I wonder if there are any Astrotracer users among the members of this group?

I just got the PENTAX GPS Unit O-GPS1, not only because of its Astrotracer function and certainly not as a repalcement of my equatorial mounts, but I do look forward to try out this neat little gadget.



Unfortunately, we are now in the period of very luminous nights - "white nights" - of Scandinavian summer and it won't be until late August that I may produce any "pretty pictures" - if any at all!

Meanwhile, do other memebers of this group have Astrotracer images, tips and tricks to share?

Apart from the morning of the Venus transit (I got 274 pictures of that) weather is extremely cloudy currently, but I did get some proof-proof-of-principle over a copuple of nights:

............
Arcturus, 100mm focal length, 30s exposures with and without Astrotracer (on two different nights)

............
Alcor-Mizar, details as above

Without knowing for sure, it seems fairly obviously to me that the strength of the O-GPS1, the incorporation of a magnetic sensor and thus the incorporation of an electronic compass, is also its main weakness:

• When you have the true azimuth and altitude and exact position and time, it is straightforward to convert to hourangle and declination and thus determine the proper motion of the sensor to counteract the celestial sphere's rotation. And the Astrotracer seems to have good algortithms for that. So far, I haven't noticed any tendency towards field rotation using a 100mm lens for my first tests.

• However, the magnetic sensor will have ever harder times to read the direction of the magnetic field, the higher the sensor points in the sky, and when it points towards Zenith the direction will be completely undertermined.

• I further wonder, if the camera's inclinometer will work all the way up to a perpendicualr position? We all know, that the display of inclination disappears long before then.

But knowing and understanding these limitations better, it should be possible to choose appropriate times of the year to photograph portions of the sky with decent results - or so I hope, at least........

B.R. / Steen G. B.

Finally some clear (although very bright) skies last night. So, I could pursue my goal of some more "systematic" tests. Again, my trgets were Arcturus (42 degrees high in the sky) and Alco-Mizar (60 degrees up), but this time, I used my old Tamron Adaptall-2 200mm f/3.5 lens (Model 04B) stopped down to f/5.6.

Examples shown are quite encouraging, I think, and about what I was hoping for with a 200mm FL. Now, I am eager to see if things will work even better, once I get out in the country side, away from my 4th floor balcony and all of its potential magnetic disturbances........?

Examples shown are all un-processed, non-resized crops of JPEGs shot at ISO 400.

B.R. / Steen G. B.

What do you think is happening with the progression of images? The first several are showing a consistent drift as if it were slightly misaligned. Since its based on GPS, the location and number of satellites may have an impact on the resolution. Not sure if this makes a difference, though.

I agree with your findings but I am not sure about the explanation.

You are right that the number of satellites and their positions change and on certain nights I have had problems with (frequent) intermittent losses of contact for periods of some hours.

However, as you have pointed out, the small trails are what one would expect with a slight misalignment (trails don't follow the stars' apparent rotation over the sky) and my prime suspect remains the compass. Perhaps more, systematic tests during my vacation - where I shall be on open land - will show?

You could try some bright stars near the meridian and equator if that's not too low on the horizon. At least that way you can identify which component of misaligment is giving the drift.

You are absolutely right. And that's exactly what I plan to do in my vacation. We are talking 34 degrees above the horizon, but here, and now from my balcony I would be facing Central Copenhagen.

Astrotracer - Tips & Tricks #1

"Meanwhile" here is a cheap and dirty trick - I shall not take offence if somebody whispers "Cheat!"


We take one of the images from before:



and do a little bit of manupulation:




Obviously, I have worked with levels and curves, as I usually have to do under my light polluted urban sky - but what
happened to the elongated stars? The answere in this case lies in Focus Magic and "Fix Motion Blur":



- not the ideal solution but still, perhaps worthwhile pursuing to give the finishing touch now and then?

Ah! - finally found this one (archieved in the wrong folder). It is taken at a compass reading of 182.5 degrees (at the utmost left in this 100% crop) and I believe the bright star is Zeta Ophicus with a declination of -10 degrees. (Everything here in this direction is totally invisible to the naked eye this time of the year from my place). It is a 100% non-resized crop, 750 pixels wide; It is the same 200mm lens and again a 30s exposure:



and the post-processed version (including removal of a hot pixel) of the above:


Hardly any drift here. Hmmm.....conclusion so far??? What is your opinion Smigol?

Can you tell me the date and time of the first Zeta Ophiuchs shot? Thanks.

Even so, the drift that I see is in RA, not DEC. Thus, I think that the error comes from not tracking enough or tracking too fast. Without a frame of reference of what the starting point is it's hard to tell. If there were two shots back to back you might be able to see the drift rate.

Thus, the only solution for this one would be to either limit your subframes until the drift is no longer objectionable or reduce the focal length of your lens or maybe check the charge level of the batteries?

Unguided the star trails would be 90 pixels long, so this is an improvement!

Date: 19 June 2012 (the day before yesterday)
Time: 23:19:49 Central European Standard Time = UTC 22:19:49 (as obtained from the GPS satellites)
Shooting Position: N55:46 E12:30

OK, so it was near the meridian at the time. Drift is in RA so my thoughts still stand. I think that the tracer just has a hard time keeping up with the required movement for the 200mm. When you tried the shorter lens, did you see better tracking? Of course, pointing to different parts of the sky will make a difference. If you look at this site: Useful Formulae you will see a quick javascript formula that will show the expected length of the star trails. Maybe we can figure out the extent of the "anti trailing" capability by examining the sub lengths.

Since I received the O-GPS1 on 7 June I have only had three nights with reasonably clear skies. Right now, it is raining cats and dogs, but during summer vaction in late July-begin August my plan is to make test shoots in "all" directions at "all" declinations from 0 - 90 with some selected primes in the 24 - 350mm range.

Preferrably, I should repeat such tests a number of times to check for apparatus reproducibility.

Then, we may have (I hope) some solid data to figure out the "anti trailing" capability.

Actually: It's nice to have such a project for northern summer time, where I usually don't have the chance to take "nice astrophotograps"!

B.R. / Steen G. B.

Just a thought, does this lens report its length to the camera or is it a manual focus lens and you have tell it the focal length? Maybe it's as simple as reporting a slightly longer/shorter length.

You're right: There is a potential source of error here.

My 200mm lens is a manual one and I have to tell the FL to the camera. A 10% error would reult in a trail of some 9 pixels for stars near the equator. Leaving Astrotacer off should - in principle - allow me to calculate the actual FL from the lenght of the trail:

s = 2*F*tang(0.0021*t)*cos(d)

I know that for the K10D that the debug menu allows you to tweak the effect of the Shake Reduction. Not sure if there would be a similar feature on the K5. If so, then that might be a way to get it tweaked, although you'd have to confirm if this setting is appropriate for all lenses.