Focal Reducer Test Report

INTRODUCTION

The first use of the Meade 0.33 Focal Reducer with my 8" LX-200 f/10 telescope and MX5C CCD camera was very disappointing.  I knew that I would not be able to use the reducer with my flip mirror system as there wasn't enough back-focus but I did not spend enough time understanding spacings, variations in focal ratio, vignetting, and optical distortion.  To be honest, I didn't know what I was doing but knew I wanted wider angle images.

After much discussion on the MX5C eGroup and help from people there, I came to understand the basics of this device and, more importantly, the limitations.   Spacing of the focal reducer to the CCD chip is critical.  The smaller the distance the higher the focal ratio.  The larger the distance the lower the ratio but the increase in light drop-off (a form of vignetting) from the center to the edge of the field becomes very pronounced.  The use of flat-fields to correct for this vignetting is mandatory (or cropping the image to eliminate that portion of the image).

So after a night of "imaging" with this device and the resulting poor, vignetted images, I set out to quantify some of the factors involved in effectively using the 0.33 reducer.  What follows is a summary of the work I did to achieve that goal.  All of the testing was done indoors.  The forecast for the night I did this was for clear skies but as has become the norm, it seems, this did not turn out to be the case so I spent the night indoors playing with focal reducers and taking images of an artificial star and a white card with a T-shirt over the end of my scope. 

Since I also have a Meade 0.63 focal reducer I included it in the testing as well.  I was disappointed to learn that with the various combination of spacers and the camera nosepiece (why are they called that?),  f/6.3 is a figment of Meade's imagination (at least when used with the MX5C camera) .  The shortest I could achieve was f/7.3 with all the spacers I had available installed.

FOCAL RATIO TEST

The first set of tests were done to quantify the actual focal ratios resulting from the different reducers and various combinations of spacers. The methodology used was to take the image, mark a point on the wide portion of the artificial star image directly across from the "star" on the left and the corresponding point on the diameter line on the right side.  These are the "pixel points" in the chart.  The difference between the points was compared to the same distance in the f/10 case which was used as the reference and all other calculations made from it. The results are summarized below:  (note last column added 3/21/01)

Notes:

• 8" LX-200 f/10 telescope with Celestron (same as Meade) 0.63 reducer and Meade 0.33 reducer, MX5C camera.  NGF-S Focuser installed for all tests.
• All distances in mm
• Distance from end of camera to surface of CCD chip is 17.5mm
• * "Measured" distance is from the lip of the focal reducer (shoulder where the threads stop and the smooth part with the printing starts) to the end of the camera shell
• Variations in distances are due to different reducers as well as threading a T-thread into the camera's Pentax thread.  This works if you are careful and don't over-tighten.

(1) This column represents final measurements using the MX7C AND a Pentax to T-Thread adpater from Adirondack Video

As can be seen,  the "0.63" reducer never provides a f/6.3 system.   The T-ring and 2 spacers (15 and 30mm) provided with the Meade 0.33 reducer were also used with the 0.63 reducer and even when all spacers were used, f/6.3 was not achievable.

FLATNESS TEST

The next thing tested was the flatness of all of the 0.33 combinations.  It was assumed that the f/10 and any of the 0.63 reducer combinations were flat so only the 0.33 reducer was tested.  I'm not sure I performed this test correctly and it should be noted that the "flats" produced are not useable as flats for correcting images.   The uniformity of light was not very good and since I was looking for a comparison of the four 0.33 configurations the relative vignetting was what was desired.  The way the tests were done was to install the 0.33 reducer with no spacer and focus on the artificial star.  A white card was placed in front of the star and indirectly illuminated with a light at the angle opposite from that the scope was pointing at.   A T-shirt was tightly stretched across the OTA of the scope and images taken.   The scope was defocused about two turns of the manual focuser on the LX-200 and all tests were done with the same focus setting (this may be a mistake in this testing but since it was going to make the work an order of magnitude harder to refocus each time,  I opted for the easy way).  At the end, I tried various other focus setting and the differences were slight. 

Here are JPEG's of the luminance frames created from each case.  Each image was taken and exposure times adjusted such that the max ADU value was about 2500 (of 4095 for this 12 bit camera).  As the speed of the system increased the exposure time was decreased to compensate.  The visualization of the lum frames was adjusted with a linear transfer function and so that the min value was at the start of the histogram rise and the max value where it began to fall (for lack of any better standard).

	IMAGE1 (f/6.1) 0.33 Reducer, no spacer (just the T-ring supplied with the focal reducer), exposure time is 1.7 seconds.
	IMAGE 2 (f/4.8) Same as Image1 but with 15mm spacer and 1.0 second exposure time
	IMAGE 3 (f/3.4) Same as Image1 but with 30mm spacer and 0.55 second exposure time
	IMAGE 4 (f/2.0) Same as Image1 but with 45mm spacer and 0.43 second exposure time

The amount of vignetting can be clearly seen in these images.  I cannot explain why the bright section of Images 2 and 3 are not in the center.  My guess is that is is collimation of all of the elements in the system or a manifestation of the non-uniform light source (it was slightly off to one side).  The result in Image 4 is totally unexplainable (by me at least).  I do remember when I tried the full up spacer combination represented by image 4 on the real sky,  I observed that the stars at the edges were beginning to look like little V's.  In this f/2 case, the field of view was very wide, of course, and the image in the center was useable.

SUMMARY

With both the 0.63 and 0.33 reducers and the spacers and T-ring supplied with the 0.33 reducer, many focal ratios from 2.0 to 7.8 are attainable, albeit with varying quality of image.

The system may be used reasonable well in the  f/3 range but flat-fielding is mandatory (as it is for f/5 operation).  Use at f/2 can be set up but real images need to be assessed to see if it is at all effective.  The whole question about over or undersampling remains and my next step is to try to understand that topic.

I have zipped up all of the raw images, JPEG's, and an Excel spreadsheet used for the focal ratio calculations and you can email me for the 2.2 MB zip file if you are interested.

Now I will wait for a (rare) clear night to put this new found information to use.