Secondary Obstruction Test

Every so often  the arguments about the effects of secondary obstruction burst forth like flowers in the spring. The usual lines in the sand are drawn with the large aperture camp saying it's not worth worrying about while the APO refractor camp claim that it's a critical performance factor. Much has been written on the subject but as far as I can tell, little real world testing has been done. As it's such a simple question to answer I decided to run my own test to see just what the effects on a real optical system are.

For the test I used my 6" F/8 Newtonian, on the bench to image a daylight object. The images were recorded with an Atik 2HS monochrome CCD camera at prime focus. Two images were recorded. One with my normal 17% obstruction and one with a 43% obstruction mask  Focus and exposure was unchanged. Here are the resulting unprocessed images, 17% obstructed on the left, 43% obstructed on the right.

(note:  All obstructions quoted are percentage of aperture.)

At first glance the 43% image actually looks better but that's because it's a bit under exposed relative to the 17% image. The difference in density is due to the additional light loss of the larger obstruction. Following are the histograms for the images, 17%, 43% and overlay, 43% shown in grey:

The 43% histogram is clearly shifted to the left but more importantly the tonal range of the data is compressed (contrast reduced) as can be seen in the following overlay where the lowest tonal peaks of both images are lined up:

This represents a very real reduction in tonal range and contrast between the two images. The real question is, how does this effect the images in question. To answer that I've adjusted the density of the 17% image to match that of the 43% image, here's the result:

Now the real differences between the two images and easily be seen (remember, only brightness has been adjusted to match the exposures). The overall difference in contrast is obvious and the visibility of fine detail, especially in the low contrast areas is clearly enhanced in the 17% image. In some areas the 17% image looks sharper simply because the enhanced contrast is making small features easier to see. What we don't see is any of the resolution improvement that some have claimed for large obstructions.

After seeing the results of the above images I found myself to be rather curious about the effect of smaller obstructions verses no obstruction at all. Thanks to  Rich Richins of Albuquerque New Mexico, here are some examples using an Orion E80 APO refractor, unobstructed one the left and 20% obstructed on the right.

At first glance these look very much the same but once you really start examining these images closely you'll notice a few things. The biggest difference can be seen in the low value tones that make up the wall in the background. The contrast and definition in that area is noticeably degraded in the obstructed image. Also of note are the fine lines between the sheets of stacked wall board on the lower left of the frame. The shadow lines there are much better defined in the unobstructed example.

Just to make sure this wasn't a fluke of focus or conditions Rich repeated the test under better lighting conditions. Again, the 20% obstructed image is on the right.

The effects are similar to the first test. There is significantly less contrast in the mid and low tones that make up the tree on the right hand side. You can also note a general brightening of the shadow areas on the wall.  In addition there's a general loss of definition in the fine details of the roof shingles that was easily seen when examining the images full size.

Conclusions

This were simple, rough and ready tests that no doubt others can improve on. However, even these simple tests do allow some basic conclusions to be made.

Both contrast and visibility of fine detail is adversely effected by large obstruction percentages. The effect is seen both in raw images and visually which was checked at the time of the test. This could be of concern to both deep sky and planetary observers and imagers because contrast is the bottom line factor in the ability to discern fine details or in extreme cases, to even detect the presence of faint or low contrast detail. The APO tests also show that even the accepted maximum obstruction of 20% has a noticeable effect on image quality.

Although it's often minimized because it's always expressed in magnitudes, light loss through obstruction can also have significant effects. The effective F stop of the test 6" F/8 system is actually about F/8.9 when the 17% secondary blockage and the reflectivity of the standard coatings (91%) is taken into account. This drops to F/9.8 with the 43% obstruction in place. That's a significant loss to the deep sky imager or visual observer who needs every photon he can get.

The last question to ask is, does it matter? To a certain extent people will see what they want to see in these examples. Also it should be pointed out that 43% obstruction is larger then anything you'll see in a common commercial telescope although it wouldn't be unusual for larger professional instruments. Visual Newtonians will typically run under 20% but some compound instruments and photographic Newtonians will run to the high 30's.  The argument can be made that the contrast difference can be made up in image processing and to a certain extent, that's true. However, I think it's clear that the better the raw image, the better the final result, no matter how much processing is applied. This is also little comfort to the visual observer who has to take what they get.  In the final analysis every optical system is a blend of compromises. The negative effects of a large obstruction is real and while it may not be the number one driving force in a design, it does need to be taken into consideration.

Tony Gondola, 9/05