Seestar S50 drifting

bakosmi

EQ Mode. I have polar aligned via Seestar APP, then confirmed PA via NINA's TPPA and i was within 1" of PA error.
Problem is no matter what I do, my S50 is drifting back and forth in RA constantly, restricting me to take only 10s subs in EQ mode, anything more is elongated.

Had anyone same issue and resolved it somehow ?

I have recalibrated compass and level, my PA is good. I am desperate. Thank you!

bakosmi

did another round of testing - 0.0 PA via app, doing 10s subs is fine , but doing 20s has more rejected than stacked... i am shooting pleiades at the moment.

seems like seestar s50 can only do 10s usable subs... :/

Lilo

bakosmi
It can 30s wo problems.
Check that it can rotate freely and the base doesn't touch something. Sometimes rubber feet could be the culprit.

BiF

Elsewhere, I complained that after the update, I can no longer do 60- and 30-second exposures—90% of subs rejected (before the update, 5-10% of subs rejected). There have been no updates since then.
Tonight (Central Europe, very strong wind!) – 30 sec, 1 500 subs, 6 subs rejected!
I don't understand...

3AC9bx99

Nothing unusual with a high frame rejection rate with the Seestar in EQ mode. The tracking is not always precise and seems to vary from unit to unit. I also get high frame rejection rates at longer exposure times and often 10s is the best I can use. Being in the Northern hemisphere it is worse for objects in the South as you get the longest star trails occurring near the celestial equator so grater frame rejection rates. For objects closest to the celestial North, star trails are less and the Seestar rejects fewer frames. I can use 30s exposures for Northern objects quite successfully with a keep rate of 80% or more. Using NINA's polar alignment helps but for me ZWO need to improve the quality of the tracking mechanism.

YF5I3f19

It seems to be kind of luck of the draw, someone can take frames 30 or 60 seconds long, others (like you and me) are severely limited to 10. In my case, even with 10sec frames I still have some of them with elongated stars, which I have to discard. The tracking is probably the aspect of S50 which is most affected by cheap components, showing a huge periodic error (the constant drift back and forth in RA) which, at present, is unfortunately something we necessarily have to live with.

bakosmi

everything is sturdy and no wind - i got Skywatcher Skyadventurer tripod + eq wedge and also have steel plate to weight the tripod down. Silicone cable which is very soft even in freeze, nothing tangled. Setup is ok. What is not ok is S50 constantly drifting back and forth In RA. I am quite desperate as I believed i can do 20-30 subs, not having thousands of files instead.

YF5I3f19

I can understand your disappointment, it was the same for me, but I believe there's not much we can do to improve this situation because it's a hardware thing. Me too, I don't like at all having to deal with thousands of files, but I'm afraid that's the way it is. On the other optimistic hand 🙂, we still have the possibility to produce very nice images with our S50, once we get accustomed to this new procedure.

bakosmi

i don't know if i can accustom to something that should be working and it's not as should.
i wll try some lower alt targets tonight.

curtismacc

As 3AC9bx99 says, drift is more pronounced closest to the celestial equator as the drift goes as the cosine of the object's declination. No matter how hard we try, we never get a perfect PA. The problem with posts on this and other forums on the frame rejection rates is that folks rarely tell which objects they are imaging so it becomes impossible to tell what part of the difference in rejection rates we get versus someone's posting is due to declination drift vs other mechanical issues including differences in one Seestar unit to another. So, I would suggest anyone posting on this topic to let us know what rejection rates you get for M81 and M42, at least for this time of the year. For M81 at a declination of 69 degrees one would get the lowest frame rejection rates compared to M42 at a declination of -5 degrees. If you don't see a clear difference between these two objects, declination drift is not your problem, which also means that PA error is not your problem. For example I am getting frame rejection rates at 30s exposures over 3 hours of 0 to 2.4% over 4 nights for M81 and 47.5 to 51% for M42. I am using NINA's 3 PT PA process to get within 3' total error combined for both axes, which is 0.05 degrees.

bakosmi

i had tracking issues with m42, m45, and ngc 2244 (rosette) - i think all targets quite high these days.
problem with nina PA is, that the thing is drifting while polar align as well - so if you finish polar alignment within 1", watch it, it will go away. so you did not even polar aligned properly via nina. that's the fun. with normal cheap rig, you get PA, you rerun it, you should be slightly off but not so much as with S50.
also i think polar alignment should not be done exactly against polaris area.
with my old righ i had best PA results aligning directly to west where it really moves.

curtismacc

bakosmi " i had tracking issues with m42, m45, and ngc 2244 (rosette) - i think all targets quite high these days." This is a common misconception. For drift caused by PA error, It is not the altitude that matters it is the objects declination. M4 has a declination of ~ -5.3deg, M45 24 deg and the Rosette ₅ deg, all positioned close to the celestial equator so you can expect higher drift than for something like IC5146 at 47deg declination and definite much more drift than M51, M81 and M82.

You are correct, it is best to PA away from Polaris which provides more "leverage" in the mathematical fit as you move further away. It is tied to the fact that the drift increases the further from the celestial pole that you point. The problem is that, as you say, drift occurs while doing the PA adjustments, both with NINA and with the Seestar App. So the more "leverage" also means more drift in the time it takes to do the PA. Applications like T-Point which can use hundreds of points and take more than an hour must account for this drift because one can get an excellent PA with it. I use it with my Software Bisque mounts and routinely use over 100 points.

3AC9bx99

One work around that may be worth experimenting with is to use the continuous capture option. Switch off Execute Enhancement after GoTo and select Continuous Capture in the live screen. Enter the number of frames you expect to capture during your session and let the Seestar store every frame during the night. If you have Auto shut Down toggled on the Seestar will switch itself off after 15 mins of inactivity at the end. You can use Deep Sky Stack feature to stack the frames or better still use an external stacking program such as in Siril. Whether you get a higher stacking rate this way compared to Live stacking is the experiment. I haven't had successive clear nights to try this on the same object. Something I intend to try out. Be interesting to see if anyone has more success with Continuous Capture.

3AC9bx99

I may be wrong but I think the Seestar will still automatically still stack all the frames with Continuous Capture so no need to use Deep Sky Stack or external stacking program. External program may give more control on what to keep.

bakosmi

I did pa to zero yesterday only using seestar app and reentered the pa to confirm with a new set of images.

I proceed with m31 which I could do 20s at approx. 30% loss rate. Was quite ok.

Then I switched to horsehead when it got to 35 alt. And it was terrible. Not a one good sub. I redid the goto but no help. I ended up again taking 10s shots.

Very very bad experience so far.

I will install today the last update and will see.

YF5I3f19

I think I have some solid, definitive numbers now regarding my unit's tracking.
I shot a number of 5s frames, very close to 0° declination to determine my unit's periodic error.
I managed to reach perfect polar alignment, considering that over a 15-minute period I got 0 pixel drift in declination (in fact I see only horizontal movement of the framed stars).
Under these conditions I measured a periodic error of 607 seconds, spanning 85 pixels which, at S50 2.4"/px pixel scale, means a whopping 204 arcseconds peek to peek.
Which means that, even shooting at 10 seconds, I still have to discard quite a few frames and this, by the way, is confirmed by what I get in practice.

curtismacc

YF5I3f19 What did you use to do the PA and how did you measure the drift? Nice data.

YF5I3f19

I used NINA for PA.
To measure the drift, I opened the sequence in Siril, picked a small reference star close to the image center and monitored its x coordinate; after one cycle, lasted 607 seconds, the reference star returned to its initial position, I read the two farthest apart x coordinates during the cycle and that was it, the difference was 85 pixels.

bakosmi

I did new testing yesterday. I polar aligned repeatedly going out of PA and measuring three points again with overall 0.1 0.0 PA in seestar app. I used i think point 2 on South, knowing i will be shooting targets in south.

I have tried 30s exposures - all of them were blurred in RA, S50 was hunting RA+-, then i tried 20s at one target (M31), reject rate around 30% (acceptable comparing to overhead when shooting 10s - S50 won't manage to shoot and save them in time, so often it stucks at 0s time, shorting the real time vs capture time).
on another target - triangulum the reject rate at 30s was 100%
so no change at all.

Result:
10s result - enlongated stars on every other frame, perfect only one in 5 (HRF range 1.5 - 2.5) - looks very periodic, consistent with observed RA movements.

20s exposure results are highly target dependent.
30s is unusable completely.
60s is probably there only to frustrate me even more.

These blinks were taken via stargazing live mode, no mosaic.

horsehead.mp4

5MB

m31.mp4

3MB

note: in the videos there are only subs that were kept by seestar app - i was checking the enlongation / drifting and HRF in NINA.

bakosmi

I checked my other subs taken with NINA only:

nina.mp4

1MB

nina2.mp4

2MB

and it seems obvious seestar app does some kind of dithering, while nina does not,
but the RA moving back and forth is present in both.

i wonder what other users have set here or if it has any effect on this issue: