Retinal Offset and Field-of-View Effects on Apparent Size Using Binoculars (1 Viewer)

Retinal Offset and Field-of-View Effects on Apparent Size and Depth Using Binoculars

Several threads have discussed perceptual effects of porro vs. roof prism binoculars. The common observation is that porros induce smaller perceived images and greater spatial depth. Unfortunately, in addition to the greater retinal disparity provided by porros, another variable often distinguishes them from roofs, namely field-of-view (FOV). Generally, (but not always) porros have a larger field. Although several observers have commented that FOV has no effect on these phenomena, my own opinions have been more guarded. This is primarily because, as an experimental psychologist, I know that virtually nothing has a null effect on visual experience. The issue, to me, is not so much whether it has an influence, but how discernable is it? The problem to this point was that I didn't have an easy way to separate the two variables, namely retinal disparity and FOV.

Fortunately, I was recently able to acquire a dated (c. 1980) Bushnell 8x30 "Broadfield Compact" roof prism with a FOV of 8.5 deg. This exactly matches the 8.5 deg. FOV spec. of my Nippon Kogaku (Nikon) "Mikron" (c.1960). The latter is what I had used to verify the smaller image and greater depth of a porro when compared with my 7.8 deg. FOV 8x30 SLC Mk II (c. 1993). Optically the Bushnell is not in the same league as either the Nikon or Swaro. Not being phase coated it presents a darker view, but does appear to have coated glass and BAK-4 prisms. Bushnell was kind enough to clean and align it on a complementary basis, for which I am most grateful. My less than perfect measurements with a ruler indicate that the 30 mm objectives and 3.75 mm exit pupils matched sufficiently to conduct perception trials. (The critic will note that the design is incomplete, since a narrow field porro is missing. This would have involved finding an 8x30 porro with a 7.8 deg. FOV to match the Swaro, however, which was beyond my level of dedication.)

It took me some time to settle on procedures, but in the end I decided on viewing a scene centered at 60 yd. (180 ft.). Admittedly, this was partly because I could do the work while seated at my desk, looking at a house and driveway diagonally across the street. Several natural and man-made objects were conveniently located so that shading, perspective, and color contrast cues were all present, and an expanse of lawn allowed for assessing depth. The view spanned a range, near to far, of approximately 25—75 yd, with binoculars focused on objects at 60 yd.

In order to assess relative perceived size and spatial depth, the scenes were viewed both binocularly and monocularly. Since the order of comparison might be expected to influence the findings, trials were conducted by first looking at the scene through the SLC for 5 sec. of mental adaptation, and then switching to either the Nikon or Bushnell. Judgments of size and depth change were based on my immediate perceptions after the switch, and not after a studied view. (This is important to note for others attempting the experiment.) A second and third set of comparisons was made looking through the Bushnell and Nikons first and then switching to the others. In each comparison sequence one would expect the Porro images to appear smaller and have greater depth than either roof, and the two roofs to appear of equal size and depth compared to each other — that is, if FOV had no effect.


60 Yard Results (binocular viewing). Results are numbered with b = binocular.

1b. Looking at the scene first with the Swaro (roof) and then switching to the Nikon (porro), there was a distinct decrease in object size and increase of depth. This is consistent with earlier BF discussions.
2b. Switching to the Bushnell (roof) from the Swaro, however, there was almost an equal decrease in size and somewhat smaller increase in depth.
3b. Looking at the scene first with the Bushnell (roof), and then switching to the Nikon (porro) there was a slight decrease in size (smaller than #1b), and some increase in depth (also smaller than #1b).
4b. Switching to the Swaro from the Bushnell, there was a modest size increase in size and slight decrease in depth.
5b. Switching to the Swaro from the Nikon, there was a substantial increase in size and decrease in depth (basically the inverse of #1b).
6b. Switching to the Bushnell from the Nikon there was a small increase in size and decrease in depth.

These trials were repeated several times, but the alignment of the Bushnell was still not right and I was fighting off headaches. However,

• The wide FOV roof prism binocular produced similar size and depth perceptions as the porro having the same FOV.
• The narrow FOV roof produced greater size and less depth than a wide-field roof binocular.

Conclusion 1: Retinal offset and FOV jointly influence size and depth perceptions using binoculars at this distance.

Method Note: It is somewhat easier to perceive a decrease in size than an increase, and an increase in depth than a decrease.


60 Yard Results (monocular viewing).

Monocular viewing should put all binoculars on an equal footing if retinal offset accounts for all the perceptual size and depth effects. In this case I only used my dominant left eye, but the right eye matches pretty closely. Results are numbered with m = monocular.

1m. Looking first with the Swaro (roof) and then switching to the Nikon (porro), there was a distinct decrease in object size and moderate increase in depth. Yup, even with one eye.
2m. Switching to the Bushnell (roof) from the Swaro, there was almost an equal decrease in size and moderate increase in depth (slightly more than #1b actually).
3m. Looking with the Bushnell (roof), and then switching to the Nikon (porro) little apparent change in size or depth.
4m. Switching to the Swaro from the Bushnell, both roofs, there was a moderate size increase and small decrease in depth.
5m. Switching to the Swaro from the Nikon there was a moderate size increase and small depth increase.
6m. Switching to the Bushnell from the Nikon there was no clear change in size or depth.

Some of these perceptions were very subtle, like 3m and 6m, however:

Conclusion 2. Size and depth clearly change as a function of FOV difference between the binoculars.

Additional trials at 60 yd. (winking). Results numbered w = winking.

After doing the monocular trials I couldn't resist winking trials, or opening and closing the second eye. A brief period of adaptation is required for each binocular that takes 5 sec. or so to accomplish.

1w. Swaro. Closing one eye diminishes apparent size and depth. Opening both eyes does the opposite.
2w. Bushnell. (Images seem smaller than Swaro.) Nonetheless, closing one eye decreases apparent size and depth. Opening both eyes does the opposite.
3w. Nikon. (Images seem close in size to Bushnell, with greater depth.) Closing one eye decreases size and more strongly decreases depth than the Bushnell. Opening both eyes does the opposite.

The perceptual size effect for the Bushnell was similar to the Nikon’s, but the winking depth effect was somewhat stronger for the Nikon. Once adapted to the Swaro's larger apparent size, the winking effects were similar to the Bushnell’s.

Method note: One problem with doing winking experiments is that the combined visual fields change along with the image. There is nothing one can do about this since it results from the geometry of the head and eyes. Pure binocular or monocular procedures are better, IMO.

If anyone has had the fortitude to read this far, it should be evident that both the retinal offset and field-of-view has an effect on perceived image size and spatial depth using binoculars. Frankly, based on these observations I’m not sure which is more dominant at this viewing distance. The story might be different closer than 40 yd., so I’ll look at that next. However, the human perceptual system is regular as rain — very complicated!

Your comments would be appreciated. Yelling and screaming will not be tolerated (very well). It’s been fun though.

Best regards,
Elkcub
PS. Picture of Bushnell 8x30/8.5 "Broadfield" (c. 1980) shown.

Good for you to have that much patience for conducting such tests.

Elk,

Good work!

Even with FOV influencing apparent size, is there any difference in the perception of detail?
It looks bigger or smaller but do I see the detail less or more? If so, is the difference significant?

Happy bird watching,
Arthur Pinewood

H

Pinewood said:

Elk,

Good work!

Even with FOV influencing apparent size, is there any difference in the perception of detail?
It looks bigger or smaller but do I see the detail less or more? If so, is the difference significant?

Happy bird watching,
Arthur Pinewood

H

Click to expand...

Hi Arthur,

My main interest here was driven by curiosity about FOV effects on apparent size and depth. I don't think it has too much impact on apprehending visual detail.

In and of itself I'd expect that detail perception is primarily determined by retinal image size, when corrected for brightness with larger objectives to keep the EP constant. If not, there are complicated tradeoffs. This business about smaller retinal images conveying the same amount of detail, IMO, is either deviously semantic, or wishful thinking, since under most conditions apprehension is not as easy or rapid with a smaller target image. That translates to visual strain for me, or the inability to appreciate the optical detail at all.

I'm thinking about doing a few studies of visual apprehension, similar to what Kimmo suggested on another thread. Until then my 10x42 SLC does everything quicker and easier than my 8x30 SLC, and I believe it's primarily the image size not the somewhat greater brightness that makes the difference. For the lower magnifications like these 8x30s, my impression is that the FOV makes the binoculars more "general purpose," and in that regard I'd rather have an EII than an SE, and I've looked through both at different times. o

Regards,
-elk

lvn600 said:

Good for you to have that much patience for conducting such tests.

Click to expand...

Thanks. I think the issue could use a bit more systematic evaluation. We are not seeing much use of wide-field binoculars nowadays, certainly not roofs like those venerable Bushnell Broadfields. A quality roof of that sort might just add some real spice to the mix.

-elk

Elk,

Thanks for an interesting experiment. A thought occured while reading it about a possible way of expanding it. With some black card stock, calipers and a sharp knife (and a bit of patience), you should be able to make some external eyepiece field-stops for your wide-angle porroprism binocular, which would allow you to vary the field of view of one and the same binocular. Thus you could eliminate more of the variables. Just an idea.

Kimmo

kabsetz said:

Elk,

Thanks for an interesting experiment. A thought occured while reading it about a possible way of expanding it. With some black card stock, calipers and a sharp knife (and a bit of patience), you should be able to make some external eyepiece field-stops for your wide-angle porroprism binocular, which would allow you to vary the field of view of one and the same binocular. Thus you could eliminate more of the variables. Just an idea.

Kimmo

Click to expand...

Hi Kimmo,

I was thinking about a similar thing for use outside the objectives, but that would only stop them down for differential illumination studies.

How would an "external" field-stop be used? Would I not have to open the binoculars to insert them? I'm missing something here. :h?:

-elk

elk,

Along the same lines as Kimmo's suggestion you might try a longer eyecup on one side of the same binocular in the monocular tests. Just a simple paper tube would work to hold the eye at a longer distance from the eyepiece effectively reducing the field width of that side compared to the other. An interesting experiment. I've done similar things, but not quite so systematically. I'll try to duplicate some of your results later today.

Henry

Henry,

Oh, I see what Kimmo was refering to. I'll give that a go.

Probably the key observations to replicate are 2b and 2m, namely going from narrow to wide roof, and back again, 5b and 5m. Note that it is best to adapt to binocular A for 5+ sec. and then record your immediate impression after switching to B.

Arthur Pinewood will try to compare his two Nikon porros, which have a similar large difference in FOV. I have a prediction on the outcome, but would rather wait to see what he or you find.

Thanks,
-elk

elk,

Yesterday I tried several experiments to try to duplicate your results. I confess I expected to confirm my view that FOV has no effect on perceived image size. Instead, I saw something surprising and learned something new; a much better outcome.

For monocular tests I made a special pair of monoculars, identical except for FOV. I removed one eyepiece from a Nikon 9X35 E (home made by combining 8X30 eyepieces and 7X35 objectives) and substituted a smaller fieldstop which reduced the AFOV of that side from 66.4 degrees to about 54 degrees. I used two other binoculars for both monocular and binocular comparisions: a Nikon 8-16X40 XL zoom and a Zeiss 8X42 FL. I picked the Nikon zoom because it has an extremely narrow 42 degree AFOV at 8X compared to 62 degrees in the Zeiss which is the only 8X roof I own.

The target I picked for judging apparent size was a blue bird box at about 130'. In monocular tests I could see no difference in the apparent size of the box between the two sides of the 9X35 and no difference between the Nikon zoom and the Zeiss. My previously held views apparently confirmed. Afterward it occured to me that perhaps the object I picked was not random. I think years of messing about with this sort of thing has taught me that objects of a certain angular size are the best for making accurate size judgements so now I automatically choose an object that spans about 4-5 degrees of AFOV. I walked closer to the blue bird box until it spanned about 30 degrees of AFOV. Now an illusion of size difference was visible so that the box appeared larger in the smaller FOV. The difference was rather small in the two sides of the 9X35 and more pronounced between the Nikon zoom (through which the box almost filled the field) and the Zeiss. It appears that two factors contribute to the strength of this illusion: the angular size of an object in the field and how close to the edge of the smaller field that object approaches. I also beleive there is a threshold of angular size for an object below which the illusion no longer occurs.

When I switched to binocular vision and compared the Nikon zoom (132mm objective spacing) and the Zeiss (67mm objective spacing) at this distance from the box (about 25') I found the parallax effects of objective spacing completely overwelmed the FOV effect. Now the image in the Nikon appeared tiny compared to the Zeiss. In fact equalizing the apparent image sizes required increasing the magnification of the Nikon to about 9.5-10X. Then I returned to the original 130' distance and found the the image of the box in the Zeiss still appeared larger than the Nikon with binocular vision. At this distance it required increasing the Nikon's magnification to 8.5-9X to equalize the apparent sizes. It appears that the angular size of an object makes no difference in the illusion created by objective spacing.

In reading through your results I found myself wondering whether your Swarovski might actually have a little higher magnification than the Bushnell or Nikon. That would explain why it always produced the larger appearing image just as well as the difference in FOV. Unfortunately I've never found a way to accurately measure the true magnification of a binocular, especially given that the magnification varies across the field. In my experience magnification differences of 2% or a bit less are visible.

Henry

henry link said:

... In reading through your results I found myself wondering whether your Swarovski might actually have a little higher magnification than the Bushnell or Nikon. That would explain why it always produced the larger appearing image just as well as the difference in FOV. Unfortunately I've never found a way to accurately measure the true magnification of a binocular, especially given that the magnification varies across the field. In my experience magnification differences of 2% or a bit less are visible.

Henry

Click to expand...

Henry,

Very clever and interesting observations, although done a bit differently than mine in several ways, which might account for disparate results. I'll need to think about it for a while and get back with you.

In the meantime I've included a quick snapshot of the scene I was looking at two days ago, as well as the cropped area that roughly corresponds with that seen by the binocs and used to make size/depth judgments. Note that the large wood cabinet is about 4'x4'x1" and the cardboard box has lettering just visible at my viewing distance.

Those objects have changed location today, but the cinder blocks are still there. The view through my Bushnell (wide) yields a distinct impression of smaller objects than through the Swaro (narrow), and only slightly larger than the Nikon (wide) if anything. Going from Bushnell to Nikon there is primarily a difference in perceived depth, not object size. I can't rule out that the Swaro actually has an 8+ magnification, but that wouldn't account for the Bushnell vs. Nikon comparisions — two binocs that share the same FOV but differ in parallax.

-elk

...When I switched to binocular vision and compared the Nikon zoom (132mm objective spacing) and the Zeiss (67mm objective spacing) at this distance from the box (about 25') I found the parallax effects of objective spacing completely overwelmed the FOV effect. Now the image in the Nikon appeared tiny compared to the Zeiss. In fact equalizing the apparent image sizes required increasing the magnification of the Nikon to about 9.5-10X. Then I returned to the original 130' distance and found the the image of the box in the Zeiss still appeared larger than the Nikon with binocular vision. At this distance it required increasing the Nikon's magnification to 8.5-9X to equalize the apparent sizes. It appears that the angular size of an object makes no difference in the illusion created by objective spacing.

Click to expand...

Henry,

As mentioned in post #1, I was also going to make evaluations closer than 40 yd. To better correspond with your procedure, I selected the view of my goldfinch feeder at approx. 50 ft. (see photo). [I made a typo in original post saying 25 ft.]

I didn't go through the whole sequence again, but I'm happy to report that we basically get similar results at shorter distance. With two eyes my Nikon generates distinctly smaller apparent size than the Bushnell, which has the same FOV. As impressive to me is the difference in 3-D effect, with the Bushnell disagreeably flat. (Wish I could could modify the FOV on my Nikon, since the tube idea wasn't practicable.) The Swaro appears much larger than the Nikon, with two eyes, but only marginally larger than the Bushnell. Using one eye, the Nikon and Bushnell produce the same object size, but the Swaro is somewhat larger. I can't discount a real magnification difference, of course, but my impression is that the FOV has a small effect even at close range.

What this adds up to me is that viewing distance is an important consideration, which is not really surprising. In fact it's a deja vu. Within 50 ft. the human is known to be able to employ triangulation for distance estimation, whereas this ability falls off quickly at greater range. By 180 ft the cues associated with FOV are more salient. My rationale for intermediate distance scenes appearing smaller with a wide ange binocular is because greater scene width is a cue that we are further away—hence, objects would be interpreted as smaller.

I'm not sure how much further we need to go with this, unless you find very discrepant results at increasing distance. I'd have to say that the industry is probably going in the right direction by not emphasizing really wide angle roofs, although in the case of porros I think it's a major asset right along with retinal disparity. This experiment has again brought me to appreciate the "wow" of the full view porro. I wonder what would happen to the SE if they widened the field stop — uh, maybe find that it's really a degraded "E."

Questions/comments. I'm not doubting what you observed. However, in your binocular experiment were the Nikon zoom and Zeiss FL matched for FOV? Then, when you matched image size with the Nikon zoom (cute concept), it seems to me the FOV must have diminished as power increased. That the FL still appeared larger at 130' than the Nikon with adjusted power and diminished field is tough to understand. (Are you sure the Zeiss doesn't have a larger real image, BTW.)

Anyway, my current opinion is that the dynamics of apparent size and depth change as a (continuous) function of observation distance. In the case of binoculars, for various reasons I would also suspect that these perceptual effects change with power — which would, incidentally, complicate the results using a zoom.

-elk

Friends,

I have been away a few days and won't have time to read and digest this thread until a little later due to a busy schedule, but two comments in the meantime.

Firstly, to Elk: my idea about the fieldstops was basically making a pair of black cardboard "washers" which you would place right on top of the eyelenses. If the diameter of the eyelens is X, and the diameter of the aperture of the washer is 0.8 times X, you would end up with an apparent field entering the eye which would also be 0.8 times the original apparent field. No? I think this would be much easier to maintain than a constant excessively large distance from the optimum exit pupil.

Henry: I have measured relative magnifications with the aid of a booster, under the assumption that by measuring the diameter of the viewfield of the booster/binocular A (or telescope) combination at a fixed distance, and comparing this to the result for booster/binocular B at the exact same distance will give a rather accurate reading of the ratio of magnifications between A and B.

Kimmo

kabsetz said:

Friends,

I have been away a few days and won't have time to read and digest this thread until a little later due to a busy schedule, but two comments in the meantime.

Firstly, to Elk: my idea about the fieldstops was basically making a pair of black cardboard "washers" which you would place right on top of the eyelenses. If the diameter of the eyelens is X, and the diameter of the aperture of the washer is 0.8 times X, you would end up with an apparent field entering the eye which would also be 0.8 times the original apparent field. No? I think this would be much easier to maintain than a constant excessively large distance from the optimum exit pupil.

Henry: I have measured relative magnifications with the aid of a booster, under the assumption that by measuring the diameter of the viewfield of the booster/binocular A (or telescope) combination at a fixed distance, and comparing this to the result for booster/binocular B at the exact same distance will give a rather accurate reading of the ratio of magnifications between A and B.

Kimmo

Click to expand...

Kimmo/Henry,

Your idea works like a charm; the FOV is reduced using this technique. But, listen to this. As soon as you mentioned "washers" I went out in the garage and located a bunch of old neoprene faucet washers from the olden days. Lo and behold two immediately appeared (one red, one blue) that are 17mm OD and 5mm ID and 3mm deep [labeled "3/8L" for those interested]. As providence would have it the eyelens on the Bushnell is exactly 17mm, and, of course, the 3.75mm exit pupil is centered. Having a slight taper on the outside, the washers literally snap into place. I can't believe it! They stay in place as I use the binoculars. I'll start looking for washers to fit the Nikons next.

So, at the moment I have a roof binoc with either a very wide or very narrow FOV. Based on what you said the ratio is 5/17 = .294, hence the FOV = .294 x 466 = 131.25. Is that correct? In any event, either one-eyed or two, the feeder mentioned above at 50 ft. (and another at 75 ft.) on a preliminary basis appears somewhat larger with ultra-narrow view and smaller with wide view. I'll have to evaluate that more systematically. As you correctly suggest this equipment mod allows several parameters to be held constant while varying only FOV.

I must admit, amigos, that some of these perceptions are subtle and easily biased by my own mental set. I'd like to get a helper or two to make judgments with the same apparatus without having any preconceptions.

Many thanks,
-elk
PS. The eyelens on my Swaro is also 17mm. This is too much. So the Swaro's smaller FOV would be .294 x 408 = 119.95. The two roofs retain a FOV ratio of .914 with external stops or without. Now, dare I take chances messing up the lens coatings on my precious one to do this ... ??

Wash the washers carefully, and being soft, they should be okay with your Swaro.

If you wish to verify the exact modified field of view, measure both unmodified and modified fields by viewing a measuring tape at a set distance.

Kimmo

kabsetz said:

Wash the washers carefully, and being soft, they should be okay with your Swaro.

If you wish to verify the exact modified field of view, measure both unmodified and modified fields by viewing a measuring tape at a set distance.

Kimmo

Click to expand...

Right, but when I got all the plumber's grease off they didn't stick as well.

Thanks for the tip on measuring the FOV. I'll do it!

Does the light from the eyepiece form a cone to the optical exit pupil that is really located on a plane behind it? If so, I assume the distance of this plane from the eyelens is the eye-relief of the binoculars. No?

Next question. As one draws away from the EP's surface does the magnification increase as well as the FOV decrease?

Next question(s). Does it make sense that the image would be much more crisp with the washers in place? It seems to be so. Is it because we're using the center of one or more lenses, or having fewer off-axis rays?

Next question. Does the magnification of one's eyeglasses add to the magnification of the binoculars?

I used to drive my teachers insane with questions. Sorry. But, they were relieved to see me graduate.

-elk
PS. The All-American Bushnell decked out for the 4th of July!

First of all thanks for a very interesting series of experiments. Now you can try out different fields of view with the same binocular, do you notice any difference in apparent brightness?

Elk,

I've continued to play with this, using Kimmo's washers and borrowing a pair of old B&L 8X24 reverse porros to maximize differences in objective spacing. The short result is that in monocular tests I continue to see no effect on apparent image size from FOV if the target object is small (5 degrees of AFOV), and only a small effect if the target object is large.

On the other hand, in binocular tests the illusion of size differences for any size object from parallax can be amazingly large and also extend to unexpectedly long distances. In the extreme case of the B&L 8X24 (37mm objective spacing) vs the Nikon zoom (132mm spacing) the apparent size difference at 15' was so large that equalizing the image sizes required increasing the Nikon's magnification to almost 11X. Back at 130' the Nikon required about 9.5X. I was still able to see a little difference in the size of a small object at a distance I estimate to have been 500-600' requiring an increase in the Nikon to perharps 8.2-8.3X. In monocular tests I could see no difference at all in magnification between the B&L and the Nikon at 8X. The field widths (Nikon 42 degrees, B&L 50 degrees) should have favored a larger apparent image in the Nikon from the influence of FOV. So all the difference I saw would have been from parallax, indicating that when it comes to this illusion the eye/brain is sensitive to very tiny changes in the lines of sight of the eyes.

P.S. I just noticed your addition to an earlier post. When magnification is increased with the Nikon zoom the real field shrinks, but the apparent field widens in partial compensation. I made my size estimates using small objects that I had found were not affected by FOV in monocular tests.

Henry

henry link said:

Unfortunately I've never found a way to accurately measure the true magnification of a binocular, especially given that the magnification varies across the field. In my experience magnification differences of 2% or a bit less are visible. Henry

Click to expand...

Henry and elk,

I hope this doesn't sound off the mark, as it is not a way to measure magnification. However it is a simple way to detect whether there are magnification differences between two binoculars, and I have done it to compare the relative brightness of binoculars in low-light situations, as well as to compare color casts. I hold both binoculars to my eyes simultaneously -- one barrel of each up to an eye. It takes awhile to adjust them, rarely more than thirty seconds, but it is relatively easy to line up the images and even superimpose them. The shapes of some binoculars cause them to interefere with each other and make the comparison almost impossible, but usually if at least one is a roof-prism model, it will work.

Jonathan,

Thanks for the suggestion. I've done this too, and did it for the binoculars I tested on this tread. I don't know how wide the tolerance is, but the eyes seem to be able to successfully merge images of slightly different magnifications. In this case I tried to place the images of the blue bird box from the two binoculars right next to each other. With both images swimming around I find it hard to make really accurate judgements about relative size. I think another problem is that distortion will effect the apparent size of objects differently depending on how much large the objects are and where they are located in the field. For instance, binocular with barrel distortion will have maximum magnification at the center of the field while a binocular with pincushion distortion will have minimum magnification at the center.

A problem for me with using this technique for judging brightness and color cast is that I often seem to see a slightly brighter image from one eye than the other (usually the more rested eye) and I always see a different color cast, warm on the left and cool on the right.

Henry