[peterh337]

I have put them at this URL
http://peter-ftp.co.uk/k1astro/IMGP1857.JPG
and you can download them by substituting 1858 1859 1860 1861 in there - a total of five big jpegs. The URL is mostly case sensitive.

Many thanks for any pointers.

I have re-done the "precise calibration" but had already done it previously.

Regarding the camera heading in the EXIF



that looks about right, within a few degrees at most, so the compass was working. But also the shorter exposures are very sharp.

[jbinpg]

@photoptimist - Stars move fastest as seen from Earth along the celestial equator, not along the ecliptic.

[Stone G.]

I have put them at this URL http://peter-ftp.co.uk/k1astro/IMGP1857.JPGand you can download them by substituting 1858 1859 1860 1861 in there - a total of five big jpegs. The URL is mostly case sensitive.

So, your image with the most trailing is #1861.


1.I first of all note, that the trailing is for stars all over the frame and thus, we are not talking wide angle field rotation issue.


2. From Phil Harvey's ExifTool I get the following information:
Geographical position:50 deg. 53 min 36s North and 0 deg 12 min 25 s West
GPS date and time taken: 2017-01-12 at 23:20:29 hrs.
Azimuth = 211.94 degrees
Pitch angle (Altitude) = 24.5 degrees
Exposure time = 117.8 seconds.


3. Now, you have conveniently placed Sirius very close to the center of your image so entering date, time and position in Cartes du Ciel (Sky Charts) I can find the actual parameters for Sirius the moment you took your picture. This is the information Cartes du Ciel gives:


Star
HR 2491 HD 48915
Flamsteed Number: 9
Bayer Letter: Alpha
Constellation: Canis Major


Visual Magnitude: -1.46
Color Index: 0.00
Spectral Class: A1Vm
Annual Proper Motion: -0.553 -1.205


SIRIUS; Canicula; Dog Star; Aschere


J2000 RA: 6h45m08.90s DE:-16°42'58.0"
Date RA: 6h45m54.57s DE:-16°44'05.4"
peterh337 2017-1-12 23h20m ( TU + 0h00m )
Sideral Time : 6h50m
Hour Angle : 0h04m
Azimuth :+181°07'
Altitude :+22°24'


Rise : 18h43m Azimuth:+116°23'
Culmination : 23h20m
Set : 3h57m Azimuth:+243°37'


In fact, you have captured Sirius the very moment (almost) it culminated and thus, was due south on that night. So, unless I have made some terrible blunder, your altitude is spot on (Sirius may not be exactly in the center and there is also the atmospheric refraction at low altitudes to consider) but your Azimuth / Compass Heading is some 30 degrees off - and that does explain the trailing that you see across the entire frame over an exposure time of 2 minutes.


As I said in my first post, disturbances or very localised anomalies in Earth's. Magnetic field may spoil your precise calibration, but there may also be other explanations such as a sudden shift in the satellites that your GPS has contact with which wouldrequire a re-calibration.

[Stone G.]

Here then, is what Geosetter tells about where you pointed your camera according to your Exif-Data. How does that match your own recollection?

Attached Images

 

[peterh337]

Yes, that's correct.

This is the anger of posting EXIF on social media and why facebook etc strips it off

But I knew that in this case.

[Stone G.]

But then your compass heading is indeed 30 degrees off what it should have been(?)

[peterh337]

I have plotted the camera heading used in the referenced image (because I know where I was in the house) and it looks like 190 degrees



So the heading may be wrong but in a different direction.

The heading in the pic above is nowhere near the 211 degrees in the EXIF.

I have just taken another pic (in as close to the previous direction as I can, by reference to where I was standing and the two conifers) which I put at the same URL as "1866", having calibrated the compass outdoors, but that shows another wrong compass heading:



There is no manual compass calibration on the K1, is there, and no way to view the compass reading without taking a photo and looking at the EXIF.

---

Last edited by peterh337; 01-14-2017 at 07:33 AM.

[phoebus]

@peterh337 you can see the compass heading on the "Electronic Compass" info screen and calibrate the compass manually by pressing the +/- button when in astrotracer mode. When i've tried astrotracer, it has worked over 2 minutes with a 105mm lens. I calibrated it with the tripod mount attached, just before taking the photos:

[peterh337]

Thanks

Without any new adjustments, the camera heading displayed in the info screen is within a couple of degrees of a magnetic compass located about 2ft under the camera.

This is another pic, correctly showing a 180 degree heading, taken from a very similar place in the house



However, the mystery deepens.

Measuring the correct magnetic heading (I am a pilot so know true v. mag) for that night shot, where the EXIF shows 181 degrees, with a compass, shows 180 as close as I can read it. I did that about 5m away from the house.

But measuring the magnetic heading with the compass where the tripod was in that shot, 1m away from the house, produces a different heading, and this photo shows it



So the issue here seems to be a combination of things:

1) Close to the house, both compasses (the real one and the camera's) are about 20 degrees out (under-reading) but they agree.

2) The camera was probably not properly calibrated last night.

Not sure this squares up with Stone G's analysis which suggests the heading error was in the opposite direction.

But I need to try another star shot well away from the house.

I also found this
http://www.ricoh-imaging.co.jp/english/products/o-gps1/calibration.html

Finally, referring back to the earlier images of the star sky, I can see how the sensor shift cannot compensate fully if the sensor can rotate only about its centre. If the sensor can rotate about a centre point which is outside its area (which should be possible if the suspension system was implemented thus) then one should get all the stars sharp even if the camera is not pointing at the centre point around which the stars rotate.

[photoptimist]

Finally, referring back to the earlier images of the star sky, I can see how the sensor shift cannot compensate fully if the sensor can rotate only about its centre. If the sensor can rotate about a centre point which is outside its area (which should be possible if the suspension system was implemented thus) then one should get all the stars sharp even if the camera is not pointing at the centre point around which the stars rotate.

The sensor is NOT limited to just rotating about the center. It can do any combination of rotation and linear motion within the overall mechanical limits of the SR mechanism.

It's easiest to imagine the sensor as a 36 x 24 mm rectangle floating in a 39 x 27 mm rectangular hole -- it can freely shift and rotate within the limits of the corners of the sensor bumping the edges of the rectangular hole. Obviously the amount of required linear motion places constraints on the possible amount of rotation and the amount of required rotation places limits on the possible amounts of rotation.

[Simen1]

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

The sensor is able to move 1,5 mm in all directions (less when rotated too). On cameras with 4,9 micron pixels (K-1, K-5 series ..) the 1,5mm means approximately 300 pixels in all directions.

[peterh337]

Many thanks for the clear descriptions.

In that case I don't understand why the astro mode would be able to compensate for some stars and not others.

[photoptimist]

Many thanks for the clear descriptions.

In that case I don't understand why the astro mode would be able to compensate for some stars and not others.

It's a fundamental mathematical issue with the geometric differences between the rotation of the celestial sphere and the imaging of motion by a rectilinear lens. If the stars were pinned to giant perfectly flat sheet of paper, astrotracer would have no problems at any focal length or any predictable motion of the flat sheet of paper. But it is as if the stars are pinned to a rotating sphere and that means they appear to move a different speeds in the center and edge when viewed through the lens.

At some level this problem occurs for the same reason a that flat map can't accurately represent a spherical globe. A flat map does an OK job for small portions of the globe (equivalent astrotracer with a long focal length) but it does a increasingly worse job as the map attempts to cover greater angles of the sphere. The map maker (the lens) mush warping the curved edges of the sphere to conform to a flat map and that means stretching them.

The stars move at an angular rate of about 72 microradians per second on the celestial equator (thanks, @jbinpg for pointing out the difference between the ecliptic and celestial equator). On the K-1 with a 24 mm lens this angular motion is about 1 pixel every 3 seconds in the center. But due to stretching of the spherical view by the rectilinear lens, the stars and their 72 microradians per second motions are stretched at the edges and will move about 1 pixel every 2 seconds. Astrotracer can't make the center of the sensor move at 1 pixel per 3 seconds while making both the left and right edges move at 1 pixel per 2 seconds.

[Gimbal]

Or put another way, wide rectilinear lenses distort the image to keep lines straight. So objects at the extreme corners are enlarged compared to objects in the center. I'm sure you have seen that before.

This also means that stars moving across the view will appear to move faster at the extreme corners, which of course the moving sensor can't mimic.

[clackers]

Many thanks for the clear descriptions.

In that case I don't understand why the astro mode would be able to compensate for some stars and not others.

Don't use a wide angle lens and expect the same results others get.

The stars themselves move differently depending on if they're at the edges or not.

https://www.pentaxforums.com/reviews/pentax-k-1-review/astrotracer.html