A current thread on eyepiece contrast brought up this matter, which I feel deserves its own treatment. Not that it hasn't been discussed here over the years!
It's something of a durable myth, among some folk, that the very fact of increased aperture by itself results in the visual discernment of subtler contrast. Some years back I went through several pages of discourse with another CNer who steadfastly asserted this. The dearth of interaction from other forumites suggested a general lack of certitude on their part. One on my missions is to beat back against such misperceptions.
In the context of this discussion, threshold contrast could defined as the limit at which a source of given surface brightness just becomes perceived as differing from its surrounds. The source could be darker (as in a dark nebula seen against unresolved starlight or bright nebulosity), but more commonly we tend to think in terms of brighter-than-sky objects such as glowing nebulae. And here we will restrict to the situation of dim, extended DSOs, where threshold contrast is especially crucial, although it's certainly important in the realm of the bright, such as planetary surface observation.
Other things being equal (basic instrument desgn, optical quality, transmission efficiency, control of unwanted light and exit pupil), threshold contrast for visual detection is invariant with aperture. If a 2" scope just permits to perceive [object + sky] surface brightness of, say, 0.07m brighter than sky, a 20" scope will do no better.
The preceding naturally assumes that the object in each case is easily larger than the minimum size for detection. And therein lies the key to this myth I'm knocking down.
There has arisen a veritable conflation of contrast transfer--as a *general* concept--with the modulation transfer function (MTF), as frequently used as a measure of optical performance. But the MTF principally concerns the small scale regime hardly much larger than the Fresnel pattern of diffraction. In other words, it has relevance where the subject is well within the bright photopic range and the exit pupil is small enough--typically less than about 2mm--to permit the discrimination of the effects concerned. Or to put it another way, the MTF for the most part concerns resolving power, at least when the optics are halfway reasonable.
Even if we are dealing exclusively in the realm of the small and bright, the MTF doesn't tell the full story. Besides diffraction, aberrations and small-scale scatter, contrast is afflicted at intermediate and large scale by such causes as optical reflections and non-optical surface scatter/reflections; this we term veiling glare. An actual measured MTF over all relevant scales would represent the full reality, but a calculated MTF (which we typically see) based only on and restricted to that range indicated by a wavefront measurement and consideration of diffracting obstructors misses the full picture.
In short, contrast transfer concerns more than the realm the tiny range the usual MTF chart covers.
Indeed, for dim, low contrast DSO observation, the common MTF chart is essentially irrelevant. For a difficult nebula against a fairly dark sky, the human visual resolving power *on the retina* is not the 1-2 arcminutes for daylight conditions, but instead is a truly awful 1/2 degree, or 1 degree, or as bad as 5-6 degrees at the limits of faintness/contrast. If f we were to choose a characteristic number, we could say that for faint fuzzy DSOs our resolving power is 100 times poorer than for lunar/planetary observing.
To see this most viscerally, install your solar filter and look at the Moon. The result will be a sunlit lunar surface dimmed to that of a moderately bright planetary nebula (not bright enough to discern color, but neither a difficult detection.) A quarter or gibbous Moon is best, for you then have still the intrinsically *very* high contrast terminator and shadows to examine. Note how terribly poor your ability to resolve familiar features now!
Now, back to the telescope's inability to improve upon threshold contrast detection when made larger. To this end, an easy to perform experiment might drive home the lesson better than can any amount of verbiage. This test is based on the following premise:
Getting physically nearer to a subject does not improve threshold contrast.
If one doubts this, the test would resolve the matter if a suitably large target comprising a low-contrast pattern is devised.
For a first order, simple appreciation, the target could be a poster. The poorer its contrast the better. Set it up at some goodly distance where the smaller features are not fully resolveable. Measure this distance.
Pull out or borrow a binocular of 5-7X (the lower the magnification, the less the distance you will have to walk--or bike . Examine the the target carefully through the bino. Now walk (or bike) toward the target, stopping at that distance whose ratio equals the inverse of the magnification. For example, if the initial viewing distance with bino is 100 feet, and the bino magnification is 5X, you will stop 20 feet short of the target. Now examine it with eyes alone.
If necessary, repeat (and get a bit of exercise . Do you find any notable differences in the two views?
A fuller test using a battery of binos would be more instructive. For instance, if one had 2X, 4X and 8X binos, comparisons could be made at respective eye-alone distances of 50', 25' and 12.5' (for the same 100' starting distance as above.)
And better yet, under conditions of low light, and using binos having at least near to 7mm exit pupils.
You see, employing a telescope is just like moving nearer to an object. This is exactly true if the scope's exit pupil is at least as large as the observer's iris.
For our earlier experiment, suppose you place beside the target a friend who looks back at you all the while you conduct your observations. Furthermore suppose your 5X bino has an exit pupil equaling your iris diameter; let's say this is 3mm, making for a 5X15 bino.
If your friend has a fixed-maginification telescope which can resolve your iris, he will see it subtend some apparent angle in the eyepiece. At the initial 100' starting distance, let's say he sees your 3mm iris subtending an apparent 1 degree in his eyepiece. Your bino's 15mm objectives will then subtend an apparent 5 degrees.
Now you move the 5X closer to the target, stopping at 20'. Your friend willl now see your iris subtending 5 degrees, which is the same angular size as your iris when 100' distant. This is why your naked eye view from 5X nearer to the target is exactly the same as the 5X magnified view from 5X farther away. In both cases the entrance pupil of the system viewing the target, as seen from the target itself, is identical.
Finally, as one point of evidence that the contrast threshold does not vary with viewing distance (and by extension, with changes in aperture), consider that as you look at the wall in your room right now and move nearer to and farther from it, its surface brightness does not vary in the slightest. Certainly NOT as a function of the inverse square law! (Which is a *huge* variance.) And so if the basic surface brightness is not varying, it must follow that any local differences in surface differences must not vary either. In other words, contrast does not vary. And if contrast is not varying, neither can threshold contrast.