[peterh337]

I did some shots last night. It works great, but after about 40 seconds the stars move. So that seems to be the sensor movement limit. I am near 50N 0W.

[Simen1]

Actually, that is your tracking limit. Do you use the Astrotracer, other tracking or no tracking?
What is the focal length and sensor size?

[peterh337]

Astro = ON.
The lens was the 24-70 on full wide. This was an image which worked fine.

[Stone G.]

I did some shots last night. It works great, but after about 40 seconds the stars move. So that seems to be the sensor movement limit. I am near 50N 0W.

There is an in-camera physical/mechanical constraint on how long you may track. That is dictated by how much the sensor is actually able to move, and the theoretical upper limit in tracing time then depends upon the focal length of your lens and the declination/altitude of the center point of your image, (: stars near the celestial equator move faster than stars nearer to the poles).


HOWEVER, in prapractise the accuracy of the Astrotracer is very much dependent upon how precise a calibration of your electronic compass that you can actually achive on your specific, geographical location. This is not trivial, because local magnetic anomalies may interfere severely with Earth's magnetic field and thus, confuse Astotracer's compass.


Try different locations, different altitudes and different directions and you may find quite large variations in maximum acceptable tracing times with one and the same lens.

[peterh337]

Interesting; thanks.

I know the sensor can move only so much, and it won't be all that many pixels.

The compass has been calibrated and is accurate. The camera was also outdoors so had a GPS fix, otherwise even with the above pic (about 20 secs, ISO1600) the stars would not be sharp.

It's interesting that a minute or so may indeed be the limit for this function. It is still very good of course... and there are other solutions for much longer exposures e.g.
iOptron SkyTracker PRO DSLR Camera Mount (3322) - Wex Photographic

The point is that everything below 1 minute was sharp, and 1 min plus all showed very visible elongation. So it looks like the compass and GPS were spot on.

[CarlJF]

Stone G. has very good points. However, even then, you should be able to get better than 20 sec. with a 24mm focal length. I've done 4 minutes exposures with a wide angle lens and 20 sec. is about what I got at 300mm...

Are you sure you've done the "precise calibration" in the astrotracer menu and not the standard one in the GPS menu ? It happens to me a few times to select the wrong calibration in the menu...

Also, with a wide angle lens, some star trails could be visible the farther they are from the center of the picture. Unless you're pointing at the celestial center, the stars don't move all at the same speed and the astrotracer compensates only for those in the center. For example, look at the 4 minutes exposure in this thread. It's obvious that the astrotracer worked fine since the center is sharp, but there are visible star trails at the edges. This is an artefact needing to be considered when using astrotracer with a wide angle lens.

[peterh337]

I reckon you have it there, Carl, because the trails I saw were in Orion which you can see is in the top RH corner. I didn't bother looking elsewhere.

The key point is that the trails were near-vertical, when actually (I am no astronomer!) Orion moves from left to right when facing south, where we are.

So the camera compensates along just one straight axis?

[Stone G.]

No, the sensor can move left-right, up-down and rotate along the optical axis. What it cannot do is to tilt in and out of the plane perpendicular to the optical axis. Hence, very wide angle lenses may track stars near the centre properly, but stars images nearer to the edges will become stretched the further away they are from the centre and the longer the exposure time is.

This is also w hat CarlJF stated with different words in his post above.

---------- Post added 01-13-17 at 03:51 PM ----------

My experiences (with K-5 and K-3) is that the best lenses suitable for Astrotracer applications are in the 24mm to 200mm FL range. That said, I am no pixel peeper and I happily use my Samyang 14mm f/2.8 lens with Astrotracer also.

[peterh337]

I was aware the sensor can move as described but though that perhaps the astro mode is limited to linear motion only.

The inability to tilt should not affect this because the depth of field (at the sensor) should be a distant second order effect only. But then as I say I am no astronomer

The motion of the stars (not planets) is like this:



and a sensor which can combine linear and angular motion should cope with that.

I am going to do some tests to establish the exposure time at which the trails start to appear.

Can you upload an image that did not work? I am doing 4-5 minutes @ 24mm without any problem. Did you calibrate? Is the back LCD in its home position?

This one 63mm and 120 seconds.

[Stone G.]

The inability to tilt should not affect this because the depth of field (at the sensor) should be a distant second order effect only. But then as I say I am no astronomer

It is not a question of DoF; it is a question of how images of stars on the 3-D celestial sphere moves on the 2-D sensor, when the optical axis is constantly pointing in the one and same direction (relative to Earth) and the sensor is bound to have a position perpendicular to the optical axis, whatever movements else the sensor may be capable of.


You are not an astronomer, but imagine a camera sitting on a properly aligned motor-driven equatorial mount and you take picture of Orion just as it culminates. Lens will be pointing due south as will the mount's polar axis. A couple of hours later, you take a 'perfect' picture one again, but now the camera is pointing lower towards the horizon, it has rotated along the mount's polar axis and the optical axis is no longer pointing due south but south-southwest. And the sensor has of course moved together with the optical axis.


This 3-D movement cannot be fully mimicked by Astrotracer.

[photoptimist]

Vertical trails for stars that were moving horizontally? Hmmm... That sounds more like a calibration issue.

The SR system on the K-1 has a range of motion of ±1.5mm which is over 300 pixels. With a 24 mm lens, the fastest stars (those on the ecliptic) only move 1 pixel every 3 seconds in the center of the image. In theory, the SR system could correct for at least 900 seconds of motion of stars in the center (and maybe 1800 seconds if Astrotracer starts the exposure with the sensor intentionally pushed to one side).

But!! Due to the differences between the equi-angular motion of the stars and the rectilinear projection of the lens, these ecliptic stars will move faster at the edges of the sensor frame (about 1 pixel every 2 seconds if my math is right). If Astrotracer perfectly corrects the motion of the central stars, I'd expect the edge stars to move about 1 pixel every 6 seconds. If Astrotracer tried to remove the average motion to minimize average trailing across the frame it could probably get the residual trailing down to 1 pixel every 9 to 12 seconds.

But here's the bigger issue: the uncorrected trailing of edge stars should be largely horizontal along the ecliptic maybe with some radial trailing for stars out in the corners (as seen in that stellar image by CarlJF). I would not expect any vertical trailing unless the camera's calibration was off.

Note that the calibration is very sensitive to the magnetic environment. The K-1 manual recommends redoing this calibration if you change batteries, change lenses, or even if you change the angle of the tilt screen. It's also possible that any steel in a tripod head could affect the calibration, too.

[Stone G.]

Well, if the OP has the camera mounted parallel to the horizon rather than parallel to the celestial equator, then a compass calibration error will result in an up-down component in the trailing, (much as you have it when you do polar alignment of an equatorial mount using the classic Star Drift Method). But that would then apply to stars across the entire frame. It woul be good, if we could see one of the OP's star trailing images.

[photoptimist]

The camera's accelerometers measure the camera's pitch and roll pointing direction relative to vertical -- those should be pretty accurate unless they've been physically knocked out of alignment. It's the magnetic sensor's measurement of compass direction that seems to be the most susceptible to problems.

You are right that it would be good to see the OP's pattern star trailing.

[CarlJF]

and a sensor which can combine linear and angular motion should cope with that.

I am going to do some tests to establish the exposure time at which the trails start to appear.

Sure, but it always depnds on the way the camera is pointed. When pointed at, or near the celestial center, as in your example, the tracking will be much more efficient. But if you look at the diagrams provided by Nass and at the DCShooter's picture in this discussion, you will see that with a wide angle lens there are also plenty of situations where it's physically impossible to correct for each an every star movements by moving the sensor. Some would need an almost linear correction, some would need a clockwise rotation, and others would need a counterclockwise rotation. Obviously, the astrotracer can't do all these movements at the same time and the choice has been made to correct for what is in the center, which kind of make sense. And this is not even considering the distorsion of the lens which can be significant on the edges of a wide angle lens, even good ones, that astrotracer can't do nothing about.

This is much less a problem with longer FL since the covered angular area of the sky will be less and the stars will thus show a more uniform movement pattern. Which can be more easily corrected by astrotracer than the complex ones often observed with WA.