6
Review of iPolar hardware and software for polar alignment
Discuss this article in our forums
Review of iPolar hardware and software for polar alignment
By: Stephen Andrew Kennedy
From: Strawn, Texas (Near Fort Worth, TX)
Experience: Relatively New (but I have an old soul…)
Executive Summary: iOptron’s iPolar upgrade, which replaces the optical scope on the iOptron SkyGuider Pro, does not accurately provide polar alignment when compared to results measured using identical equipment operating Sharpcap Pro and the specified guide camera and scope. Data was collected while simultaneously running both the iPolar alignment software and SharcapPro on the same equipment. The deficiency is consistent with an issue in the iPolar system’s calculation of altitude parameters in polar alignment after plate solving.[1]
Equipment and software:
Tripod: Robus RC-5570 Vantage Series 3 Carbon Fiber Tripod
Mount: SkyGuider Pro Mount with factory-installed iPolar alignment hardware
Base: William Optics replacement base for SkyGuider Pro
Scope: William Optics RedCat 51
Guide scope: ZWO 30mm f/4 MiniScope
Guide camera: ZWO ASI 120 mm mono
Camera: Canon EOS Ra
Software: iOpton iPolar Software version 2.41 (current version at time of review)
Software: SharpCap Pro version 3.2.6448 (current at time of review)
Software: PHD2 2.6.9dev1
Issue Presented: After several months of operating factory-installed iPolar alignment equipment on the iOptron Skyguider Pro, the author observed consistent unsatisfactory polar alignment results with light exposures greater than 90s. Specifically, the author observed alignment degradation that was measurably constant over time beginning within minutes of achieving polar alignment even when a guide scope was applied with PHD2 tracking software. As discussed below, because PHD2 tracking software cannot adjust declination axis parameters of the SkyGuider Pro, the author hypothesizes that the flaw in the iPolar alignment system likely resides in altitude calculations after plate solving, but the author cannot completely rule out other inaccuracies in measurement, especially given the quantification calculations described below as registered by SharpCap Pro. The author acknowledges that the best practice to determine the exact issue would require inspection of iPolar software source code, which is not available to the author.
Discussion: The author observed that the iPolar alignment system could not maintain polar alignment for more than a few minutes when collecting exposures greater than 90 seconds. The author employed a guide scope and PHD2 software to correct observed deficiencies for exposures of more than 90s but found no significant improvement. At the suggestion of a cloudynights subscriber, the author compared SharpCap Pro polar alignment capabilities with the iPolar system. The author ran both polar alignment systems simultaneously over three consecutive evenings with identical errors observed in iPolar alignment during each test.
The experiment parameters included a routine of first achieving polar alignment with the iPolar alignment system, which uses a camera embedded in the SkyGuider Pro mount, and iPolar software version 2.41. Upon achieving polar alignment with the iPolar system, the author then checked polar alignment using SharpCap Pro version 3.2.6448 and the ZWO 30mm f/4 MiniScope and ZWO ASI 120 mm mono camera. SharpCap evaluates polar alignment as “poor,” “fair,” “good,” and “excellent” and provides specific numeric calculations on whether the alignment requires “up/down” and “right/left” adjustments, which the author understands to be altitude and azimuth calculations, respectively. On each test, after polar alignment had been achieved according to the iPolar system, SharpCap rated the polar alignment to be “poor” with both the up/down and right/left alignment errors to be greater than 1.00.00 while the iPolar alignment system software, running in the background, registered polar alignment to be on target. The author then captured 240 second duration exposures of the Rosette nebula and observed star trails after the second exposure even when using the PHD2 system and guide camera. Within approximately three to four minutes, the PHD2 guiding graph registered right declination corrections being sent to the mount which exceeded 4 on the graph and never corrected to less than 4 thereafter.[2]
After one hour of guiding with the iPolar system, the author changed the test parameters and conducted polar alignment using SharpCap Pro and the referenced guide camera. Upon achieving polar alignment with SharpCap, the author then compared the polar alignment parameters with the iPolar alignment system, which was still running in the background. The iPolar system immediately indicated that Polar alignment was off target. However, because the iPolar alignment system does not provide numerical measurements of alignment, but instead includes an image of the alignment, the calculated deficiency per iPolar is unknown. For reference purposes, the author includes images of the polar alignment measurement of SharpCap (showing “good” quality alignment) versus the simultaneous off-target measurement registered by the iPolar system in figures one and two below. The author then proceeded to collect 240 second exposures of the Rosette nebula over one hour using the PHD2 guiding software. The author observed no star trails in any of the images, and the guiding graph registered right declination corrections in the range of 1 to 3 during the hour long test period. An example of a 240 second exposure of the Rosette nebula at 1600 ISO using SharpCap as the polar alignment system during the test period is provided in figure 3.
Conclusion: Unless there is an equipment-specific anomaly unique to the author, which seems unlikely given the successful results in operating the SharpCap Pro polar alignment system with the same equipment, the author concludes that the iOptron iPolar alignment system is not reliable and cannot recommend using the iPolar system for polar alignment.
Stephen Andrew Kennedy
Attorney at Law
Strawn, Texas
Figure 1 below is the polar alignment measured by SharpCap on 11/20/2020 at 10:07 PM Central time.
Figure 2 is the polar alignment measured by iPolar on 11/20/2020 at 10:08 PM Central time.
Figure 3 is an image of the Rosette Nebula taken on 11/20/2020 at 11:35 PM Central time after nearly one hour after achieving polar alignment with SharpCap (converted from RAW to JPEG with Camera Raw, no adjustments to contrast).
[1] The author acknowledges that further data is required from other SkyGuider Pro devices to confidently exclude the possibility of an equipment-specific malfunction. However, the author notes that the SkyGuider Pro equipment works to specification when Shapcap Pro software is used with a guide camera to achieve polar alignment.
[2] The PHD2 software cannot correct declination errors for the SkyGuider Pro because the mount is limited to right-ascension adjustments.
- mrlovt, Maxtrixbass, stevesells and 3 others like this
64 Comments
SharpCap report shows how to move polar axis to be dead on north. The units are degrees, minutes and seconds of arc. Your screen shot showing 0.1.39 means you're off of north by 1 minute 39 sec. Not bad at all. A minute is 1/60th of a degree. And basically all of your error can be eliminated by moving axis too right by that amount. 5 sec down is basically nothing.
When SharpCap shows the error of 1.00.00 it means you're off by a degree, or twice the diameter of the moon.
Have you tried re-calibrating iPolar? (two exposures at different rotations about the RA axis).
If that doesn't work, you can manually set the center position. I did this same comparison with SharpCap and manually & iteratively adjusted the "center of camera" position until both pieces of software gave very similar results.
Besides the good question above, I did not see anything in your article that supports the conclusion that this is an error in calculation of atmospheric effects. I am not saying it is not, just that I saw no support for that statement.
I will say that on both the CEM40 and CEM70 I have found the alignment to be within anywhere from 1-5 arc minutes in most cases, without any significant effort, just get the plus pretty close to centered in the circle.
Thank you for your comment. This will confirm that troubleshooting included recalibration of the iOptron polar alignment system on three separate occasions with no observable improvement. As for manually and iteratively adjusting the center of camera position, the purpose of the review was to evaluate the product from the factory after calibration and advise potential users of how well the product performs when compared to a polar alignment product known to be reliable. The scope of the review was not to determine whether the user can modify the product to make it work within specifications, but rather, to review the product out of the box. Manual modifications of the camera could void warranties associated with the product, however, I have not analyzed the specifics of the warranty. For products out of warranty, a forum post advising how to manually adjust the center of camera position may be of value.
Manually adjusting the center position of the camera is done in software. Basically, telling the software which pixel is at the center of rotation. Definitely not modifying the product or voiding a warranty. It's a feature built into the software.
Edit: Setting the center position is exactly what the 2-image calibration does. But you can tweak it by a few pixels manually (in the software) to make it more accurate.
Re center that begs another question: Can an iPolar be installed (from the factory) somewhat off-axis? Wouldn't that lead to bad alignment? And if so -- other than bad alignment, how would one tell?
Unfortunately it does all sorts of filtering out non-stars, otherwise you could do something like swing it through 180 degrees while taking a photo and look at the star trails. If it's off axis, would there even be a non-moving center?
Note this does not negate the findings of the review, since if true it is clearly a problem with the iPolar (at least as installed). I'm just curious if there is a way to find out definitively WHY it is performing so badly for the author? Not just that it is.
Thank you for the clarification. I understood the phrase "manually and iteratively adjusting the center of the camera position" to mean a repetitive process of physically adjusting the position of the camera, which would could void the warranty. The software provides for 2-image calibration which was an exercise completed on three occasions to no effect. Other than the 2-image calibration, no other modifications were attempted to compensate for the deficiency. The purpose of the review was to compare the functionality of the iPolar alignment system from a factory-installed SkyGuider Pro when compared to a software product that uses a third party guide camera and scope attached to the same mount.
An offset of the axis of the electronic polar scope from the RA axis is what the calibration accounts for. If the scope was installed perfectly, there wouldn't be a need for calibration. With the scope offset from the RA axis, the star trails will rotate around a point that isn't the center of the sensor. The calibration procedure finds the pixel that it is at the center of rotation. I did the calibration and then polar aligned with SharpCap. Switch back to iPolar, notice it's not perfect, type in a new center position a pixel or two in the correct direction. Rotate around the RA axis to make sure it's stable at any angle.
You can also do this procedure without SharpCap and basically perform the calibration procedure on your own. Rotate around RA, visually see where the point it thinks is center rotates around, and adjust the center position until you bring the two into alignment. (I'm not talking about where it thinks north is, just where it thinks the center is compared to where you see the center is)
No idea why it's performing poorly for Stephen. I've noticed that the iPolar calibration will sometimes try to take the second image really quickly after I just barely move the RA axis. Calibration will be less accurate if the two images are only a few degrees apart. The software should wait for you to tell it that you're ready for it to take a second image, but it doesn't. I agree that the iPolar software could use some work, especially to make it work quickly right out of the box.
Stephen, let's say you have it polar aligned according to iPolar. What happens when you rotate 90 deg around RA? Does it go out of agreement? And what happens if you go another 180, so 90 deg the other way from the starting position? I'll be it drifts around, which would mean the center position is off in the software.
Another issue could be sag in the main scope with respect to the mount. In that case, the mount could be well-aligned, but SharpCap would give a different answer. I doubt that's the case for Stephen, as he saw star trails when aligned with iPolar but no trails when aligned with SharpCap.
I'm having trouble visualizing it, but I was not thinking offset, but not aligned. In other words, a line from the center of the lens to the center of the sensor may not be in the center of the axis, but it is parallel. What if instead it isn't, what if when the axis were polar aligned perfectly the whole iPolar assembly were pointing left 5 degrees or some such. I'm not sure the centering procedure can deal with that (I'm not sure it cannot either, as I said, having trouble visualizing it).
Is that true? I thought if the axis of the mount were properly aligned, that a mis-aligned scope still would track properly and drift align as polar aligned properly.
So do you mean what if the axis of the iPolar is at an angle to the RA axis of the mount?
I mean that the flexure changes with rotation. So not just misaligned, but changing as you move around the axis. For example, if you're at 90 degrees, a heavy scope on a weak mounting could bend down a bit. But then as you come up to 0 degrees, there's no bending. And between 0 and 90, the bend is somewhere in between. In this case, where SharpCap directs you for polar alignment would be different depending on which angles you used for the before and after positions. If you have a very rigid setup and the misalignment is constant throughout the range of rotation, I think you're right.
Correct. Basically the iPolar is a semi-cylinder that fits inside of a tube, and I think iOptron depends on the tube holding the iPolar along the same axis. What if it got in kind of cockeyed somehow. What would the impact be, or would it just magically work out?
I'm not sure. I kind of think it would work out, just like a main scope that is slightly misaligned. Cylinders are pretty easy to machine, though, and if it runs the length of the mount, how big of an angle could it reach, even with a little play?
Just grasping at straws in terms of the OP's problems.
I'm not completely convinced. I get how a main scope doesn't matter, but that's because the axis of rotation is aligned. In this case it's to get the axis aligned. But I don't know.
However the iPolar scope is aligned, it should rotate around the mount's RA axis. Hopefully, the axis of the iPolar is perfectly aligned with the RA axis. If it is angled, it will still rotate around the RA axis, but that will swing the iPolar about the RA axis. Like two chopsticks glued together at an angle, and you rotate one while holding it otherwise motionless. The other will swing around it. Like here, where P1 is the axis of the iPolar.
In the case that they are aligned but the iPolar is translated, it would look like this, where B is the iPolar and Q is the RA axis.
If you have both...idk, just use SharpCap!
I can only relate my own recent experience with the iPolar. I have an AP900 GTO mount permanently installed in my observatory. I retrofitted the iPolar camera with the recommended adapter. At first, I found the software to be rather "jumpy." I later learned that much of the jumpiness was due to strong atmospheric refraction. Later I achieved excellent PA and found that guiding at 2800mm with my Edge 11HD and a Lodestar through an OAG, PHD2 called for only small corrections. I took 35 180s subs with no trace of any trailing stars. Just sayin'
Michael
Attached Thumbnails
@barrabclaw, ok, I think i'm convinced. The center calculation will always be a center of rotation, so as long as the iPolar remains at whatever cant it is calibrated at, it should still polar align.
It would be interesting to know why it wasn't working for the OP. What could iOptron have done incorrectly in that one?
I looked at my guiding assistant runs from last night, and did two separate GA runs (was testing backlash really), east and west of meridian, one said 1.6' off, one 3' off. That's consistent with every attempt; I can't imagine how to get a full degree off. Something was clearly broken on the OP's.
I use factory-installed iPolar units in my CEM60 and CEM40EC. After polar-aligning using the iPolar software, I use the PHD2 Guiding Assistant to check the polar alignment. For the CEM60, I routinely get less than 1.5 arc minutes, and about twice that on the CEM40EC. I can guide my 2000-mm FL SCT with 0.25-0.35 arc-second rms errors. In my experience, iPolar is extremely reliable and a big time saver.
--Michael
I love this analysis. Great job Mr. Kennedy. Stuff should perform right away. Based on your observations under a double blind testing scenario your evidence is convincing.
Agreed. Not sure what it is but I'm pretty sure it has to do with the calibration and setting the center position. It would be really easy to tell if that's the case by seeing what the error is at two positions in RA, 90 degrees apart.
Isnt it possible to club iPolar camera with Sharpcap?
This way we eliminate the doubt on iPolar Camera being misaligned.
Some discussion above about how the iPolar camera is pointing.
I have a PoleMaster.
It really doesn't care where the camera is pointing. It can be translated, it can be pointed off axis. The software can cope.
It pulls up a screen with Polaris and surrounding stars. You use the arrow keys to rotate circles on top of surrounding stars. That tells the platesolving what's up. You rotate the mount and ID one of those stars at the start and the finish. That tells the software where the RA axis is pointed. Remember, it doesn't care where the _camera_ is pointed, exactly. It figures it out. It puts a green circle on the screen with the star on it. You rotate the mount back. The star staying on the green circle is confirmation the software has the pointing location of the RA axis right.
The PoleMaster then places a target where the RA axis is pointed. Places another target on the North Celestial Pole. You use the mount alt-az adjustments to superimpose them.
It. Just. Works.
Fast and easy. Polar alignment is better than it has to be.
Leaving the mount setup in my observatory I can skip the rotation step after doing it just once. Takes me 30-60 seconds to tweak polar alignment for the night, the mount is just sitting on the ground, rather than on a buried pier.
fyi, "the iPolar alignment system could not maintain polar alignment" -> This sentence does not make much sense, but I think I understand what you're saying.
Nothing in the review says whether the author used the iPolar camera in conjunction with SharpCap to achieve polar alignment. Doing so would have confirmed whether the issue came from the iPolar hardware or the iPolar software. Note: SharpCap is most definitely able to connect to the iPolar camera, which is seen by Windows as a ... webcam, which it kind of is (albeit with a slightly bigger lens)
I have a CEM70 mount, and adjusting the iPolar camera within the RA shaft via the set screws is actually a feature of the system. Assuming the author did not use the iPolar camera in conjunction with SharpCap to achieve polar alignment (see my previous point), it is not entirely impossible that the iPolar camera is loose, which could most definitely explain what he's been observing. I don't own the SkyGuider Pro, but I would expect that there is a way to replace the iPolar camera, and therefore check that it is tightly installed. It is a safe assumption that at this price point, some tinkering and troubleshooting may be necessary...
A lot of people (myself included) are getting incredibly accurate results with the iPolar camera + iPolar software (confirmed by PHD2 drift alignment, on numerous occasions, in my case) The iPolar camera hardware is desperately simple (like I said, it is pretty much a webcam with a slightly bigger lens...) so I doubt the problem comes from there (although if the lens or the sensor is loose, that could also explain it, but I think it's unlikely) The iPolar software is not great, but it kind of does an okay job (the plate solving is desperately slow...) With this in mind, it is clear that this review is not only inconclusive, but it appears to be flawed, and I would hate that people base their purchase decision on this review alone.
Well said. Using the iPolar camera with SharpCap, as a couple people have suggested, is a great idea for isolating the issue. If I had to bet, I would bet that it works well with SharpCap and that the problem is the calibration in the iPolar software. I've had it try to take the second image immediately after I move the RA axis instead of waiting for me to finish the motion. If the axis only rotates by a few degrees, the calibration will be less accurate than a motion through 90 degrees. A calibration error will also show up by attempting polar alignment with the iPolar in the iPolar software at 0 deg and then again at 90 deg. But the iPolar scope+SharpCap will be a valuable test. Hopefully the author gets a chance to try it out! Should only take a few minutes.
I'd like to add my experience to this saga......
The short answer is that the iPolar software (at least for some of us) simply does not work.
TLDR: DO NOT USE iPOLAR SOFTWARE. USE THE iPOLAR CAMERA WITH A DIFFERENT POLAR ALIGNMENT SOFTWARE (e.g. Sharpcap).
Before I begin I'll answer: Yes, I have downloaded the latest iPolar software. Yes, I have checked the firmware of all my related equipment and they are up to date.
Camera calibration:
I received a new CEM40 a few days ago. I have been trying to get the camera to calibrate per the (incomplete and incorrect) instructions in the manual. I have made 20+ attempts at camera location calibration. All attempts resulted in WILDLY differing camera location results. After each calibration attempt the software would place the red cross on my screen, but would always place it an odd location; sometimes the left edge, sometimes the right edge, sometimes the top edge, sometimes the bottom edge, sometimes in the right, bottom quarter, sometimes in the upper, left third. Despite repeated attempts at camera location calibration, I could never get a result that provided a camera alignment location that resulted in enough screen to use for alignment purposes.
I manually entered camera coordinates and then iterated different polar alignments hoping I would eventually converge at a camera location calibration value that I could rely on. I was never able to have success. When I manually entered calibration values I was able to get a polar alignment result. HOWEVER, after this polar alignment result, I could never get a successful one star or two star alignment on the mount. So, the GOTO feature would fail. I attribute this to the incorrect camera alignment values.
After two days I was frustrated, but felt like I would eventually converge on a solution using manually entered calibration values. Except........(begin the playback of ominous music)
iPolar Software Idiocy
The Polar software decided to get even more unstable last night.
I decided to do some more manual calibration, polar alignment iterations last night with an eye towards eventually converging on a calibration solution. I entered a new set of manual camera calibration values, got the red cross on my screen in a location that gave me enough room to try a polar alignment and then began the process..........For unknown reasons the iPolar software decided it wanted to do a "Rave". The red dot showing the position of the pole would appear and then immediately disappear and move all over the screen. It would appear very close to the red cross for 1 second, then it would disappear. Two seconds later it would reappear at the top edge of the screen for two seconds. Then it would disappear again, and be replaced by the maroon arrow telling me to lower the scope. Two seconds later the maroon arrow would shift positions and tell me to raise the scope, then it would disappear and the red dot would be back right next to the cross. Then it would disappear and the maroon arrow would be back pointing at the top right corner for two seconds and then disappear again; only to reappear pointing at the bottom left corner. Sigh......
I got rather irritated (meaning I went on a profanity laced tirade that would make the Father from "A Christmas Story" blush) and gave up for the evening.
Finally, a Success....
After calming down...... I spent $15 to upgrade to Sharpcap Pro. I went back out to the mount and decided to polar align with Sharpcap. I decided to use the built-in iPolar camera.
Using the iPolar camera and Sharpcap software I was able to get a very solid polar alignment in under 10 minutes. I was very careful and methodical, and it could have been done in less than half the time. The Sharpcap software was able to plate solve after two observations (one with the scope in Zero position, and one with the RA 90 degrees to the "east"). Sharpcap gave me very solid and stable correction instructions and I was able to polar align with less than 1 minute of error.
After polar aligning with Sharpcap I did a one star alignment with my mount. It required a bit of an adjustment and then the GOTO feature behaved beautifully.
So, the iPolar camera worked fine. I was able to get a very good polar alignment using the camera. The iPolar software is currently useless for my mount. It will not calibrate a camera position nor will it provide stable pole locations for an alignment.
If you're having camera calibration issues or alignment issues with iPolar I suggest abandoning the software. Use the iPolar camera and a more stable software solution (e.g. Sharpcap). If iPolar works for you, that's GREAT!!!!! If it doesn't then RUN AWAY!!!!!!
Out of curiosity, with iPolar, did you find that the stars (not just the +) were blinking in and out, more and less on each iteration?
I've found there is a sweet spot of gain + exposure. Too much, or too little, and the star detection is erratic, which either refuses to plate solve, or worse plate solves to wildly wrong results.
I find this out by trying to hurry dusk, and get polar alignment while the sky is too bright, and adjusting settings to try to et around it; so somewhat self inflicted.
If you are seeing erratic number of stars, before trying to align, adjust the gain and exposure to yield the most stable view of the stars - one where the number of stars stays the most consistent. Then try again.
That may of course have nothing to do with the trouble you (@dcm_guitar) had, but some of your description sounded familiar.
I find if I just leave the settings alone (once I found good ones) and wait for it to get decently dark, star count is mostly stable, and I get excellent alignment with no effort.
PS. That's not an apology for badly designed software -- good software should help the user be optimal, not require them to guess.