Recent improvements in complementary metal oxide semiconductor (CMOS) sensors for digital cameras have made them excellent performers for deep sky imaging, compared with the previous generations of CCD and CMOS chips.
However, most new CMOS sensors do have smaller pixels (3.76 microns) and relatively small sensors compared to CCDs. The combination of small pixels and long focal lengths often results in oversampling, which leads to a host of complications without significant increase in details.
We can compensate for this oversampling by using a reducer. While the solution is effective, it’s not perfect. We will look at how one type of reducer affects the field of view, vignetting and illumination, spot sizes, resolution, and image brightness.
Planewave Instruments 0.66 reducer summary
Here we test the Planewave Instruments 0.66x reducer (part #125166,), paired with the Planewave CDK 12.5-inch telescope, and a popular APS-C sized CMOS sensor ( IMX571). The reduction factor of the Planewave reducer is quite high, transforming the f/8 CDK to f/5.3.
Overall, the reducer performed quite well, although the short back focus will be a challenge for some imaging trains. You can see the large CDK reducer below, next to the popular two-inch Astrophysics 0.67x one (at left), and then attached to a CMOS camera and filter (at right).
Field of view (FOV)
The area captured in the frame is increased with the reducer. Below is a comparison of the simulated FOV with M101 (the Pinwheel Galaxy) with the reducer (white frame) and without (red frame).
The increased FOV with the reducer allows smaller APS-C sized chips and more economical 36mm filters to capture larger regions of the sky. Capturing the same FOV without the reducer would require a full frame camera and 50mm filters, resulting in two to three times the imaging train cost.
Vignetting is light loss in the corners. Here we compare the illumination with and without the reducer. These images are analyses of an evenly illuminated test panel, showing the resulting illumination of the sensor. The numbers represent the light loss relative to the center of the image.
From the plots above, we see that the extreme corners of the sensor receive about 5 percent less light with the reducer compared to without it, which is an acceptable result.
Spot Size Performance
Below is a plot indicating spot size performance. The blue line indicates the size of the stars (spots) in microns at varying distances from the image center (in mm) without the reducer. The red line shows the size of the stars with the reducer (provided by Planewave).
The furthest corner of an APS-C sensor is 14mm off-axis (distance from the center), which is marked by the purple vertical line. The chart shows that there is minimal increase in star size up to the corner of an APS-C sensor frame.
Below we look at real image data to compare the corners of an APS-C sized sensor, to the center of the image.
The mosaic above is a closeup of the image center and corners. The table below gives the measurements of star size (FWHM) in each tile of the mosaic, along with a measure of the roundness (eccentricity). For eccentricity, a value below 0.4 is considered very good in practice.
Here we see the center and the corners are almost equal, indicating that the reducer is not bloating or elongating the stars in the corners as some others do.
For best performance (smallest spot size, sharpest stars), the optimal spacing between the reducer and sensor is found to be 47mm.
A decrease in focal length can, in some situations, reduce resolution. However, these larger telescopes have longer focal lengths, resulting in image scales much higher than the seeing (atmospheric steadiness) will allow when paired with small pixels.
In this case, the native telescope image scale would be 0.3 arcseconds / pixel, which is usually too high a sampling rate (over-sampled). Even with the reduced focal length and change in image scale, the resolution is still greater than the sky in most parts of the world will allow due to light pollution.
The final Image above of IC1396 (the Elephant’s Trunk Nebula) was processed with only 11.5hrs of total integration time using SII, OIII, and H-alpha 3nm data.
The tests confirm the excellent performance of the reducer. This combination of large aperture and fast f/ratio allows imagers to capture larger objects in less time with a higher signal-to-noise ratio, and it has transformed my personal imaging experience. The only major drawback, besides its cost, is the lack of backspacing, which may limit its use in some circumstances.
Full resolution images (SHO, Hubble Palette) are available at https://astrob.in/o831ao/D/