so he may be 24-48hrs in responding but there are very knowledgeable users on-line that may also provide some additional insights. Both NINA and its various plug-in developers (including Hocus Focus) are in a state of constant development, adding functionality, updates/upgrades/fixes/etc. It is very much a moving target and the developer and plug-in developers do want to hear from users and get useful input (saying something is a piece of c**p is not useful input). As an example, in NINA, the current development beta had its fifth version update (the third in as many weeks). Development Version 3 which is utilizing new technologies (moving to .net framework ver 7 to enable the new features) is just getting underway. These nightly build versions are very much 'bleeding' edge and the current Version 3 NB is not yet recommended as dependable for routine use. Hocus Focus just published an update as well, which I had installed this AM and will be testing/running later this week as I changed out some items in my optical train that may (will) affect sensor tilt since they affect the back focus adjustment in the Photon Cage. I'll say this about the 'accuracy' of any of these mathematical model based tools - which includes ASTAP, CCD Inspector, and HocusFocus. None of them are perfect and like much of amateur astronomy how reliable/accurate you find them is more influenced by the users physical rig installation than it is by mismatches between reality and the models these tools are built around. If your system has flex (and when talking about measurements in micron/fraction of a micron, ALL systems have flex), there is going to be variance from the "ideal" result from run to run no matter how small it may be. Although not mentioned in this thread, the quality of seeing at the time has a tremendous impact on the focus data generated, its repeatability and thus a person's perception of its reliability/accuracy. So is using one of these tools (such as Aberration Inspector) worth while? Relative to doing nothing yes, but it depends on how critical/serious you are about your astrophotography. If you're talking about one vs. another, that is a matter of user preference, comfort with the UI provided by the developer, and the ease of interpreting the data for the purpose of making adjustments. Not too dissimilar to asking someone if Ford is better than Chevy or Ram. From my own perspective, I like ASTAP as a plate solver but abhor its user interface for further functions, but that is a personal preference. I found the Aberration Inspector relatively easier to implement and it gave definite indications of amount and direction of adjustment needed (definite for values of definite where mathematical models are concerned). Again, a user preference and not an indicator of accuracy. As for as the 3d display and graphic display of tilt HocusFocus - those are the 'shiny' bits that don't really mean much - IMO. The display (and the documentation) both say that neither are intended to indicate actual measurements/corrections but relative tilt. I don't pay any attention to them and keep them turned off since that is something that doesn't need to run for my purposes. Others may feel differently, again user preference. As far as the models go and their validity - EVERYTHING you deal with in software is based on modeling to some degree. All models are based on assumptions and to make them workable some things have to be taken as the boundaries just to keep it from bogging down in endless variations. The purpose is to get to the 80% solution, as it is with everything else. The remaining 20% of any task, model, performance requirement takes exponentially more effort/expense for very little return. Value judgement to do that work is dependent on how critical the purpose is. Life saving/life sustainment - and most everyone is going to say yes that is worth it to see how many decimal places you can shave off the delta from perfection. Hobbyist/Quasi-scientific freeware tools for amateurs - probably not. IMO - YMMV. |
|
20.43
#...
·
·
4
likes
|
|---|
|
On the topic of tilt mitigation in general, I will say that one of the biggest breakthroughs is the ability to separate stars in the linear stage of processing, stretch separately, and seamlessly blend back in. Tilt issues compound with field curvature to make really ugly stars. If your system is pretty close to tilt free, odds are only the extreme corners still suffer. This final stage of tilt mitigation is the most challenging part of the process. By stretching stars less, they look much better so you might even find its not necessary to beat your head against the wall squeezing every last drop of performance out of your imaging system. No matter how crazy you go with this, make sure you make some images along the way. Don't only test and work on fixing tilt. This last bit of advice is coming from someone who is the ultimate pixel peeper. |
|
6.15
#...
·
|
|---|
|
I do want add one additional piece of information for context concerning the micron measurements in the Aberration Inspector. What NINA/Hocus Focus have (and is the base measurement until the user provides additional data on the focuser) is the encoder readouts as the focuser moves. That is the only direct "measurement" that is displayed indicating the variation of focus points across the sensor. To present a measurement in microns, the user must provide the step size of their focuser in microns - the measurement of movement from one encoder dot to the next should be in microns. Not all electronic focusers have that information in their documentation. This information is the smallest measured movement your focuser can make, measured in microns. How accurate/repeatable that information is depends on the manufacturer. Until the user provides that information to Aberration Inspector (and dependent on how accurate that information is), it does not "know" anything other than encoder steps. The micron measurements it provides are a calculated value based on the 'hard' encoder data and the step size information provided by the user. In other words, they are another mathematical 'model' whose accuracy is dependent on many factors including just the user accurately entering the correct data on the micron step size of the focuser. I believe that is why the graphical representation are based on the delta in the encoder values and not micron measurements - at least they appear more closely aligned to those values regarding the scale shown. YMMV. |
|
20.43
#...
·
·
2
likes
|
|---|
I believe that is why the graphical representation are based on the delta in the encoder values and not micron measurements - at least they appear more closely aligned to those values regarding the scale shown. YMMV. Whether the algorithm uses step count or microns is not really important. It's relative either way. The delta is calculated using the curve vertexes. Either using the center as the reference (ASTAP) or an average of the corners (Hocus Focus). These outputs describe the relationship(s) of optimal focus position for the center and corners. It is not telling you how many microns to adjust your sensor plane. It's just telling you how many microns the delta is for the optimal focuser position for respective sensor areas. The positive or negative value is the guide for which direction to angle the sensor. How much is dependent on your optical system. |
11.99
#...
·
·
1
like
|
|---|
These outputs describe the relationship(s) of optimal focus position for the center and corners. It is not telling you how many microns to adjust your sensor plane. It's just telling you how many microns the delta is for the optimal focuser position for respective sensor areas. This is my understanding also. Optimal focal plane position does not necessarily correlate directly with backspacing. Therefore, the unit of measure of the movement of the focal plane is not critical. |
|
20.43
#...
·
·
1
like
|
|---|
Arun H:These outputs describe the relationship(s) of optimal focus position for the center and corners. It is not telling you how many microns to adjust your sensor plane. It's just telling you how many microns the delta is for the optimal focuser position for respective sensor areas. Correct. The only correlation is the direction of the adjustment required. |
5.96
#...
·
|
|---|
ASTAP has an image aberration inspector which will show tilt in your images. Be mindful that it shows tilt within the entire system and not just the camera. I have the 2600 both MM and MC and neither have tilt. I built a rig as shown in this video, https://www.youtube.com/watch?v=Gw2oTwb4GjY. *** Thanks Jeffrey! It may not be tilt specifically with the camera. It could be somewhere in the optical train (spacers+OAG+filter drawer). I tend to keep my optical train intact and keep it connected to the telescope but will check how tight everything is before imaging. I suspect I need to be more careful about how I screw in spacers/OAG and camera. *** |
|
5.96
#...
·
|
|---|
Arun H:These outputs describe the relationship(s) of optimal focus position for the center and corners. It is not telling you how many microns to adjust your sensor plane. It's just telling you how many microns the delta is for the optimal focuser position for respective sensor areas. *** I've only played around with tilt adjustment a couple of times using the Hocus Focus plugin with limited success. I wasn't so concerned with absolute numbers. As you say it's a measure of optimal focus position rather than adjustments needed. The hocus Focus analysis seems to be very long winded and timing consuming and I certainly don't want to have to do that every time I imagine. That's why I'm potentially leaning towards using the aberation inspector for tilt adjustment. *** |
|
20.43
#...
·
·
4
likes
|
|---|
That's why I'm potentially leaning towards using the aberation inspector for tilt adjustment. Robert, I've put together a document on tilt and spacing analysis. Parts III and IV are specific to single frame analysis using aberration inspector. See attached. 2022-11-29 A guide to fixing tilt and spacing with objective analysis-Chris White.pdf |
|
5.96
#...
·
·
1
like
|
|---|
That's why I'm potentially leaning towards using the aberation inspector for tilt adjustment. *** That's super helpful. Thanks so much Chris! *** |
|
5.96
#...
·
·
1
like
|
|---|
I thought this thread was about the ZWO 2600. Maybe a separate thread for the PlayerOne should be started if people want to talk about that camera. *** Ye, I wasn't intending this post to instigate a slagging match between people which is a bit disappointing but regardless there are some really helpful and insightful comments. Your analysis and document on tilt adjustment is outstanding. I'll give it a proper read this evening after work. *** |
|
2.39
#...
·
|
|---|
As a participant in that slagging match I apologize for the distraction.I thought this thread was about the ZWO 2600. Maybe a separate thread for the PlayerOne should be started if people want to talk about that camera. As the instigator of that slagging match I apologize for the distraction. Great info here indeed, and I really appreciate the effort Chris White put in to develop and share this excellent information with the community. |
|
5.96
#...
·
·
1
like
|
|---|
Jason Coon: *** We are all human. Thanks Jason. Your input is much appreciated too. I appreciate all the advice! Thank you*** |
|
2.39
#...
·
|
|---|
| When doing these Tilt adjustments do filters make any impact? I realize they do for focusing, but if you are simply working out Tilt and backfocus issues is it fair to assume that flat for one filter is flat for all? Is it better to do this with no filter (or a clear focusing filter)? |
0.00
#...
·
|
|---|
Just wondering how people have gone about tilt adjustment with the 2600MC camera? I used a Astrodymium Rotating Tilt Adjuster to get it properly adjusted. I don’t like to play with ZWO’s tilt plate because it can be like going down a rabbit hole. It worked well for me. |
|
20.43
#...
·
·
1
like
|
|---|
Jason Coon: This depends. As long as your filters are all the same thickness you should be fine regardless of what system you are using. With very fast focal ratio scopes though you could run into a problem if your filters are different thicknesses (example of color filters are a different thickness than your narrowband filters.) A different thickness filter requires a different optimal backspacing. Backspacing error results in field curvature issues. Field curvature compounds with tilt and makes it look worse. Filters don't directly impact tilt adjustments though you may see more of an issue on wider band pass filters than narrower. (Lum vs red for example). I think the best filter to use is red. I think the software does a better job measuring star size when it's more tightly focused with less scatter. More important though is a dense enough star field for your focal length and sensor size. Also more important is long enough exposures thay you resolve those stars. While it seems like a waste of time you will get better analysis with longer exposures. So 20 to 30 second is better than 5 second. |
|
2.39
#...
·
|
|---|
Chris White- Overcast Observatory:Jason Coon: Thank you! I'm doing these adjustments on both a RASA at f/2 (with an OctoPi adapter) and an EdgeHD reduced to f/7 (With an ASG Photon Cage). With both I shoot in OSC. So it sounds like I should be doing this with my dual narrow band filters in place. At 30-seconds on the EdgeHD I can see how that would improve start detection, but with the RASA I'm concerned that would over saturate the stars, so I'll scale that accordingly (it's about a 12:1 difference). I really appreciate your help! |
|
20.43
#...
·
·
2
likes
|
|---|
Jason Coon:Chris White- Overcast Observatory:Jason Coon: I'd use the dual filter with the RASA testing if that is how you mostly image. it will be inportant to get backspacing exact. For the edge it probably doesn't matter. In fact it would probably be easier to not use the filter. Spacing is more forgiving at f7. Don't use less than 6 to 10 seconds regardless. I use epsilon and have tested a broad spectrum of settings. I prefer 10 seconds and will sometimes use 30 seconds. It doesn't matter if you saturate some stars. You really need more stars. That's the key to better analysis. |
|
11.02
#...
·
|
|---|
Jason Coon:Chris White- Overcast Observatory:Jason Coon: Remember that focusing the RASA (and the Edge HD for that matter) will move the primary mirror. So as you are making any adjustments be sure that the scope is well focused before and after each change and limit the changes to be very, very small. |
9.03
#...
·
|
|---|
| Hi,Setting up a new scope (C11) for my ZWO ASI 6200 and recall all the fun flattening with my Stellarvue. Recall someone had figured out how to put a spring into the screw system to provide tension to obviate the push pull with two screws.Anyone have a good measurement/source for such tiny springs? Same issue I am sure with the 2600.JY |
5.73
#...
·
·
1
like
|
|---|
|
Hi Jeryyyyy. I don't know if it will work for your ZWO, but Starlight Xpress uses a rubber O ring between camera and tilt plate. I'm not sure if there is a machined groove for the O ring. I have never removed the tilt plate. Unlike the springs, an O ring will block light from entering. Lynn K |
