Description
Guiding oscillation occurs when the autoguider over-corrects: instead of stabilizing the star, it causes it to oscillate back and forth around the target position, much like a driver who swerves by steering too aggressively.
The result is kidney-bean shaped, short zigzag, or doubled stars -- a consequence of a too-aggressive or poorly tuned guider, not a mechanical defect in the mount.
Typical causes include: too-high aggressiveness, a guiding exposure that is too short (the guider chases the seeing), a bad calibration, or poorly handled backlash that causes oscillations in the declination axis.
This is a tuning defect and therefore free to fix. Distinct from periodic error (unguided, cyclic), from drift (unguided, monotonic), and from poor seeing (turbulence).
Visual signature
Stars take on a kidney-bean, short zigzag, or doubled appearance, without a clear single elongation direction.
The pattern is erratic from one frame to the next: it follows the guider's over-corrections and the turbulence, without the cyclic regularity of periodic error.
The effect is uniform across the field (mount and guiding origin), and often most pronounced on the worst-tuned axis (RA or Dec).
The guiding graph (RMS, PHD2 plot) confirms it: rapid oscillations around zero with repeated overshoot, rather than a slow, smooth error drift.
Differential diagnosis
Not to be confused with periodic error: periodic error is cyclic and intrinsic to the unguided mount; guiding oscillation is erratic and stems from a guider that is over-correcting.
Distinct from tracking drift: drift is a steady displacement in one direction; oscillation is a back-and-forth motion around the target.
Separate from poor seeing: seeing broadens stars diffusely and symmetrically, whereas guider over-correction deforms them into zigzags or kidney beans. A guider that chases the seeing combines both effects.
Not to be mistaken for wind vibration: wind causes sharp, isolated deformations, not an oscillation pattern tied to the guiding loop.
Probable causes
- Guiding aggressiveness set too high (over-correction)
- Guiding exposure too short (the guider chases the seeing)
- Incorrect or outdated guiding calibration
- Declination backlash not properly handled
- Poor seeing amplifying unnecessary corrections
- Guide star too faint or saturated (noisy centroid measurement)
Course of action
- Reduce guiding aggressiveness progressively until the graph stabilizes
- Increase the guiding exposure time (2-4 s) to average out the seeing
- Recalibrate near the meridian and the celestial equator
- Handle declination backlash (use unidirectional mode if needed)
- Select a guide star that is well exposed -- neither too faint nor saturated
- Set a target RMS appropriate for the main camera's plate scale
- Enable an adaptive algorithm (predictive PEC, multi-star guiding in PHD2)
The Doc's advice
An oscillating guider is almost always a sign of too much aggressiveness or a guider chasing the seeing. The counter-intuitive fix: slow down. Raise the guiding exposure to 2-4 s to average out the turbulence (you are not correcting the seeing, you are ignoring it), and lower aggressiveness until the graph settles. Calibrate near the meridian and the equator. On declination, manage backlash and consider unidirectional mode. Aim for an RMS that matches your plate scale: there is no point chasing 0.3 arcsec when your resolution is 2 arcsec/pixel. A calm guider beats a reactive one every time.
Think you can see this defect in your image?
Run a diagnosisFrequently asked questions
Why does my guider make the stars oscillate?
Because it is over-correcting. When aggressiveness is too high or the guiding exposure is too short, the system reacts to every micro-movement (including turbulence) and sends corrections that are too strong: the star overshoots the target, the guider corrects in the opposite direction, and the cycle repeats. On the captured frames, this produces kidney-bean or zigzag stars. The solution is to calm the guider: less aggressiveness, longer guiding exposures, and a proper calibration.
How do I set guiding aggressiveness?
Reduce it progressively until the guiding graph stops overshooting and oscillates less around zero. Aggressiveness set too high causes over-corrections; too low allows error to accumulate: there is a sweet spot specific to each mount. Combine this with a guiding exposure of 2-4 s, which averages atmospheric turbulence instead of chasing it. In PHD2, the guiding assistant and predictive algorithms help find stable settings. Aim for consistency rather than reactivity.
Should I guide fast to better track the seeing?
No -- that is the classic mistake. Seeing (atmospheric turbulence) makes the guide star dance at high frequency, and attempting to correct it only injects that motion into the mount: the guider "chases the seeing" and degrades the stars. Conversely, increasing the guiding exposure to 2-4 s averages out the turbulence so that only the true tracking error (drift, periodic error) is corrected. You do not correct seeing with the guider; you ignore it.
What RMS should I aim for in guiding?
There is no absolute answer: the right target depends on your plate scale (resolution in arcseconds per pixel). With a wide-field setup at 2-3 arcsec/pixel, a total RMS of 1 arcsec already yields round stars; at a long focal length with 0.5 arcsec/pixel, you need to aim considerably lower. The practical rule: guiding RMS should stay well below your plate scale. Chasing 0.3 arcsec on a setup that samples at 2 arcsec/pixel is pointless and a source of frustration. Match the effort to the actual resolution.