ACP contains a built-in adaptive telescope pointing corrector that is similar to the MaxPoint™ pointing corrector from Diffraction Limited. The big difference is that we have integrated pointing correction into ACP in a way that makes its operation automatic, adaptive, and transparent. No other software can do this! The pointing models that ACP uses are compatible with MaxPoint, so you can read an ACP model into MaxPoint and use MaxPoint's analysis features!
If you want to use the telescope mechanical analysis capabilities of MaxPoint with ACP's pointing models, see Using MaxPoint. You can also convert ACP pointing models into the Software Bisque TPOINT-importable format using the ModelToTPOINT.vbs standard script. For more information, see Analyzing With TPOINT.The corrector engine learns about pointing errors whenever ACP is called upon to synchronize the telescope coordinates. When correction is enabled, sync operations don't actually reset the telescope coordinates. Instead, the desired and actual coordinates at the time of sync are sent to the corrector as a mapping point which immediately updates the model used to predict the telescope's pointing behavior. Each time a sync is performed, whether during automated image acquisition, manually through ACP's controls, from one of ACP's utility scripts, or from a client (such as MaxIm DL, TheSky or Starry Night) connected to ACP's ASCOM telescope hub, the pointing engine is immediately updated with a mapping point.
When ACP is first started with a new telescope and with correction enabled, it needs a minimum of six accurate syncs at fairly widely spaced points in the sky before the correction starts to take effect. For most mounts, after ten or so syncs, the pointing accuracy will be below two arc minutes, often below one arc minute, regardless of mechanical flexure or polar alignment errors (within reason!). If the mount has poor slewing position repeatability, more mapping points will be required. A mount with poor repeatability will never achieve excellent corrected pointing.
The syncs must use accurate coordinates. Hand-centering stars in an image will not reliably produce the accuracy needed to make the best use of the corrector. There is a utility script, TrainCorrector, that can build a model, however you don't need to train the corrector, though! Read on...Since the oldest mapping points are discarded, the model can adapt to slow changes in the telescope's mechanical characteristics. There is a tradeoff, though. If you keep relatively few mapping points in the model, the corrector will adapt faster, but the overall accuracy will be lower. More mapping points means slower adaptation and more accuracy (up to a point). Unlike some other correction systems, ACP's corrector typically reaches its maximum accuracy with 50 or fewer points for decent mounts.
The first sync given to ACP after ACP connects to the telescope will be sent through to the telescope (instead of creating a mapping point as described above). This happens each time ACP is (re)connected to the scope. The reason for this is to remove any random encoder bias from the correction solution. On some scopes, particularly low-end ones, there is a constant error in one or both axes after they are started up. Syncing removes this error, keeping it out of the pointing corrector math.
The COR annunciator light will flash, indicating that the model is not active, until you have completed the initial sync.During routine operation, you don't need to know about this or do anything special. Just start up, connect and begin acquiring images with ACP. This info is included here because scenarios come up where users need to know about it. It's hoped that this info will reduce the frequency of questions related to this.
A frequent concern is "What happens if I observe many targets in a small area of the sky? Will the pointing corrector fill up its list with points in that area only?". No, ACP is smart enough to avoid this. If it finds a point within 5 degrees of a new one, it replaces that nearby point with the new one. That way, the mapping information is kept fresh, but points will remain distributed around the sky.
Once you have provided the corrector with an initial set of mapping points, ACP makes corrections by altering the coordinates going back and forth between it and the telescope, according to its mechanical model. If you don't alter anything mechanically (including payload and balance), the model will remain valid. If you shut down ACP, it assumes that you have also shut down and restarted the telescope and will do one real sync the next time it is connected.
A permanent record of the model (in the form of a MaxPoint-compatible list of mapping points) is kept on disk in the file Public Documents\ACP Config\Active.clb. The next time ACP is started, the model is restored and pointing correction is immediately effective. Further observations will serve to refine the model.
The pointing corrector takes your desired coordinates and adjusts them to get the imperfect mount to point where you want it. It also takes the imperfect telescope coordinates and reverses out the correction before displaying them. Thus, you only see "ideal" coordinates in ACP's displays, and you only use ideal coordinates when specifying a slew or sync.
Based on the mapping points supplied (via ACP sync operations), the corrector calculates the following mechanical errors in the mount:
| Error Term | Description |
|---|---|
| RA Bias | Error in the RA axis encoder (static after telescope startup) |
| Dec Bias | Error in the Dec axis encoder (static after telescope startup) |
| Polar Azimuth | Polar axis azimuth (left-right) misalignment (static) |
| Polar Altitude | Polar axis altitude (latitude, up-down) misalignment (static) |
| Axis Alignment | Deviation from exactly 90 degrees between the RA and Dec axes (static) |
| Collimation | Deviation of the actual optical axis from the tube axis (static) |
| Dec Flexure | German Equatorial Only: Flexure of the declination axis of the mount (varies with hour angle and declination) |
| Fork Flexure | Non-German mounts only: Flexure of the fork (varies with hour angle) |
| Tube flexure | Flexure of the optical tube assembly (varies with hour angle and declination) |
The mapping points are processed by the corrector which calculates a "best estimate" of the eight active model parameters. It does this each time a new mapping point is added (via a sync in ACP or one of its hub clients) and uses all of the observations added so far, up to a configurable maximum. The next time ACP is started, further observations will serve to refine the model.