SharpCap Forums

Hi folks,

it's been a while since I have posted details on a new or improved feature here - part of that has been due to lots of bug fixing going on, also there have been a lot of small tweaks across different areas of the software that didn't really merit a post here. Anyway, this week's update - 4.1.12281 - has some improvements to feature tracking that are worth writing about...
It's probably easiest to start with the changes to the options available in the Feature Tracking Window as shown above. You might note that there are two new options available (and a bit of a re-arrangement to fit things in nicely). The new options are

* Having the option to use 'Phase Offset' detection to track the image as well as the traditional 'Image Features' and 'Center of Mass'. Phase Offset detection using the same Fourier based algorithms used for multi-point alignment in solar system live stacking, and should work well for solar/lunar images - most likely better than the 'Image Features' option which was previously the best choice for solar/lunar tracking.

* Having the choice to use 'Continuous' movements for calibration, rather than the original stepped movement approach. Continuous movement should avoid issues where the mount is slow to react initially to movement requests (leading to it not moving as far as expected in a 1s pulse) or has significant backlash that might affect each pulse.

In future, I expect that continuous movement and phase offset detection might well become the defaults for feature tracking calibration, but that will depend to some extent on feedback regarding how these work with a range of different cameras, mounts, etc.

In the background, I have also changed the code so that feature tracking will not move the mount at all until 3 position measurements in a row have been outside the 'dead zone' distance - this avoids issues with the software trying to respond to a single bad measurement and then moving back again to correct when further good measurements are taken.

Finally, another change behind the scenes will remove the need for calibration of feature tracking (and of the Solar/Lunar Framing assistant). Both feature tracking and the solar/lunar framing assistant need to calibrate to work out the orientation of the camera relative to the movement axes of the telescope mount and the scale of movement of the image for a given mount movement. SharpCap now stores this information centrally for the duration of a session, so if you calibrate Feature Tracking then you will find Solar/Lunar Framing automatically pre-calibrated (or the other way round). Not only that, plate solving gives the same information, so if you plate solve then that will set the orientation and scaling information for both feature tracking and Solar/Lunar Framing (well, not really solar, since you can't plate solve in the day...). In theory, the solar/lunar mosaic tool should interact with this central store of information too, but that is still to-do (it's more complicated due to primarily working in Ecliptic rather than Equatorial co-ordinates).

The stored data regarding the orientation of the camera sensor and image scale is maintained under the following conditions:

* SharpCap remains open
* You continue using the same camera (it can be closed and re-opened)

If you re-orient the camera or do something that would change the image scale (insert/remove barlow lens for instance), you should re-calibrate - otherwise SharpCap will still incorrectly use the old data, as it doesn't know you have rotated the camera or changed the optics. If you have a GOTO mount connected then SharpCap will correct the values for differing positions in the sky and also for changing side-of-pier on an equatorial mount.

cheers,

Robin

Hi Robin,

I have a question on Feature Tracking. If I calibrate using the full resolution of my camera and then later go to a lower resolution to select a smaller area, do I need to recalibrate?

Thanks,
JohnD

Hi,

that should be fine - moving to a smaller ROI (or from ROI to full frame) shouldn't be a problem - the existing calibration will still be active and will work.

cheers,

Robin

Hi Robin,

I tried the new tracking method and it works great! Thanks for the quick response I wasn't expecting to be able to use it this quick!

One small thing I experienced is that it still took me twice to go through the calibration. The first failed with something like "-120 vs 130". And my calibration graph looks a bit look a right arrow ->- and it's not as nearly as clean as yours. I suppose that's seeing effect.

I will keep testing and report if I find anything interesting.

Best,
Siyu

Hi,

good news that the new approach is working for you!

The arrow shape calibration graph may mean that the mount is pausing tracking when moving the Declination axis, so that the result seen is a superposition of the requested declination movement and the (untracked) movement of the sun/moon. Setting a higher calibration speed should help make the two lines more nearly opposite and avoid the failure if this is happening. Alternatively, going the other way and choosing to use the guiding rate for calibration may help, as the mount should continue tracking while the guiding movement is active.

cheers,

Robin

Hi,

By increasing the calibration speed do you mean increasing the step size (from 0.4 to 0.8 in this case)? This works but the shape is still weird. I tried lowering the movement speed to x4 and x1 but the first 2 calibration movements are almost 90 degrees. The continuous movement on the other hand seems to work amazingly well it's fast the graph is better looking.

Though for some reason I had to increase the max move duration today to keep up. I suppose it's because I didn't PA at all.

Hi,

no, not quite, I was suggesting using a higher movement speed in the dropdown at the top left. However, if continuous movement is working well for you in calibration then there is no benefit in moving back to stepped calibration (I am thinking of making continuous the default in a future update).

There will be a connection between the movement speed (top left) and the maximum move duration - if you select a lower movement speed and you have a lot of drift then you will need a larger maximum move duration otherwise the adjustments will not keep up.

cheers,

Robin