Took some shots of Orion the other day with my Pentax and Astrotracer. ISO 800, 50 sec, F2 with 35mm lens. When I started processing them in the Pentax Digital Camera Utility (DCU5) most of the brighter stars came out looking like golf balls! - not nice sparkly stars at all! I thought I had the focus pretty good, and all the stars were round, but golf balls? Strange.... any suggestions?
Hi Jimmy,
You're not giving us much to go on. As Rob said, post an example.
how did you focus?
did you turn off autofocus?
or could you have possibly moved the focus?
Barring something screwy with DCU processing, focussing is the most likely problem. It probably isn't astrotracer. I use astrotracer. You get short streaks towards the edge of frame because the function doesn't trace perfectly. But not golf balls.
Focussing
Turn off AF
Use live view
point the lens at a bright star, Sirius, Canopus or Alpha Centauri. Use live view on maximum magnification to view and focus the image. If the stars are not perfectly focussed, they won't be visible until they come into focus.
After focussing, do a quick test exposure 2s and make sure the bright stars are focussed.
Astrotracer
Makes sure you do the precision calibration carefully rotating the camera in all 3 axes.
Keep exposures short < 2 mins for a 50mm lens. Better to stack afew exposures than to over do one exposure.
I took the attached image with astrotracer, processed in Lightroom cc and Photoshop cc. It is a stack of four x 2 minute exposures, all taken at ISO3200 with a 50mm f5.6
Unless the Astrotracer can actually rotate the sensor (I assume not) it will not be able to avoid some trailing of stars in the corners due to field rotation. I would assume that the sensor can move in X and Y side to side and up and down to track the sky, but not by rotating it about its centre.
Unless the Astrotracer can actually rotate the sensor (I assume not) it will not be able to avoid some trailing of stars in the corners due to field rotation. I would assume that the sensor can move in X and Y side to side and up and down to track the sky, but not by rotating it about its centre.
The sensor does rotate but the sky does not rotate about the same axis unless you have the camera centered on the SCP. Were it simply X-Y, the stars would rotate about the centre with longest trails at the edges. The trails at the edges are longest but the trails radiate outward from centre not rotate about the centre.
The processor probably calculates how to minimize movement over the whole field. It doesn't replace a good AP set up. Probably more as a replacement for what a Polarie or SW Star Adventurer might be used for.
It is good for someone like Jimmy that is starting out or if you are traveling in the northern hemisphere and want to do some widefield work, or a quick grab while travelling by car in a remote setting.
At any rate it is very unlikely to be causing the problem Jimmy reported.
Thanks folks for your suggestions. I am actually trying right now to find out how to shrink my photos to a size I can send. The JPG photos are around 4MB. How can I make them smaller? Cropping?. I focus by setting up in the daytime focussing on a far away hill (probably 5 kilometers away) Then I get masking tape and stick down the lens at the best focus. My lenses are Samyang manual only so I have to set focus manually. I set the Pentax to MF mode. The Astrotracer seems to work really well, because there are (mostly) no star trails, and some shots I have taken at 120 sec exposure.
I will try and post a photo asap
Thanks folks for your suggestions. I am actually trying right now to find out how to shrink my photos to a size I can send. The JPG photos are around 4MB. How can I make them smaller? Cropping?. I focus by setting up in the daytime focussing on a far away hill (probably 5 kilometers away) Then I get masking tape and stick down the lens at the best focus. My lenses are Samyang manual only so I have to set focus manually. I set the Pentax to MF mode. The Astrotracer seems to work really well, because there are (mostly) no star trails, and some shots I have taken at 120 sec exposure.
I will try and post a photo asap
Focus changes with temperature. The method you're using is unlikely to give good results.
Generally to get the file size down and display the full image, we use a combination of reducing the pixel dimensions of the image and JPEG compression.
Look for something called image size in the menu to resize the image then save as will give you a compression number, usually 1-10 or 1-12 or a percentage.
On this occasion however to assist diagnosis, I suggest you try just cropping the image so that we can see the fully magnified golf balls.
You can focus without Bahtinov mask and live view, but it means taking a lot shots and adjusting focus on enlarged frames (I was using magnifier utility... these days I use Bahtinov, Live view and magnifier for screen on camera).
Thanks for that. Very interesting tests, Bojan! So it seems that using a fast lens (ie f1.4) and faster speed is not really any good! I might as well put on my regular DA* Pentax and take the shots at F5.6 or so and risk a longer exposure. The lower the f stop, it seems the more purple fringing around the stars!
Calculation.... 10 sec at f1.4 = 20 sec at f2 =40 sec at f4 =80 sec at f5.6 Sounds feasible! I willl try tonight
Yes focus and lens aberrations seem to be the problem.
I doubt you could use a Bahtinov mask with a 50mm lens though I've never tried. The technique I described in my earlier post works very well for short fast lenses and yes stop down a little. No faster than f2.8, maybe f4 or 5.6. I think Bojan's comment about taking lots of shots was probably made because he uses Canon cameras which have limited magnification in live view. So he has to take a test exposure and then examine it in preview mode to gain higher magnification.
You haven't said which model Pentax you are using but most have 10x up to 16x magnification in live view. On bright stars, mag -1 to 1, the focusing can be done live without taking test shots and focussing can be achieved very quickly. On fainter stars, mag 1 - mag 3, the stars are invisible when off focus and snap into focus. I prefer the bright star method. Much easier.
Thanks for that. Very interesting tests, Bojan! So it seems that using a fast lens (ie f1.4) and faster speed is not really any good! I might as well put on my regular DA* Pentax and take the shots at F5.6 or so and risk a longer exposure. The lower the f stop, it seems the more purple fringing around the stars!
Calculation.... 10 sec at f1.4 = 20 sec at f2 =40 sec at f4 =80 sec at f5.6 Sounds feasible! I willl try tonight
Fast lenses are good, but they still need to be stopped down (for most lenses). Generally a fast lens stopped down will outperform a slower lens at the same f-stop until you get up to like f8, as seen in the 50mm tests in the link below. so if you have a f2.0 lens and a f2.8 lens both stopped down to f4, the f2.0 lens will generally be sharper.
My Pentax is a ripper! It is a KP Latest new beaut. I will try the f1.4 stopped down to f5.6 tonight - IF the clouds/rain/wind/ obliges. Yes it seems like I was using the lens too much open, thinking to get rounder stars on short exp's . I am fiddling around with Pentax'es Digital Camera Utility, and it seems quite good. I guess that perserverence is the game I will try to be more careful with the live focus next time
Depending on how much you stop down the lens and whether it has straight vs rounded aperture blades, you may also introduce or make more obvious diffraction spikes around some of the brighter stars.
Thanks, guys! At last I have something resembling an astro photo .I actually did find that when I cranked the aperture to f4 and 5.6, I got some spikes on the brighter stars as you said! Here is M42 with some colour in it, and...also some nebulosity just below Alnitak - is this the Flame 2024 that my atlas says? this photo was 123mm focal length, f4, 60 sec, ISO 1600
any suggestions?
I will try to be more careful with the live focus next time
