•

# Why a too-small iris results in reduced aperture

54 replies to this topic

### #26 GlennLeDrew

GlennLeDrew

Voyager 1

• topic starter
• Posts: 12743
• Joined: 17 Jun 2008

Posted 27 August 2013 - 05:36 AM

Jon,
Your diagram does not address the matter of the reduced iris. You've made the iris large enough to accommodate the full light bundle from the objective.

Execute another version, but 'pinch' down the observer's iris to 1/2 the diameter as made here. You will find that the central 1/2 of the objective is now contributing to image formation.

And by the way, the label, "half field angle" is not correct, as it's referring to the semi-angle of the light cone for an image point, which is a reflection of the t/ratio. The two sets of rays you draw come to a common focus at the field center. And so the top set are really paraxial, or parallel to the optical axis, being rays from the on-axis image point. Rays which emanate from any angle off axis do not come to focus on the optical axis.

### #27 GlennLeDrew

GlennLeDrew

Voyager 1

• topic starter
• Posts: 12743
• Joined: 17 Jun 2008

Posted 27 August 2013 - 05:45 AM

Frank,
You raise valid points, of course. But it would be preferable that the fundamentals are assimilated before drilling down into that kind of detail. If the very concept of the reduced aperture is not fully understood first, further confusion can only follow.

### #28 freestar8n

freestar8n

Fly Me to the Moon

• Posts: 5191
• Joined: 12 Oct 2007

Posted 27 August 2013 - 06:08 AM

I think that in many situations the details of how pupils are defined are unimportant - but when it comes to aperture reduction and so forth - more rigorous adherence to the formalism has clear benefit. That's why the formalism exists in the first place. In order to talk about the entrance pupil of a system, you need to establish what is acting as the physical stop, and then determine the entrance pupil diameter and location, you need to do a ray trace of that stop through the optical elements in front of it. Once you establish the resulting image of that stop, it has a size and location - in object space - and *that* is the entrance pupil.

When a small camera lens, or human iris, is placed some distance behind the eyepiece, its size *and* location both determine the size and location of the entrance pupil. If the limiting aperture is not at the location of the exit pupil of the telescope, then the corresponding entrance pupil of the system will not be conjugate with the objective - and the objective will act as a cross between a field stop and an aperture stop.

There is nothing hard to understand about where the pupils end up - because it is just a ray trace. And I think it is all much clearer if you stick to the formalism that has been established - because it becomes critical in discussions like these.

You have to ask - what is acting as the physical aperture stop, what is the resulting image of it, where is it, and how big is it. If the aperture stop is not at a location conjugate with the objective, as happens with an afocal digital camera, these details become essential.

Frank

### #29 GlennLeDrew

GlennLeDrew

Voyager 1

• topic starter
• Posts: 12743
• Joined: 17 Jun 2008

Posted 27 August 2013 - 07:08 AM

Frank,
To help in the obtaining of an understanding of the *basics*, I'm limiting to the ideal case of a system whose entrance pupil is always the objective. That is, we're not concerned with restrictors between objective and field stop for any and all field angles. Furthermore, we're assuming there is no impediment to getting the iris in the plane of the exit pupil.

Would you agree that this is a valid approach to follow in the quest for an understanding of the *fundamentals*? Until this has been achieved, is it not wise to defer the plunging into the subtler details regarding internal stops and longitudinal displacement of iris and exit pupil?

I gather you're still skeptical about the flashlight test, or even its superior variant, the beam-expanded laser test. If so, we two have a seemingly intractable and fundamental disconnect in communication on the matter of pupils. Can you imagine the poor neophyte's head scratching when straying from the more important basics too soon?

### #30 GlennLeDrew

GlennLeDrew

Voyager 1

• topic starter
• Posts: 12743
• Joined: 17 Jun 2008

Posted 27 August 2013 - 08:02 AM

To show how a small iris, as it moves around within the confines of the larger exit pupil, continues to project as the same-diameter reduced aperture, I've overlaid additional light ray bundles to the bottom panel in my original illustration. This was the quicker expedient, and I beg your forgiveness for the clutter.

IMPORTANT NOTE: I have not moved the eye downward the required distance of 1/2 the iris diameter. You are unfortunately asked to imagine this instead.

As noted, the eye/iris is located centrally--on the optical axis--within the larger exit pupil. The envelopes of light bundles for image points from the top edge (blue), middle (white) and bottom edge (red) of the FOV are drawn as lines only. Because the iris is centered, all light bundles which contribute to image formation on the retina pass through the central part of the objective.

For reference, note the three relevant diameters indicated (drawn inside the eyeball circle); the centered iris diameter and location as drawn, the iris position when offset to the bottom edge of the exit pupil, and the full exit pupil diameter actually produced by the optics.

Added in solid fills of the same color scheme as before are the light bundles which would be accommodated when the iris is moved downward until it just 'touches' the edge of the exit pupil. The portion of the objective now utilized is at the top edge.

But note this important fact; the view angle has not changed. The field of view is centered on the very same point, and does so always. As will be stressed again, we are merely sampling a smaller portion of the full envelope of light the instrument can field.

Let's say the exit pupil is 6mm and the iris is 3mm. The system is thus stopped down to half its actual diameter, and the image is 1/4 as bright. If the objective were, say, 60mm in diameter, it's now working at 30mm. As long as the iris wanders about within the confines of the exit pupil, there will be no pupil clipping, and so the image brightness remains constant. While this is going on, the image of the iris projected onto the objective moves around within the confines of the objective, always as a 30mm diameter aperture.

Once the iris crosses the edge of the exit pupil, its image crosses the edge of the objective, the resulting pupil clipping causing image diminution.

Here's how to see all this in operation. On a piece of paper or cardboard, cut out a hole no larger than 1/2 the objective diameter. (A smaller hole makes for a more dramatic demonstration!) Place the sheet in front of the objective. While looking at the exit pupil, move the hole about in front of the objective. Note how the reduced exit pupil moves about, just like a miniature version of what's going on up at the front end. This can be easier to see if you first make a reference surface from a piece of tracing paper which has a circle drawn upon it equal to the unobstructed exit pupil diameter, held motionless in the plane of the exit pupil. And a long focal length eyepiece, making for a larger exit pupil, makes things easier to see as well.

As you know, the hole in the card is simply an aperture mask, like those used for stopping down scopes to yield sharper planetary views. Just as the smaller, offset objective mask results in a correspondingly smaller exit pupil similarly offset, so does a small iris at the exit pupil do the very same in reverse. At all times--in both situations--one is merely sampling a smaller portion of the full light bundle the scope is capable of accommodating. In terms of image brightness and resolving power (neglecting for the moment the matter of increased aberrations which can be introduced by the peripheral areas of optical elements), neither the optical system nor your eye cares a whit what portion of the entrance/exit pupil is being utilized.

Let's imagine that on a bright day your scope is producing an 8mm exit pupil. You install an objective mask which produces a 2mm exit pupil. Let's say your daytime-restricted iris is also 2mm. To see anything like a full-brightness image, you must locate your iris pretty darn precisely (to within 1/2mm anyway). If the objective mask was placed at the objective's edge, the reduced exit pupil would be offset to the edge of the 'native' 8mm exit pupil.

Now we remove the mask, and the exit pupil becomes 8mm. Does the image get any brighter? No, bacause your 2mm iris is still stopping down the system. If your eye meanwhile remained perfectly motionless, the very same portion of the objective which was used with the mask in place is still being used now. You could hold your hand up in front any other part of the objective and it will have absolutely zero impact on the view. Moving your iris about within the confines of the 8mm exit pupil only results in a different portion of the objective contributing to image formation at any one moment.

In this example, with mask in place or your iris by itself stopping down the system, the aperture is reduced to 1/4, the working f/ratio of the objective increases by a factor of 4, the image becomes 16X dimmer (compared to the full exit pupil), and for the on-axis iris at least, all telescope system aberrations are reduced. This holds true wherever the iris falls within the exit pupil. Whether it's the eye or an afocal camera behind the eyepiece, these are inviolable facts, as long as the iris can be brought to the plane of the exit pupil.

### #31 dan_h

dan_h

Vanguard

• Posts: 2267
• Joined: 10 Dec 2007

Posted 27 August 2013 - 08:04 AM

There is nothing hard to understand about where the pupils end up - because it is just a ray trace.

If that were the case, this thread would not exist.

dan

### #32 freestar8n

freestar8n

Fly Me to the Moon

• Posts: 5191
• Joined: 12 Oct 2007

Posted 27 August 2013 - 08:19 AM

I think it should be pretty obvious to anyone, using simple descriptions and few details, that viewing through an eyepiece with an iris smaller than the exit pupil, will result in a loss of throughput - and an effective loss of aperture. But there is confusion in the subsequent discussion about all the aperture "being used" - and the challenges of afocal imaging with a digital camera.

Your original post here doesn't come across as intended to be simplified - it sounds like it is meant as a formal description of pupils and how they work. In that regard I think it is good to point out, for example, that the exit pupil is not defined as "an image of the entrance pupil." Both pupils are images of the physical aperture stop, as seen from different spaces, and as formed by the subset of elements in the path to the aperture. That is a very different description from yours, it is consistent with the formal definition, and it elucidates concepts critical to a non-confused discussion of what happens in afocal camera imaging.

The flashlight test is definitely a bad way to measure aperture - in general - precisely because it assumes the entrance pupil is near the objective. This is not the case when you have a limiting stop in an unusual location - particularly near the focal plane - as happens with e.g. binoviewers or using a camera afocally.

As an aside - remember that you did change the flashlight test to the "collimated laser" test - due to its inadequacies. I thought we had made at least that much progress on the topic - but now I see people talking about the flashlight test again - so there was some ground lost. In many topics such as these, a clear understanding of pupils and how they work is essential and can't be glossed over, which is why I was somewhat surprised you took on the task to give a tutorial on pupils.

For both the flashlight test and afocal imaging, the location of the pupil is critical, and the simplifying assumptions that make it easier to visualize will prevent an actual understanding of how it all works.

I can see maintaining a simplified version of this topic in a beginner section, but not in this forum where more advanced topics come up all the time - and there is generally a desire to adhere to formalism and nomenclature.

Anyway - basic concepts of pupils and how they work should be Optics 101 - and are normally in the introductory material of an optics text. I don't know why you consider it to be an advanced topic in the first place.

Frank

### #33 freestar8n

freestar8n

Fly Me to the Moon

• Posts: 5191
• Joined: 12 Oct 2007

Posted 27 August 2013 - 08:38 AM

If that were the case, this thread would not exist.

My point is that even with the complete definition of pupils and how they work, determining their sizes and locations is a matter of simple ray tracing and linear imaging. It's not like it introduces quantum optics or even aberration theory. It's a simple definition and a ray trace - in both directions - from what is determined to be the acting aperture stop.

Frank

### #34 GlennLeDrew

GlennLeDrew

Voyager 1

• topic starter
• Posts: 12743
• Joined: 17 Jun 2008

Posted 27 August 2013 - 08:55 AM

Frank, while a bit off topic, the flashlight test makes absolutely no assumptions whatsoever about where a restricting stop might lie in the system. That's the beauty of it. It simply defines the full envelope of on-axis light that can enter the objective and make it out the eyepiece. No matter where a restrictor might lie.

Parallel light into the eyepiece and parallel light out the objective traces and delineates the one and same envelope of light coming into the objective and which makes it out the eyepiece.

I introduced the laser refinement only so as to ensure the tiniest point of light at the common eyepiece/objective focus, which ensures the sharpest shadow edge, even for a restrictor which might lie not at all far from the focus. For most optical systems, as long as the flashlight source delivers a reasonably sharp and unambiguously measurable circle of light out the objective, it can be taken as sufficiently reliable.

The sharpness of the shadow is what is an indicator of validity.

I can see that a new thread, illustrated, is what's required.

This flashlight test issue between you and me is the equivalent of this current thread between Jon Wall and myself.

That most everyone is remaining on the sidelines while we 'titans' clash (in a good way!) tells me that such topics as these are not fully understood by the vast majority of amateurs. Hence my desire to aducate, but hopefully in a way resembling measured steps. If I diverge from canonical definitions, that's a result of my own ignorance. And due to the kinds of simplifications and stripping down to essentials I'm striving for.

### #35 freestar8n

freestar8n

Fly Me to the Moon

• Posts: 5191
• Joined: 12 Oct 2007

Posted 27 August 2013 - 09:15 AM

I don't consider the flashlight test entirely off topic - it's actually pertinent. And I fully support an opportunity to educate about the concept of pupils. But I recommend sticking to the formalism more - not just to be fancy - but without it, you can come to very wrong conclusions.

If you recall, it was trivial for me to place a small aperture near the focus of an sct and measure the entrance pupil with the flashlight test - and get a very real looking measurement that was completely wrong. There was nothing subtle about it - it was a great looking, and completely wrong result. The way to understand why the test failed so demonstrably was to draw a simple diagram showing the size and location of the entrance pupil - which emphasized the need for a very accurately collimated beam - which then led to the collimated laser test. In order to make the flashlight beam collimated, you have to hold it far from the eyepiece - in which case very little light gets through - again because of the flux received by that small "entrance pupil" (the exit pupil of the eyepiece in reverse). So when the flashlight is properly placed, the image with large aperture scope is much fainter - which again leads to the need for a laser.

If you place a flashlight near an eyepiece the beam will not be collimated and it will be divergent - which will generate a large error in the shadow size if the entrance pupil is far from the objective. This is optics/pupils 101, and I strongly encourage you to read up on it - both as it relates to the flashlight test, and to afocal camera imaging.

The definition, size, and location of the pupil is critical when it is created by an aperture stop that isn't at a "normal" location conjugate to some primary element. This happens all the time and should be handled properly - and it really isn't hard to do if you assume knowledge of ray tracing - which I believe is pretty safe in this forum.

Frank

### #36 Jon Isaacs

Jon Isaacs

ISS

• Posts: 54938
• Joined: 16 Jun 2004
• Loc: San Diego and Boulevard, CA

Posted 27 August 2013 - 09:35 AM

I think it should be pretty obvious to anyone, using simple descriptions and few details, that viewing through an eyepiece with an iris smaller than the exit pupil, will result in a loss of throughput - and an effective loss of aperture.

It's not so clear to everyone and historically this has been an issue that has been discussed with some heat. In particular, the issue of the resolution loss caused by an aperture masked by the iris. Several years ago Alan French was in a heated discussion on this very topic in the Astromart forums.

In the current CN binocular forum, there are those who gained understanding from Alan's careful explanation of this particular issue though he was one against many..

'nuff said.

Jon

### #37 freestar8n

freestar8n

Fly Me to the Moon

• Posts: 5191
• Joined: 12 Oct 2007

Posted 27 August 2013 - 10:19 AM

OK - well it does seem odd to me consider how basic it is compared to much more advanced topics discussed here. I'm all for elucidating it especially if people are interested - but if this is a known topic of contention, then I think it is even more important to treat it more rigorously. I don't know about the discussion you are referring to - so I can't comment on the treatment it gave on the topic.

I'll try to put some diagrams together, but at the same time - the basic concepts of pupils are treated in most any text on optics - but people seem to regard the entrance pupil as some very obvious lens or physical hole - when it often isn't so simple.

Separately - the questions about coupling a camera to a telescope afocally cannot be discussed properly without explicitly dealing with the location of the camera pupil relative to the eyepiece. Since that topic is intertwined this the one regarding the more basic aperture reduction caused by a small human pupil - I think it makes sense not to oversimplify the discussion.

Frank

### #38 Jon Isaacs

Jon Isaacs

ISS

• Posts: 54938
• Joined: 16 Jun 2004
• Loc: San Diego and Boulevard, CA

Posted 27 August 2013 - 10:34 AM

Separately - the questions about coupling a camera to a telescope afocally cannot be discussed properly without explicitly dealing with the location of the camera pupil relative to the eyepiece. Since that topic is intertwined this the one regarding the more basic aperture reduction caused by a small human pupil - I think it makes sense not to oversimplify the discussion.

This discussion did start out addressing aperture masking and the human eye.

Jon

### #39 MitchAlsup

MitchAlsup

Vanguard

• Posts: 2173
• Joined: 31 Aug 2009

Posted 27 August 2013 - 10:57 AM

Hi Mitch,
How does the lit retina create the background glow? How bad is it? I have not noticed before but sounds interesting and good to know for eyepiece selection.

Light bounces off the iris and then illuminates the cornea from behind causing what looks a lot like skyglow.

### #40 freestar8n

freestar8n

Fly Me to the Moon

• Posts: 5191
• Joined: 12 Oct 2007

Posted 27 August 2013 - 10:59 AM

OK. Then - if the eye pupil is physically located at the exit pupil of the eyepiece, and if the eye pupil is smaller than the original exit pupil - is there anyone here who still thinks the effective aperture, for collecting light or for diffraction resolution, has not been reduced?

Note that this is separate from the full aperture "being used" to reduce vignetting as you look around the scene or something.

Is there anyone here who thinks the full aperture and speed of the lens is being used in this situation, for on-axis image quality and brightness?

If so - can you describe why that would be?

Frank

### #41 freestar8n

freestar8n

Fly Me to the Moon

• Posts: 5191
• Joined: 12 Oct 2007

Posted 27 August 2013 - 11:07 AM

Jon,
Your diagram does not address the matter of the reduced iris. You've made the iris large enough to accommodate the full light bundle from the objective.

Oh - I just realized something much more fundamental about that diagram. All the parallel input rays on the left should be emerging parallel on output. The diagram is incorrect and combines the role of the eyepiece with the role of the eye.

In the corrected diagram, the large parallel input bundle would become a smaller, but still too big, parallel output bundle that slams into the human iris and only lets the central part through - with a corresponding loss of light.

Does anyone disagree with that?

Frank

### #42 Ravenous

Ravenous

Surveyor 1

• Posts: 1584
• Joined: 14 Nov 2009
• Loc: UK

Posted 27 August 2013 - 11:24 AM

In the corrected diagram, the large parallel input bundle would become a smaller, but still too big, parallel output bundle that slams into the human iris and only lets the central part through - with a corresponding loss of light.

Does anyone disagree with that?

Frank

I suspect you're right - I wasn't able to understand where the extra lines at the eyepiece came from. I didn't want to ask for an explanation, as things were getting too argumentative anyway...

But I think Glenn's original post is long overdue. Newbies reading around here may get misled by some of the other posts.

### #43 GlennLeDrew

GlennLeDrew

Voyager 1

• topic starter
• Posts: 12743
• Joined: 17 Jun 2008

Posted 27 August 2013 - 01:20 PM

Heh heh. I missed that aspect in Jon's diagram, where the emergent rays out the eyepiece, which formerly passed through the objective edge zone, are not parallel to the central rays. I was concentrating too much on the action ahead of the eyepiece.

This cardinal mistake (to make nothing of the assertion of full aperture use in spite of a restricting iris) by an experienced telescope maker is ample enough reason to cover this whole business.

Frank,
For the flashlight aperture test, from word one I always stressed the importance of placing the flashlight no nearer to the eyepiece than about 10 eyepiece focal lengths, to assure reasonably parallel light going into the exit pupil. This ensures that the tiny image of the light is formed near enough to the focus to make for a reliable result.

And as I mentioned in our previous discourse, if one has any doubt as to parallelism of emergent light, just measure the illuminated circle near the objective and at least a full focal length beyond this. If the two measures are the same, the result is very accurate.

That's the crux of the matter. If the circle is sharp, and is known to be parallel, and emerges sufficiently parallel to the optical axis, its diameter truly is the working aperture. Parallel light straight into the eyepiece, parallel light out the objective; that's all you need. No requirement to beetle one's brow about pupil or stop locations.

I will be cooking up a fuller treatment on this!

### #44 Nils Olof Carlin

Nils Olof Carlin

Vanguard

• Posts: 2009
• Joined: 26 Jul 2004

Posted 27 August 2013 - 02:03 PM

You can use an oversized exit pupil, if it gives you a wider true field - it may not in practice, you should check.

If too large, the obstructed part (if any) of the exit pupil will be wide enough to be bothersome by blocking much of the eye pupil. Not least when you observe the sun or moon, and your eye pupil is much smaller than it is during deep sky observing. A low-power EP good for deep sky may be useless here.

The eye is a globe more or less, rotating in its "socket" - this means that the iris will move sideways when you shift your gaze (it is very roughly 1/2" in front of the center of rotation). Matching the exit pupil accurately to the iris will mean that even the smallest shift of gaze will give you a (noticeable) loss of light, until you have re-positioned your eye.

Choosing exit pupils that are at least 1 or 2 mm larger or smaller than your dark-adapted pupil will make observing easier.

Nils Olof

### #45 Mark Harry

Mark Harry

Fly Me to the Moon

• Posts: 6558
• Joined: 05 Sep 2005
• Loc: Northeast USA

Posted 27 August 2013 - 02:49 PM

I think the exit/entrance pupil has been hammered into threads concerning singlet aperture telescopes a bit unfairly in the past referred threads, and a bit too often.
There is a balance that has to be realized between aperture, and light gathering that has been glossed over; and is backed up with what is exactly detectable by eye, vs resolution on reduced aperture.
Specifically it's all touted as theoretically possible, but what is actually visible. And by eyeball, each orb has varying degrees of acuity.
******
I repeat the question- How many of the "titans" have actually made a singlet telescope that actually has achro-description performance??????????????????????????????????????????????????????????????????
Or who has equated definitions with achro performance between what is worse/better than described conditions/capabilities in what the human eye perceives???
----
If they haven't, there should be an obvious silence.
The disgusting post of an enlarged image of the -TIP- of the antenna (full of lateral color) doesn't mention the central part of the image where the lateral color is -NILL-.
Such biased posting in the mask of a forum should be discouraged, or banned; and shows a lack of fundamental understanding of human oriented vision, and a total lack of consideration of civilized man to bolster an ego oriented bent on making their fellow man look stupid.
(been said before- if it works for those experimenting, who are -YOU to say what the definition of visual acuity should be?????)
Wake up; moderators.

Mark

### #46 Boot

Boot

wildly diverse musical tastes

• Posts: 7993
• Joined: 04 Jun 2007
• Loc: Outstanding in Left Field

Posted 27 August 2013 - 02:57 PM

Well said, Mark.

### #47 Mark Harry

Mark Harry

Fly Me to the Moon

• Posts: 6558
• Joined: 05 Sep 2005
• Loc: Northeast USA

Posted 27 August 2013 - 03:48 PM

Thanks, Joseph.
M.

### #48 GlennLeDrew

GlennLeDrew

Voyager 1

• topic starter
• Posts: 12743
• Joined: 17 Jun 2008

Posted 28 August 2013 - 01:53 AM

Mark,
This thread was motivated by an evinced lack of understanding that a restricting iris reduces the working aperture commensurately. This was confirmed by a post made here, in which a supporting diagram was badly in error.

Furthermore, my issue was with the unknown, or at least uncertain, variables involved with the images used to support claims of performance on the Hypo. A reasonable deduction based on what scant evidence was provided had the actual working aperture reduced to but a fraction, which is known to reduce aberrations. I wanted to see images which assuredly used at least near to the full available aperture.

Someone could easily post images taken with a stopped down achromat which could be taken for apochromatic performance. Would this reflect actual performance? No.

And so the aim to provide at least the basis for understanding the relationship between pupils in a telescope. For all parties.

Finally, this thread has not been concerned with flagellating any particular optical system, being content to confine itself to optical concepts.

### #49 Mark Harry

Mark Harry

Fly Me to the Moon

• Posts: 6558
• Joined: 05 Sep 2005
• Loc: Northeast USA

Posted 28 August 2013 - 07:40 AM

I want it to be understood;
Glen did -NOT- post the pic I mentioned in my previous post!
Mark

### #50 Crayfordjon

Crayfordjon

Vendor - Zerochromat

• Posts: 1850
• Joined: 17 Jun 2009
• Loc: UK

Posted 28 August 2013 - 10:44 AM

Freestart8, the fault is mine, I omitted to draw in the full aperture of the exit pupil dia, the iris dia shewn is for the human iris which is typically around 2mm, whereas the exit pupil is 10mm. The diagram in not in error, but was drawn hastily, I feel that some of the more veherment critics are all to ready to knock people off their pedestals, an occupation all too familiar with those who are intellectually challenged.

## Recent Topics

 Cloudy Nights LLC Cloudy Nights Sponsor: Astronomics