AstroBin is the home of astrophotography - Share full resolution images, preserve acquisition details, discover equipment, discuss techniques, and learn from the community.Discover AstroBin

Differential Flexure

Tony GondolaSonnyEAnthony (Tony) JohnsonMichele GuzziniJohn Hayes
37 replies531 views
Tony Gondola avatar

There have been more than a few threads here and elsewhere talking about differential flexure, usually in reference to why OAG is the better way to go. One thing I’ve never seen is someone actually measuring it. On a very basic level, it would seem that a good general test would be to point your system at the lowest altitude you would normally image and take a long exposure, say 10 min. If guiding remained good through that exposure yet the stars in the sub are not round, that would certainly indicate that something is moving. No way to tell if it’s the image train, guide scope or anything else but it would demonstrate that there is a problem. If the result of the test is round stars then all is good and if you benefited from an OAG, it wasn’t because of DF but something else.

I could also imagine running two instances of NINA so that you can plate solve the main camera and guide camera. Start off pointing at the zenith, plate solve, step down to a lower elevation and repeat, down to your personal elevation limit. You could then take that data see if the pointing between the main scope and guide scope diverges with elevation.

Thoughts?

Well written Helpful Insightful Respectful Engaging
andrea tasselli avatar

Mirror flop (for cats and reflectors in general) can give the same results as “differential flexure” in that context. Unless the guide scope is mounted on foam or some other crappy material I can’t see a physical justification for actual differential flexure in the classical sense nowadays (let’s say for typical exposures not exceeding 600s). Incidentally, that would manifest itself as a shift in the direction of the gravity vector whilst the former (DF) would manifest in a shift not strictly in the direction of the gravity pull, generally speaking.

Tony Gondola avatar

Good point and I would expect that it could represent that way if you took the data. Measuring the difference in pointing would only tell you that something is shifting, not what is shifting. That said, I would expect that the shifts caused by a mirror mount with to much freedom of motion would show as a sudden change in pointing where that due to flexing due to gravity would tend to show something more like part of a sine curve as you go through different elevations.

Helpful Engaging
andrea tasselli avatar

There wouldn’t be just a sharp difference in position but, once reached a critical inclination, there would an initial faster drift followed by a slowing down as the mirror adjusts to an equilibrium position (and I know for direct experience). Differential flexure harks back to the time when we used to take hour-long integrations or even before when people used big slow refractors to take emulsion-plate exposures of several hours. Allan Sandage used to ride the cage of the Hale great reflector for nights at a time, guiding on off axis reticule to avoid such a thing (he was particularly proud of not having to take a leak the whole night long). But I digress; the point of differential flexure is that, if a thing at all, would need to manifest over some length of time and our typical integrations are shorter than the time it take for the gravity vector to change enough to manifest a different tracking path between main and guide scope.

Engaging
Tony Gondola avatar

That was my initial take on the subject also, that with typical integration times used today that DF is much less of an issue. Still, I think it would be interesting to actually measure it as a sanity check if nothing else. This is what happens when you don’t have any clear nights for several weeks at a time. The rabbit holes become tempting!

One thing I didn’t realize when I posted on this is that NINA already has a plug-in to measure and and correct DF. It might be interesting to add this to a session, just to see what the system looks like.

Well written Respectful Engaging Supportive
Alex Nicholas avatar

To what end?

I think that with regard to differential flexture in an imaging system, knowing ‘how much’ is not valuable… Knowing it exists is 100% important, but knowing that its say, 5”/hr or 1.23” per degree of scope tilt…

The complication will be (as I’ve found in systems in the past) that differential flexture may affect you worse at certain declinations, depending on where and why the flexture is occuring…

I think, as long as you can identify that your guiding is reporting no star drift, and your images are showing star trails, then you know… and you fix it. Spending time measuring ‘how much image degrading differential flexture exists’ seems like a completely unnecessary set of mental gymnastics… Identify that there is flexture - identify and rectify flexture…

Its kind of the same as if I indentify that I have a stone in my shoe… I’m not going to spend hours trying to think of all the places from which the stone may have originated, or trying to discern the size of the stone by how much it’s impeding my ability to walk… I’m just going to remove the stone and get on with my day.

Tony Gondola avatar

In my defense, I bring up the subject partly out of boredom (no useable nights) and partly in the quest of learning. To me it’s an interesting exercise in understanding our systems better. I also like to be on solid footing when subjects such as OAG come up. Honestly, I don’t see how investigating something like this is a waste of time even if it doesn’t directly apply to my current system.

Well written Respectful
John Hayes avatar

Over the years, I’ve seen numerous efforts to measure differential flexure. I think that the main reason to do it might be to figure out how to at least minimize it within any given structure. Telescopes like most SCTs that rely on moving the primary mirror will exhibit non-repeatable image motion simply due to the way that the primary mirror is mounted (on a slider). Larger scopes, like my ASA600, will show mechanical flexure that might be more repeatable so that you could analyze it and maybe make it more rigid. But, in the end, it doesn’t matter. Guiding through the scope is always the best way to deal with flexure so it’s not worth a lot of effort to try to measure it.

John

Well written Helpful Respectful Concise Engaging
Bill McLaughlin avatar

After over 30 years of guiding and having tried nearly every combination (AO/OAG, OAG, Guide scope and hybrids) I have always come back to the same conclusion and that is guide scopes can work but sooner or later will cause you problems.

The worst part is that due to the myriad of combinations of variables (mount, guide scope mount, focal lengths, type of scope, position in the sky and more) what may work one time will not work another. Just when you think everything is perfect and reliable, flexure will suddenly bite you again.

My conclusion is that with very few exceptions they are just not worth using and for that reason investigation into variables is really an academic exercise. Might be fun as an exercise but that is about it.

Just my .02

Well written Respectful Concise Engaging
Tony Gondola avatar

I think it’s clear that OAG is the best way to go, just do it and forget it. In my case though, it’s not possible as I don’t have enough back focus to accommodate it. I’m not having a visible problem with my subs at the moment but it would be nice to know where the limits are. The NINA plug-in should make the assessment part pretty easy so I’ll probably integrate that into my next few sessions just to see what I get.

Well written Concise Engaging
SonnyE avatar

I think it is clear you are so bored, you are goosing butterflys.

Tony Gondola avatar

How do you goose a butterfly?

Michele Guzzini avatar

@Tony Gondola

I am the maintainer of NINA Flexure Compensator plugin. I personally use it on a 12” GSO RC telescope, this is just an example with this setup, 300s exposure

Flexure Compensator OFF4a21a397-86f3-4b9b-9452-a638f1c5999e.jpg

Flexure Compensator ON509b9ecc-532a-4c21-97e1-edafe17decb6.jpg

Helpful Concise Engaging
John Hayes avatar

Michele Guzzini · Jun 30, 2026 at 12:28 PM

@Tony Gondola

I am the maintainer of NINA Flexure Compensator plugin. I personally use it on a 12” GSO RC telescope, this is just an example with this setup, 300s exposure

Flexure Compensator OFF4a21a397-86f3-4b9b-9452-a638f1c5999e.jpg

Flexure Compensator ON509b9ecc-532a-4c21-97e1-edafe17decb6.jpg

You are almost there. Now compare that to OAG or ONAG, showing the performance over different regions of the sky and with some FWHM analysis.

John

Michele Guzzini avatar

@John Hayes

I think there may be a misunderstanding of the goal of the plugin. It is not meant to replace OAG or ONAG, nor to prove it is a better solution than them. Those are obviously the preferred solutions when they are possible.

The purpose of the Flexure Compensator is to provide an additional tool for cases where OAG/ONAG cannot be used, are not practical, or simply to help diagnose whether guiding issues are caused by differential flexure.

In my own case, it made a previously unusable setup usable. Of course, fixing the mechanical source of the problem is always better, nobody disputes that. But sometimes software compensation is a practical approach to mitigate real-world limitations, just like autoguiding compensates for mechanical tracking errors instead of requiring a mount with zero periodic error.

It's just another tool, and everyone can choose the approach that best fits their setup.

Well written Helpful Insightful Respectful Engaging Supportive
Tony Gondola avatar

Michele Guzzini · Jun 30, 2026, 12:28 PM

@Tony Gondola

I am the maintainer of NINA Flexure Compensator plugin. I personally use it on a 12” GSO RC telescope, this is just an example with this setup, 300s exposure

Flexure Compensator OFF4a21a397-86f3-4b9b-9452-a638f1c5999e.jpg

Flexure Compensator ON509b9ecc-532a-4c21-97e1-edafe17decb6.jpg

Thanks for showing that!

John Hayes avatar

Michele Guzzini · Jun 30, 2026 at 03:02 PM

@John Hayes

I think there may be a misunderstanding of the goal of the plugin. It is not meant to replace OAG or ONAG, nor to prove it is a better solution than them. Those are obviously the preferred solutions when they are possible.

The purpose of the Flexure Compensator is to provide an additional tool for cases where OAG/ONAG cannot be used, are not practical, or simply to help diagnose whether guiding issues are caused by differential flexure.

In my own case, it made a previously unusable setup usable. Of course, fixing the mechanical source of the problem is always better, nobody disputes that. But sometimes software compensation is a practical approach to mitigate real-world limitations, just like autoguiding compensates for mechanical tracking errors instead of requiring a mount with zero periodic error.

It's just another tool, and everyone can choose the approach that best fits their setup.

That’s all good. My point wasn’t that the tool is not useful or not a good idea. My point is that you could go just a little further to show how well it works relative to guiding through the scope.

John

Michele Guzzini avatar

John Hayes · Jun 30, 2026, 04:05 PM

Michele Guzzini · Jun 30, 2026 at 03:02 PM

@John Hayes

I think there may be a misunderstanding of the goal of the plugin. It is not meant to replace OAG or ONAG, nor to prove it is a better solution than them. Those are obviously the preferred solutions when they are possible.

The purpose of the Flexure Compensator is to provide an additional tool for cases where OAG/ONAG cannot be used, are not practical, or simply to help diagnose whether guiding issues are caused by differential flexure.

In my own case, it made a previously unusable setup usable. Of course, fixing the mechanical source of the problem is always better, nobody disputes that. But sometimes software compensation is a practical approach to mitigate real-world limitations, just like autoguiding compensates for mechanical tracking errors instead of requiring a mount with zero periodic error.

It's just another tool, and everyone can choose the approach that best fits their setup.

That’s all good. My point wasn’t that the tool is not useful or not a good idea. My point is that you could go just a little further to show how well it works relative to guiding through the scope.

John

Yes, that would definitely be an interesting test. Unfortunately, at the moment I don’t have an OAG available to test it on the same setup. Setting up a proper comparison would also require quite some time, since the telescope is a remote setup (not far away, but still not at my place), which makes testing more complicated.

Since the plugin is free and open to everyone, it would be great if someone with the right equipment could perform a more extensive and rigorous comparison.

Well written Concise
SonnyE avatar

Michele Guzzini · Jun 30, 2026, 03:02 PM

@John Hayes

I think there may be a misunderstanding of the goal of the plugin. It is not meant to replace OAG or ONAG, nor to prove it is a better solution than them. Those are obviously the preferred solutions when they are possible.

The purpose of the Flexure Compensator is to provide an additional tool for cases where OAG/ONAG cannot be used, are not practical, or simply to help diagnose whether guiding issues are caused by differential flexure.

In my own case, it made a previously unusable setup usable. Of course, fixing the mechanical source of the problem is always better, nobody disputes that. But sometimes software compensation is a practical approach to mitigate real-world limitations, just like autoguiding compensates for mechanical tracking errors instead of requiring a mount with zero periodic error.

It's just another tool, and everyone can choose the approach that best fits their setup.

Could this plugin be used as a check for flexure in refractor setups?

I don’t believe I have any problems in my equipment because of my mechanical setup, it is way beyond most for rigidity. But it sounds useful to add. Just to have present in application.

I’m using a 130mm aluminum tubed refractor with a 3.2” focuser assembly. I don’t use calculations as much as I prefer to weigh my image results for results of overall performance of my equipment on any given target. And so far, so good by normal benchmarks.

But I would be interested in your plug in for my NINA control. Just to see if it could be helpful for my particular setup. Might be a welcome refinement.

I’ve gone ahead and added it to my mount computer to try out in NINA. Thank You Michele, I’m looking forward to my future results trying out your plug-in.

Sonny, thinking from outside the box. 🤔

Respectful Supportive
Michele Guzzini avatar

SonnyE · Jun 30, 2026, 05:13 PM

Michele Guzzini · Jun 30, 2026, 03:02 PM

@John Hayes

I think there may be a misunderstanding of the goal of the plugin. It is not meant to replace OAG or ONAG, nor to prove it is a better solution than them. Those are obviously the preferred solutions when they are possible.

The purpose of the Flexure Compensator is to provide an additional tool for cases where OAG/ONAG cannot be used, are not practical, or simply to help diagnose whether guiding issues are caused by differential flexure.

In my own case, it made a previously unusable setup usable. Of course, fixing the mechanical source of the problem is always better, nobody disputes that. But sometimes software compensation is a practical approach to mitigate real-world limitations, just like autoguiding compensates for mechanical tracking errors instead of requiring a mount with zero periodic error.

It's just another tool, and everyone can choose the approach that best fits their setup.

Could this plugin be used as a check for flexure in refractor setups?

I don’t believe I have any problems in my equipment because of my mechanical setup, it is way beyond most for rigidity. But it sounds useful to add. Just to have present in application.

I’m using a 130mm aluminum tubed refractor with a 3.2” focuser assembly. I don’t use calculations as much as I prefer to weigh my image results for results of overall performance of my equipment on any given target. And so far, so good by normal benchmarks.

But I would be interested in your plug in for my NINA control. Just to see if it could be helpful for my particular setup. Might be a welcome refinement.

Sonny, thinking from outside the box. 🤔

Reflectors or refractors can both suffer from differential flexure when using an external guide scope, because the two optical trains are not rigidly coupled at the micrometer (pixel) scale. The problem can become more significant with longer focal lengths and smaller pixel sizes.

Long story short, if you are guiding with an excellent RMS but your stars are still showing some trailing, differential flexure could be the cause.

If your stars are already pinpoint, honestly I don’t see any advantage in adding Flexure Compensator to your imaging sequence.

Well written Helpful Respectful Engaging Supportive
SonnyE avatar

Michele Guzzini · Jun 30, 2026, 05:37 PM

SonnyE · Jun 30, 2026, 05:13 PM

Michele Guzzini · Jun 30, 2026, 03:02 PM

@John Hayes

I think there may be a misunderstanding of the goal of the plugin. It is not meant to replace OAG or ONAG, nor to prove it is a better solution than them. Those are obviously the preferred solutions when they are possible.

The purpose of the Flexure Compensator is to provide an additional tool for cases where OAG/ONAG cannot be used, are not practical, or simply to help diagnose whether guiding issues are caused by differential flexure.

In my own case, it made a previously unusable setup usable. Of course, fixing the mechanical source of the problem is always better, nobody disputes that. But sometimes software compensation is a practical approach to mitigate real-world limitations, just like autoguiding compensates for mechanical tracking errors instead of requiring a mount with zero periodic error.

It's just another tool, and everyone can choose the approach that best fits their setup.

Could this plugin be used as a check for flexure in refractor setups?

I don’t believe I have any problems in my equipment because of my mechanical setup, it is way beyond most for rigidity. But it sounds useful to add. Just to have present in application.

I’m using a 130mm aluminum tubed refractor with a 3.2” focuser assembly. I don’t use calculations as much as I prefer to weigh my image results for results of overall performance of my equipment on any given target. And so far, so good by normal benchmarks.

But I would be interested in your plug in for my NINA control. Just to see if it could be helpful for my particular setup. Might be a welcome refinement.

Sonny, thinking from outside the box. 🤔

Reflectors or refractors can both suffer from differential flexure when using an external guide scope, because the two optical trains are not rigidly coupled at the micrometer (pixel) scale. The problem can become more significant with longer focal lengths and smaller pixel sizes.

Long story short, if you are guiding with an excellent RMS but your stars are still showing some trailing, differential flexure could be the cause.

If your stars are already pinpoint, honestly I don’t see any advantage in adding Flexure Compensator to your imaging sequence.

OK, thank You for your input, and the Plugin. I have added it to my NINA plugins for a go. I do have very pinpoint star fields. But think it can’t hurt to try your plugin out to observe for any refinements. Your plugin makes my total of 3 plugins I’m using in NINA. The PHD, Hokus Focus, and now yours.

My entire assembly from the clamp up is built for support and rigidity. Thank You for your input.

Incidentally, about 16 years ago I did try to incorporate two different OAG’s into my first equipment. Neither worked. I contacted the supplier of the better one I tried and he stated to me it was not intended for use with a refractor, only SCT or reflector type telescopes. I took my guide camera of then and mounted up a 50mm guide scope. Instant guiding sucess!

Flash forward and today I still use my 60mm guide scope and a high resolution (ASI290MM) camera for my guiding. Not interested in OAG’s. Even though now days many do use them with refractors. Bravo refinements!

SonnyE, outside the box because I never got in the box. I just pick things out of it. (But avoid the cat droppings.)

Respectful
SonnyE avatar

Hey Michele,

I had a session of the Sadr region I’d taken the day before your post. https://app.astrobin.com/forum/topic/240772?post=324614&page=1

So, I took a session of the Sadr Region after adding your Plug In to my NINA.

Here are the two files:

📷 Sadr 60s x 200 imagesSadr 60s x 200 imageshttps://app.astrobin.com/i/7t0t4l/

📷 Sadr 40sx200Sadr 40sx200https://app.astrobin.com/i/j1o5ia/

This is probably as accurate a before and after as you can find. I do nothing to my images but stack them and save them as a jpg for web presenting. (Although, I did crop the stacking artifacts in these images.)

In my opinion, with my equipment, the plug in made a difference. The top one is 60 second exposures, while the bottom one is 40 second exposures. No other changes beyond the two other than that.

But I think I see an improvement after installing your plug in in my NINA. I’m going to continue using it in my NINA.

Thank You for sharing your Plug-in with the community.

Incidentally, You are welcome to download them and use then as examples if you would like.

Well written Respectful Engaging Supportive
Michele Guzzini avatar

SonnyE · Jul 5, 2026, 01:39 AM

Hey Michele,

I had a session of the Sadr region I’d taken the day before your post. https://app.astrobin.com/forum/topic/240772?post=324614&page=1

So, I took a session of the Sadr Region after adding your Plug In to my NINA.

Here are the two files:

📷 Sadr 60s x 200 imagesSadr 60s x 200 imageshttps://app.astrobin.com/i/7t0t4l/

📷 Sadr 40sx200Sadr 40sx200https://app.astrobin.com/i/j1o5ia/

This is probably as accurate a before and after as you can find. I do nothing to my images but stack them and save them as a jpg for web presenting. (Although, I did crop the stacking artifacts in these images.)

In my opinion, with my equipment, the plug in made a difference. The top one is 60 second exposures, while the bottom one is 40 second exposures. No other changes beyond the two other than that.

But I think I see an improvement after installing your plug in in my NINA. I’m going to continue using it in my NINA.

Thank You for sharing your Plug-in with the community.

Incidentally, You are welcome to download them and use then as examples if you would like.

You should compare apples to apples, exposures must be the same duration. Keep in mind that using Flexure Compensator will have the downside that it requires some time to make it’s own exposure and plate solving. So I reccomend to use it only when differential flexure is an objective issue.

SonnyE avatar

Michele Guzzini · Jul 5, 2026, 04:59 AM

SonnyE · Jul 5, 2026, 01:39 AM

Hey Michele,

I had a session of the Sadr region I’d taken the day before your post. https://app.astrobin.com/forum/topic/240772?post=324614&page=1

So, I took a session of the Sadr Region after adding your Plug In to my NINA.

Here are the two files:

📷 Sadr 60s x 200 imagesSadr 60s x 200 imageshttps://app.astrobin.com/i/7t0t4l/

📷 Sadr 40sx200Sadr 40sx200https://app.astrobin.com/i/j1o5ia/

This is probably as accurate a before and after as you can find. I do nothing to my images but stack them and save them as a jpg for web presenting. (Although, I did crop the stacking artifacts in these images.)

In my opinion, with my equipment, the plug in made a difference. The top one is 60 second exposures, while the bottom one is 40 second exposures. No other changes beyond the two other than that.

But I think I see an improvement after installing your plug in in my NINA. I’m going to continue using it in my NINA.

Thank You for sharing your Plug-in with the community.

Incidentally, You are welcome to download them and use then as examples if you would like.

You should compare apples to apples, exposures must be the same duration. Keep in mind that using Flexure Compensator will have the downside that it requires some time to make it’s own exposure and plate solving. So I reccomend to use it only when differential flexure is an objective issue.

Well, if my weather would cooperate I would. But have had some crap weather here.

Last night was crap clouds and fireworks. Maybe tonight?

Then of course, we are dealing with variables like seeing.

My point being, I liked the Sadr 40s better than the “naked” (without your plug in) 60s.

As soon as my local weather cooperates, I can do more than a one-off trial.

But I think I see a change with your Plug-in initially.

I’ve gone to great lengths to make my imaging train as ridged and stable as I can. But rather than just poo-poo your plug-in and go about my merry way, I thought I’d try it out. For a clear average it will take some time. So I think I will continue to use it.

Thank You for it.

Supportive
Tony Gondola avatar

SonnyE · Jul 5, 2026, 03:59 PM

Michele Guzzini · Jul 5, 2026, 04:59 AM

SonnyE · Jul 5, 2026, 01:39 AM

Hey Michele,

I had a session of the Sadr region I’d taken the day before your post. https://app.astrobin.com/forum/topic/240772?post=324614&page=1

So, I took a session of the Sadr Region after adding your Plug In to my NINA.

Here are the two files:

📷 Sadr 60s x 200 imagesSadr 60s x 200 imageshttps://app.astrobin.com/i/7t0t4l/

📷 Sadr 40sx200Sadr 40sx200https://app.astrobin.com/i/j1o5ia/

This is probably as accurate a before and after as you can find. I do nothing to my images but stack them and save them as a jpg for web presenting. (Although, I did crop the stacking artifacts in these images.)

In my opinion, with my equipment, the plug in made a difference. The top one is 60 second exposures, while the bottom one is 40 second exposures. No other changes beyond the two other than that.

But I think I see an improvement after installing your plug in in my NINA. I’m going to continue using it in my NINA.

Thank You for sharing your Plug-in with the community.

Incidentally, You are welcome to download them and use then as examples if you would like.

You should compare apples to apples, exposures must be the same duration. Keep in mind that using Flexure Compensator will have the downside that it requires some time to make it’s own exposure and plate solving. So I reccomend to use it only when differential flexure is an objective issue.

Well, if my weather would cooperate I would. But have had some crap weather here.

Last night was crap clouds and fireworks. Maybe tonight?

Then of course, we are dealing with variables like seeing.

My point being, I liked the Sadr 40s better than the “naked” (without your plug in) 60s.

As soon as my local weather cooperates, I can do more than a one-off trial.

But I think I see a change with your Plug-in initially.

I’ve gone to great lengths to make my imaging train as ridged and stable as I can. But rather than just poo-poo your plug-in and go about my merry way, I thought I’d try it out. For a clear average it will take some time. So I think I will continue to use it.

Thank You for it.

It would be good also to post a detailed crop rather than and full image. I think that would show any differences a lot better.