Hello.

Could you please shed some light on the measurement of field curvature and it's effect on star images?

I have an imaging setup using an APO refractor, focal reducer/flattener and large format color ccd camera that is producing elongated stars in the corners. The elongation is in a radial pattern and the stars display a prism-like effect (red-white-blue). I suspect that the problem is with the focal reducer; perhaps an incorrect distance between the reducer and the sensor. I have been experimenting with this distance to determine if I can eliminate or minimize the aberation.

By shortening the distance between the reducer and sensor I have decreased field curvature from 20% to 10% (measured with CCDInspector) but without any significant change in star shape or chromatic error.

Is field curvature simply a measurement of FWHM values in the center of the image compared to the edge?

What effect does star elongation have on this measurement?

Could field curvature produce radially, elongated stars in the corners of the image or are field curvature and star elongation independent of each other?

Does CCDInspector measure star elongation?

Any help would be appreciated.

I have attached a full resolution, cropped image of the elongated stars taken from the top, left corner.

Thank you.

Gordon

Elongation_Crop.tif (62 KB, 42 downloads)

Hi Gordon,

quote:

I have an imaging setup using an APO refractor, focal reducer/flattener and large format color ccd camera that is producing elongated stars in the corners.

The larger star sizes in the corner, as well as, elongation is due to field curvature present in all refractors without field flattener. Some reducers are made to also help correct the field curvature and other aberrations, but these should be properly designed to fit your specific optical system and spaced at exact distance from the chip to perform well. You may want to find out from your reducer manufacturer what the proper distance is. This can be a very sensitive adjustment, as a difference of greater than 1-2mm can often result in reduced correction and possibly more aberrations.

quote:

The elongation is in a radial pattern and the stars display a prism-like effect (red-white-blue). I suspect that the problem is with the focal reducer; perhaps an incorrect distance between the reducer and the sensor.

The prism-like effect is due to chromatic aberrations. APOs are good at correcting this, but most, if not all, are not perfect.The reducer can contribute to this as well.

quote:

I have been experimenting with this distance to determine if I can eliminate or minimize the aberation.

By shortening the distance between the reducer and sensor I have decreased field curvature from 20% to 10% (measured with CCDInspector) but without any significant change in star shape or chromatic error.

Is field curvature simply a measurement of FWHM values in the center of the image compared to the edge?

Yes, CCDI measures field curvature as a percent difference in FWHM values from best focus to worst focus in the field of view. So, different image scale will affect field curvature number as displayed by CCDI. If you change the distance from CCD chip to the reducer, this will change the image scale, and so will affect the curvature measurement.

quote:

What effect does star elongation have on this measurement?
Could field curvature produce radially, elongated stars in the corners of the image or are field curvature and star elongation independent of each other?
Does CCDInspector measure star elongation?

Star elongation increases FWHM value of a star, as measured by CCDI. CCDI measures FWHM all around the star shape, and reports the median FWHM value which will increase as star elongation increases.

Aspect ratio is the measure of star elongation in CCDI. An aspect ratio reported is the percent difference between largest and smallest star radii as measured by FWHM.

Regards,

-Paul

Thanks Paul.

Good information as always.

I have one follow-up question.

As I change the distance from the reducer to the sensor, can I use both the percentage of curvature value and the aspect ratio value to compare results?

Thank you.

Gordon

Sure, Gordon, you can use both.

Aspect ratio at the edge of the image would tell you how elongated the stars are relative to round.

The curvature number is dependent on the actual FWHM value, which will change as you adjust the distance from reducer to camera.

Your goal is to find the position with the lowest curvature value, and the lowest aspect ratio of stars in the corners. If you can do this, the image quality should be the best you can achieve with this reducer.

Regards,

-Paul

Thank you Paul.

Gordon