Hi

I've been learning both patience with the unco-operative weather (gales and rain) and technique with night-time timelapse (it seems a lot steeper learning curve than I ever expected after getting some nice results with the northern lights...) over the last few months, while trying to shoot star-lapses over a mountain.

But have just hit a new (and weird) issue. All the advice I can find suggests that an exposure of less than 20secs should eliminate the rotational effect of the earth's movement when shooting the stars for a night-time timelapse sequence.

Yet in a couple of my early attempts, there is something weird going on - of which half appears to be a rotational effect of the stars motion. Despite the exposure being only 13s... (Lens: Sigma 20mm f/1.8 lens at f/1.8mm, Camera: Canon 5mk2 - ISO1600)

Small version of the whole image ():
   

Crop of 1395 pixels from the Top R corner:
   

Crop of 1395 pixels from the Top L corner:
   

Crop of 1395 pixels from the centre of the image:
   
(which seems much less affected by both issues)

This effect is more pronounced on this setup than on the other lens and camera I've been using - Sigma 17-50 f/2.8 on Canon 650D, but that effect is still there.

And on some tests I did yesterday it seems some element of rotation is present even at 4s exposure...
   
(f2/8, 4s, ISO800, very small crop of an area near the top left of the image that shows the effect)

Although it also seems that f/8 at ISO3200 for 4s shows very little movement... certainly the best results (of what I've seen of the test shots so far) against a wider aperture, longer shutter speed or lower ISO. (Sorry, don't seem to be able to add any more attachments here to show this)

The second half of the issue appears to be a short bi-directional line of bright light from the stars at 90degrees to the angle of rotation - is this to be expected or is something even weirder going on here? (best seen in 1st or 2nd attachment - crops of Top L or Top R of image)

Any advice much appreciated, as this is completely baffling me and I'm in serious danger of letting someone down who needs the footage if I can't find a solution to what is causing the dashes...

Those are lens errors (bat wings)... quite normal with a lot of lenses - more or less depending on the lens - more at the corners, more if you shoot with faster apertures - but most of the times more or less present when shooting stars...

Many thanks for the quick reply Gunther.

Just to clarify - are both issues a lens error, it just looked like one was rotational in effect (and matches up with the rotation of the stars in a starstacked image I made from the whole TL)?

Assume there's nothing can be done to prevent this in future TLs other than change lenses?

Or shoot at f/8 with a very high ISO? (3200 or more, probably more like 6400 - very dark skies!)

This is to be expected when you shoot with a tripod or alt-az tracking mount. Unless you are shooting with an Equatorial mount, field rotation will occur. This will be more pronounced towards the edges.

I know this subject from my other hobby, which is long exposure deep sky objects photography.

Google "field rotation" to learn more on the subject, for example, read this entry.

Best wishes

Sedat

Thanks Bilgebay. That's a useful (if complex!) article...

And lead me to all sorts of other astrophotography articles, which told me that "bat wings" are "coma" (apparently pronounced "comma") lens aberrations, that they affect all wide angle lenses used wide open to some extent, and that contrary to everything I've read so far, even for night time timelapse virtually all lenses DO need to be stopped down a stop or two to reduce this to acceptable levels, and that is also where the "f/2.8 or better" rule of thumb comes from, so its still fairly wide aperture when its stopped down a bit..

But it also looks like it does have a small amount of rotation despite shooting at what should have been a safe exposure length...

So looks like I need a different lens that handles coma better, shoot slightly stopped down, use an even higher ISO (will 6400 be silly?) and even less shutter speed.

(Although I knew that N-S vs E-W alignment made a difference to the max exposure time for a single image not to show rotation, that article makes it sound like it's a lot more of a difference than I'd thought... Even if I can't yet get my head around the formula for how to calculate the max safe exposure for either camera (max sensor size 5184x3456)...

Incidently my setup is fixed, although I'd love to use a slider, I can't use an equatorial mount as it would do very strange things to the horizon - I need to include the mountains within the composition, not just the stars!)

This post was last modified: 2014-03-04, 02:17 by bilgebay.

Hi MW,

Yes, stopping lenses down does help greatly with comma and color aberrations as well as vignetting. Here is a good list of lenses that can be used for shooting stars and their suggested f-stops for good to excellent results. This website is also very good for seeing many lenses' capabilities before investing a lot of money for them. You can see a review of your lens here

f/4 seems to be a good aperture for this lens. You should get great results from it.

As for your max exposure calculations, this site can be of help.

Best

Sedat

Hi BB,

Thanks for that - but could you repost the link to the list of good lenses and suggested f-stops. Doesn't seem to work for me?

Sorry... updated the link! Please let me know if you cannot see the lenses again.

Thanks, that worked.

Interesting list, several I'd not have expected (relatively slow apertures).

Currently trying my 20mm out at f/4 (& much closer to home). Fingers crossed...

Good luck! Keep us posted.

Sedat