Star trails and Long exposures help

[Maho]

Hello everyone

I'm new to astrophotography and have noticed that when I take an exposure longer than 15-20 seconds, I start seeing star trails.

I live in a Bortle 8 site and would like to take longer exposures but have no idea how to a avoid the trailing issue.

Below pic is of Vega which was captured with a 20 sec sub and even here you can see elongated stars.

I have a Skywatcher 130PDS, Ioptron CEM25P and a Canon 550d.

What should I double check? Is it balancing? Collimation? Polar alignment? I usually do a 2 star alignment as I find it difficult to see lots of bright stars from my location? Should I try a 3 star alignment?

Any tips and advice would be greatly received.

 

[vlaiv]

If you don't guide - well, start guiding.

Mounts have imperfections in gears and that results in something called periodic error. Sometimes mount tracks a bit faster and sometimes a bit slower (think ellipse instead of perfect circle). It happens on a period of revolution of out of shape gear - hence the name periodic error.

You can try periodic error correction if your mount is capable of it (either in hand controller or if you connect it to computer), but that is limited in scope of what it can do. It can minimize the error - but not remove it completely in most cases.

Solution is to start guiding.

If you are already guiding - well, you are either not guiding good (wrong parameters) or there is differential flexure between guide and imaging scope. Former can be addressed by setting proper guide parameters while latter requires better connection between components.

[ollypenrice]

The star alignment routines are entirely unconnected with your tracking accuracy. They are only related to the precision with which the GoTo operates.

Accurate polar alignment will help but, as Vlaiv points out, you simply need to autoguide. The software, PHD2, is free and intuitive. You just need a guidescope and guide camera. I would just use an ST4 cable connection to connect guide camera to mount.

Olly

[AstroMuni]

32 minutes ago, Maho said:

Polar alignment? I usually do a 2 star alignment as I find it difficult to see lots of bright stars from my location? Should I try a 3 star alignment?

The polar alignment if accurately done should minimise tracking errors. As vlaiv has said, guiding will help for longer exposures. IMO you should be able to get atleast close to a minute without adding a guider. Are you controlling your mount via laptop? If so you can get it to platesolve and realign every few exposures. Thats what I do as I also dont have a guidescope.

Edited September 2, 2021 by AstroMuni

[adyj1]

Everything Vlaiv (and Olly) says is right, but I have to say that I'd expect to get better than 15-20 second exposures without guiding. As a beginner the step up to guiding can be a bit daunting (and normally results in a hike in complexity), so if you don't have the necessary guiding equipment to hand, I'd concentrate first on your polar alignment.

 

[vlaiv]

1 minute ago, AstroMuni said:

IMO you should be able to get atleast close to a minute without adding a guider.

That really depends on periodic error of the mount.

I've found that in most cases - Periodic error is causing larger drift than polar alignment.

 

[adyj1]

1 minute ago, vlaiv said:

I've found that in most cases - Periodic error is causing larger drift than polar alignment.

... that depends how bad your polar alignment is 😉

I completely accept it could be PE, but in my very limited experience PA has normally been the culprit for short(ish) exposure trailing.

 

[vlaiv]

8 minutes ago, adyj1 said:

... that depends how bad your polar alignment is 😉

I completely accept it could be PE, but in my very limited experience PA has normally been the culprit for short(ish) exposure trailing.

 

Here is small animation I made from one unguided session. This is Heq5:

Steady drift to the left is due to PA (in DEC axis), up down periodic motion is of course in RA due to PE.

You really need to mess up polar alignment in order to get trailing in short exposures like 20s. According to this:

http://celestialwonders.com/tools/driftRateCalc.html

if your PA is half a degree out (width of a moon in the sky),  you'll get two pixels of elongation in 20s exposure (at DEC 0°):

Canon550D with 4.3µm pixel size and 130PDS with 650mm of FL gives 1.36"/px

[adyj1]

1 minute ago, vlaiv said:

if your PA is half a degree out (width of a moon in the sky),  you'll get two pixels of elongation in 20s exposure (at DEC 0°):

Vlaiv, the points you have made make sense and I can't argue with the numbers (and as usual I've learned something from your posts) .

The irony is that the 'P' in CEM25P is supposed to mean 'guaranteed low periodic error of ~+/- 10 Arc seconds or less.' 😆

Ady

[adyj1]

10 minutes ago, vlaiv said:

Here is small animation I made from one unguided session.

That clip does seem to illustrate the point well. What length of time does the clip cover? (At first I thought you must have mounted it on a horse 🤣 )

Edited September 2, 2021 by adyj1
typo

[vlaiv]

5 minutes ago, adyj1 said:

Vlaiv, the points you have made make sense and I can't argue with the numbers (and as usual I've learned something from your posts) .

The irony is that the 'P' in CEM25P is supposed to mean 'guaranteed low periodic error of ~+/- 10 Arc seconds or less.' 😆

Ady

Level of periodic error is one side of story. Smoothness of it is another.

If whole 10" error is covered in said 20s - then it will create 7px+ trailing in that image. If we assume that worm period is say 8 minutes - that is about 500s, and that PE is uniform - in that period 10" is covered twice ("up and down"), so 10" for 250 seconds - that makes something like 0.04"/s of drift, or in 20s exposure - 0.8" or less than half pixel.

If that is true nature of the PE - well, one could use much longer exposure, but question is - is mount up to specs and is 10" P2P error condition met (SkyWatcher mounts often have 35"+ P2P error), and what is the shape of error - it does not need to be sine wave - it can be multiple ripples and still be less than 10" in amplitude.

6 minutes ago, adyj1 said:

That clip does seem to illustrate the point well. What lengtho of time does the clip cover? (At first I thought you must have mounted it on a horse 🤣 )

I think that it is at least hour. Each up down wobble (whole cycle) is about 10 minutes as worm period on Heq5 is 638s. I think that video has at least 4-5 cycles or so.

[Richard_]

I started out with an unguided Star Adventurer last December and was able to get 60s subs without star trailing. However, there were what appeared to be star trails in some of the subs but this is likely due to backlash in the gears. Once I started guiding, I was able to take 300s subs with ease and I found that I wasn't throwing subs away due to the supposed star trails.

Maho, I'd recommend double checking your polar alignment routine. Are you doing it correctly? What do you use to polar align: the built in polar scope or something like iOptron's iPolar? If you're using a polar scope, check to see if the reticle is calibrated correctly or not. When I rotated the polar scope of my Star Adventurer out of the box, the crosshair part of the reticle wasn't rotating around the central point which means it was out. After correcting this (i.e. adjusting the set screws), the crosshair would then rotate around the central point and my polar alignment was more accurate.

EDIT: Take my above response with a pinch of salt, I just realised what telescope you're using. Do the star trails get longer with longer exposure times (e.g. how does a 120s sub look like compared to the 20s ones)? If there's no noticeable shift in star trail length, perhaps it's coma you're seeing in your stars? Are you using a coma corrector with your telescope?

Example of coma

https://www.lonelyspeck.com/wp-content/uploads/2015/07/coma-comatic-aberration-star-example.jpg

Edited September 2, 2021 by Richard_
Reason included in post

[Seelive]

You can easily check whether it's the polar alignment or periodic error that's the likely cause of the trailing by taking a series of short exposures (say 10 to 30 sec) over a length of time equal to several rotations of the worm wheel (so for the the Star Adventurer a multiple of 8 minutes) then stack the resulting images WITHOUT applying any alignment.  Any polar alignment error will show as a slow but constant drift in a particular direction, periodic error will show as an oscillation either side of that drift.  The following image was taken with my old home-made tracking mount (starting on the left) with no period error correction applied and then midway though the series of exposures periodic error correction was applied to the motor drive (a magnified view of an image taken using a 100mm lens and a camera with 3.7um pixels):

Since the periodic error effectively describes a sinusoidal waveform, if you happen to expose during either of the two 'zero crossing' points of the sinusoid the trailing will be significantly greater than if you happen to expose during either of the peaks of the sinusoid.  Since the worm rotation of the Star Adventurer is 8 minutes, depending upon the imaging focal length and pixel size you could possibly get away with a 30 sec exposure at the crests with negligible visible trailing but the same exposure at the zero crossings would show some trailing?

Edited September 2, 2021 by Seelive

[AstroMuni]

3 hours ago, vlaiv said:

You really need to mess up polar alignment in order to get trailing in short exposures like 20s.

And hence my advice to the OP to check PA

I like your detailed calculation and explanation though. As always its very comprehensive 👍

[Maho]

Thank you all, really appreciate it.

Autoguiding/guiding is a bit daunting so I will have to read up a bit more on what guidescope I will need. Software side seems more daunting than the hardware, looks like I will be watching more YouTube videos!

My polar alignment routine is a bit rubbish to be honest. I can just about see Polaris from my light polluted area (London) and when I look through the inbuilt polar scope, it's even less visible, and feel that it is a bit of a guess work at times. That's something I have to work on. Tonight promises  to be clear skies, so will be setting up early. My mount does have PE correction, so will be re-reading the manual.

 

Good to rule out that the star alignment process has nothing to do with star trails. 

[adyj1]

37 minutes ago, Maho said:

My polar alignment routine is a bit rubbish to be honest.

Don't worry, It is one of the hardest things to get right when you're starting out.

Take things at your own pace and keep asking questions - even if you think they are daft... The easiest way to learn is have an experienced astronomer looking over your shoulder telling you where you are going wrong - the hardest way is to have to use forums to try to explain something that doesn't work, you don't understand, and seems very complicated 😆. 

Once you get it cracked, the universe awaits 😉  

[Richard_]

3 hours ago, Maho said:

Thank you all, really appreciate it.

Autoguiding/guiding is a bit daunting so I will have to read up a bit more on what guidescope I will need. Software side seems more daunting than the hardware, looks like I will be watching more YouTube videos!

My polar alignment routine is a bit rubbish to be honest. I can just about see Polaris from my light polluted area (London) and when I look through the inbuilt polar scope, it's even less visible, and feel that it is a bit of a guess work at times. That's something I have to work on. Tonight promises  to be clear skies, so will be setting up early. My mount does have PE correction, so will be re-reading the manual.

 

Good to rule out that the star alignment process has nothing to do with star trails. 

The others have provided some great information on guiding etc but for now, practice your polar alignment procedure and practice some more!

Back when I was starting out, it would take me about 30 minutes to polar align (and that's with clear Bortle 5 skies!). Now, I use the ASI Air Pro (other software will work just fine) and I'm polar aligned in about 5 minutes with even better accuracy than before. So practice definitely pays off (as does using software rather than Mk1 eyeball)! When you're comfortable getting to grips with the polar alignment routine, and you're getting no star trails on shorter subs, have a look into increasing your sub length and reading up on guiding.

Get the basics right and everything else falls into place

[Seelive]

4 hours ago, Maho said:

... My polar alignment routine is a bit rubbish to be honest. I can just about see Polaris from my light polluted area (London) and when I look through the inbuilt polar scope, it's even less visible, and feel that it is a bit of a guess work at times...

I have my mount permanently fixed to its tripod and have 2 East-West marks in my garden where the tripod legs are positioned each time. As my garden slopes quite considerably I then use a spirit level on the mount eyepiece storage plate to level the mount. Actually, it isn't truely level (it doesn't need to be) but it just means that every time I set up, Polaris will be well within the Polar Scope field of view (so I don't need to hunt for it). I always polar align before I can see Polaris with the naked eye - I've always found I can see Polaris through the Polar Scope before I can see it visually - and it takes no time to align Polaris to the correct place on the graticule.

[Maho]

By way of an update, I used this weekend to sort out my setting up, including taking extra care on polar alignment and balancing. I can now take exposures of around 1 minute without any trailing. Very little I can do about the light pollution though.😫

 

Thanks again for all the help and advice. 🤝