Binocular Evolution II: Curvature and Distortion (1 Viewer)

[tenex]

[Part I: Field of View is here]

We have today a dramatic contrast between conventional binocular optics, with typical amounts of field curvature and pincushion distortion, and a new approach dubbed Swarovision that appeared in 2010. Often referred to as "flat field", this design (and emulations of it since) actually involve two different factors that are often confused -- with the result that discussions of it tend to correct that confusion, and stop there. Instead I want to state this distinction at the outset, and then go on to ask further questions that don't seem to get addressed.

* Field curvature. Some eyepiece designs produce less of this than others, and a larger sweet spot. Modern ones may even include a specific "field flattener" element for the purpose, giving a broader area of acceptable sharpness. Swarovision took this a step further, offering usable sharpness to the very edge of the field. (Other aberrations also had to be well managed to achieve this.) Opinions vary as to how useful this really is in practice, but it's certainly a technical achievement and a marketing point.

* Distortion. Postwar binoculars typically employ noticeable pincushioning because the alternative (barrel or even zero rectilinear distortion) produces some angular compression (AMD) near the field edge, leading to squashed shapes and a "rolling ball/globe" effect (RB) that bothers some people when panning. (Of course pincushioning itself bends straight lines, so it's a trade-off. Minimal linear distortion was also common prewar.) Swarovision tilts the balance back toward neutral, but on closer inspection the distortion profile is no longer even uniform; it varies one way and then back again as it approaches the field edge.

With all this understood, further questions arise:

1. What is such complex distortion trying to accomplish? Is it an attempt to offer the best of both worlds, reducing the bending of straight lines near the edge while also minimizing the compression effect to keep panning comfortable? Why then does its profile vary among models, even of the same line? (Not to mention, how is any of this even possible? Surely some very complex aspheric surfaces are needed in the eyepiece.)

2. How did these two things come to be done together in Swarovision? Are they really separate innovations, or is the purpose of the distortion profile somehow related to having sharper field edges? (Does that make distortion effects worse? Did the calculations required for one make the other necessary as well?)

3. Why is this approach taking over the alpha binocular market when many don't need it, and some even dislike it? The RB effect has been reduced in more recent designs, but some are still sensitive to it. Some feel that sharpness throughout gives an unnaturally "flat" view, though it's not obvious why since any binocular offers a much narrower field than normal vision to begin with, and the eye itself has only a limited central area of sharpness that doesn't even cover that. Are they perceiving something subconsciously through peripheral vision? (I'm not drawn to such designs myself and would like to know why, since I don't share these objections and can't point to anything else specific that bothers me.)

I could say more, but possibly that's enough to get a discussion started...

 

[Binocollector]

I often wondered why the pre-WWII designs managed to get a pretty flat field (as well as no pincushion distortion) in something as small and simple as my old Hensoldt 6x24 porro "Großfeld" but nowadays they need 6 lense eyepiece designs or something along those lines to achieve basically the same optical impression.
I do like a flat field and rarely ever see the rolling ball effect. But I also like my old Japanese wide angle porros. However the smaller sizes like 8x30 and 7x35 have excessive amounts of pincushion distortion.
A pic through my Hensoldt. Yes, a bit of curvature creeps in at the edges but not much. However that might in part just be because of the lower magnification and therefore large DoF.
I took another pic through the APM 6x30 and apart from the obvious differences in brightness (no coatings on the Hensoldt) the image is roughly the same (one is taken from higher up though).

 

[kimmik]

@Binocollector

Field curvature of the objective depends on focal ratio, and is made more apparent by wider FOV eyepiece unless said eyepiece corrects for it.

The Hensoldt 6x24 looking at a photograph looks like a slow focal ratio objective. similarly the APM 6x30 not only uses bigger objective, but also shorter body, so presumably a faster focal ratio.

An f/9 doublet objective telescope for example can be mated directly to camera without any flattener, as there is so little field curvature and aberration.

If one wanted f/30 (1.3m long at 42mm objective) you don't even need a doublet.

Once you want shorter lighter telescopes - aka modern roof binoculars, you run into all the 7 optic aberrations.

When you get to a certain shortness, the bino stops getting lighter then gets heavier, due to the increase in glass elements for compensation, but not saving as much in housing mechanics. From here on, the only saving is volume/space. eg NL

Moving further from the current premium designs require either new glass (i doubt), or new surfaces (most likely). Tenex’s thread on aspherics discussed the possibility of diffractive optics, polymer freeform elements, as ways to decrease weight while maintaining optical quality.

 

[tenex]

I suppose what I didn't quite manage to say by the time I split the two parts of this thread for clarity is that given all the trade-offs and constraints on optical design, I would rather see FOV made a priority than edge sharpness or complex distortion profiles which I don't particularly care about (unless further discussion somehow changes my opinion). And it's quite easy to notice too.

I wonder what customer research was conducted to make these decisions. Had EL SV been launched alongside an "SLW" model with 70° AFOV, which would have been the greater success?

 

[William Lewis]

When I first tried a swarovision binocular I really didn't like the rolling ball effect. I was much happier with my very distorted at the edges 8x32 porro (opticron srga) - see below.

More recent versions of swaro vision seem to have removed a lot of the rolling ball effect and unless I'm looking for the tricks they've played to achieve this, i.e not very straight things that are in fact straight - it doesn't bother me. Also birds aren't long and straight but it is easier to see them if they are not blurred in with a leaf. SLC 8x56 below, note it's focused on the distant trees, not the blurred ones to the left

 

[kimmik]

What a beautiful countryside William, is that walking distance from you? Sublime.

Tenex - agree with your opinion that wide is more interesting than flat. How much field curvature are you willing to tolerate, or rather is ideal for you?

I have a feeling such compact yet wide NL42/SF32/NV42 would not be good without some degree of field flattening. NV and SF are both field flattened, but not as aggressively as NL.

 

[wdc]

So what is an acceptable distortion profile for a contemporary wide(r)field roof that is not flat field? Are there any existing role models? Leica Noctivid or Ultravid? Nikon MHG?

 

[tenex]

So what is an acceptable distortion profile for a contemporary wide(r)field roof that is not flat field?

I'm satisfied with the traditional answer, a comfortable (uniform) balance between pincushioning and AMD. Leicas have more pincushioning than I'd prefer, but it's OK. I don't know of a binocular with variable distortion that's not flat field, and was asking whether there's a reason these two go together.

Tenex - agree with your opinion that wide is more interesting than flat. How much field curvature are you willing to tolerate, or rather is ideal for you?

SLCs seem ideal, with a large sweet spot dropping off only gradually. And if they would need a bit of further flattening to achieve the same impression with ~70° AFOV instead of 60°, that's my idea of what a field flattener is for.

I can understand the appeal (to some) of sharpness to the very edge, how the proposal would come up in a design meeting, and how it lends itself to marketing. But variable distortion puzzles me. I'm not sure how it came up or why it was worth such effort; the average buyer would never notice it, and the average salesman could never describe or explain it.

 

[GrampaTom]

Tenex,

This one add anything to your quest? Distortion of the new Zeiss Victory SF: A paradigmatic shift on the binocular market?. Id love to see it redone with the SF 32 and NLs

 

[tenex]

Thanks, I hadn't looked at that in a while. But here (and on the original page) Holger is talking about a single overall parameter to characterize distortion, whereas designs like Swarovision vary it radially one way and then the other. He notes this but doesn't pursue it:
"Another problem arises from the fact that the perception of the globe effect is not solely related to the k-factor, but also depending on the detailed shape of the distortion curve (from the center toward the edge), and these curves are not published by the manufacturers."

So-called mustache distortion has been discussed before here, for example: Swarovski NL 8x42 - First Impressions
These distortion profiles have evidently been refined over the last decade to reduce odd effects that they created themselves, but I still have to wonder what the point was in the first place. I'm not aware that there was popular demand to get rid of pincushioning, there being few utilitypolewatchers, so this doesn't seem an easily marketable advantage. A perfectly nice distortion balance can be struck more easily without these tricks. (Unless that somehow looks bad when the field is sharp to the edge... does a binocular exist with classic pincushioning and a flat field?)

 

[Binocollector]

For getting rid of pincushion distortion, you don't need field flatteners and I'm guessing they might be a part of the problem causing strange distortions which need to be corrected. But the field flatteners were introduced to get sharp views at the edges. At least I thought that was the reason for uneven distortion profiles?

 

[William Lewis]

What a beautiful countryside William, is that walking distance from you? Sublime.

Tenex - agree with your opinion that wide is more interesting than flat. How much field curvature are you willing to tolerate, or rather is ideal for you?

I have a feeling such compact yet wide NL42/SF32/NV42 would not be good without some degree of field flattening. NV and SF are both field flattened, but not as aggressively as NL.

The first one was when I was away camping in Wales earlier this year. The second is part of my local area me and a friend have started monitoring for the bto.

 

[dorubird]

To nuance the discussion even more, the geometric distortion in the eyes of each of us comes into the equation. That's why some people are more sensitive than others to the globe effect, because they have a slightly different perception of given geometric distortion in their eyes (aspect also mentioned in Merlitz study)

 

[Binocollector]

Very true. I myself rarely are bothered by the globe effect or don't even see it. Like in the Komz 7x30. I think that one is perfectly suited for panning the landscape. But too much pincushion distortion (which the Komz doesn't have) bothers me when scanning. Not that much or I wouldn't use my vintage wide angles but I notice it more than globe effect. I'd be more than happy if one manufacturer decided to make a bino calculated to the angular condition without pincushion distortion but with a slight amount of field curvature. Basically a modern version of something like my Hensoldt "Grossfeld" pre-WWII. So for me the combination of "field curvature + no pincushion" feels more comfy in use that "flat field + pincushion". And wasn't the reason for the new Zeiss and Swarovski designs something similar? Wide views, well corrected edges with just enough pincushion dist. to remove the globe effect for most users (which varies) and not have lines at the edge bend happily in any direction depending on which way you pan with the bino.

 

[kimmik]

Complex/second order distortion is probably not intentional, like residual secondary chromatic aberration after correcting for primary chromatic aberration.

I can find it in the EL SV 8x32, but not in the NL 8x42. Henry has also previously posted pictures of his circle test for both showing to this effect. I would like to see EL 8.5x42 and whether there is less or same amount of rolled edge distortion.

Also from his previous post of NL xray vs EL cutaway, I measured the NL field flattener at 28mm, EL at 23mm. This was the biggest optical change I could spot in the NL (several smaller changes elsewhere). Perhaps the bigger flattener element has placed the rolled edge distortion so far out that it is now removed by the field stop?

NL Pures Optically superior?

The NEW Hermann seems to be more assertive. I like it. I like it when you tell somebody they don't know what they're talking about. Most of the time it is true, but I don't have the guts to say it. You tried an NL and you didn't like, so you didn't buy one. You cut to the chase and don't offer...

www.birdforum.net

With the 7.6deg FOV of 8.5x42, a rough calculation without trigonometry: (28mm/23mm)x7.6deg=9.25deg. Assuming identical objective lenses for EL and NL, the new flattener would provide 9.2degrees of corrected field, pretty much spot on the 9.1 degree spec for 8x42. I verified this with trigonometry and the answer is the same.

If on inspection the 8.5x42 shows less rolled edge than 8x32, that would suggest to me that swaro uses the same flattener lens for the entire EL series. They were probably pushing the limit of their ability to make large field flattener at that time, and waited 11 years to make bigger flattener for the NL series.

Why should the flattener be hard to make? I hypothesize it to be the only, or one of two aspheric surfaces in the entire system.

 

[Binocollector]

Perhaps the bigger flattener element has placed the rolled edge distortion so far out that it is now removed by the field stop?

Aren't you maybe mixing two different things here? I'm not the biggest expert in this (we might ask Stepanie aka AltaVista or Bill about this, maybe Holger) but from what I know the rolling ball has nothing to do with flat field but with the absence of pincushion distortion which was the norm pre-WWII and was making some people dizzy or nauseaus when panning. Therefore the pincushion distortion was introduced which now makes the lines bend at the edges, sometimes to an annoying degree. When I pan up and down with my wide angle 7x35, the edges of the image seem to be jumping up and down. When panning from side to side it is fine of course because that was the intention of the pincushion distortion. But looking at something close to you and then at the horizon can be an annoying experience. Something that doesn't happen with binos that work according to the angular condition. And some do both -- show straight lines and no curvature, like the Komz 7x30. Or no curvature but slight pincushion, like NL Pure from what I gather, never looked through one. Or other "normal" flat field binos in my possesion that do have slight pincushion dist.
The field flattener does just remove (more or less) the field curvature.
Before the introduction of pincushion distortion you had binos that showed perfectly straight lines even at the very edge but a certain amount of curvature (as my vintage Hensoldt 6x24 proves -- perfectly straight lines but slight softening of the edges).
But I might be wrong with the details here. The field flatteners certainly introduce new kinds of abberations. The Komz 7x30 as analyzed by Holger with the help of the Swarovski lab shows different amounts of distortion in different areas of the image. Only slight but they are there. I don't see them when using it. But the amount of rectilinear distortion changes closes to the edges and then reverses. If I understood the analysis of Holger on his homepage correctly. (The formulas go way over my head.)

Case study: The distortion of the BPO 7x30

Case study: The distortion of the BPO 7x30

www.holgermerlitz.de

 

[kimmik]

One surface/lens does more than one thing. It may be marketed as field flattener, but it also does field distortion/coma/astigmatism/ca and so on. A smart design optimises the entire system with one surface.

As you've previously suggested, the field flattener is the most likely culprit for the complex distortion in EL series.

 

[William Lewis]

Aren't you maybe mixing two different things here? I'm not the biggest expert in this (we might ask Stepanie aka AltaVista or Bill about this, maybe Holger) but from what I know the rolling ball has nothing to do with flat field but with the absence of pincushion distortion which was the norm pre-WWII and was making some people dizzy or nauseaus when panning. Therefore the pincushion distortion was introduced which now makes the lines bend at the edges, sometimes to an annoying degree. When I pan up and down with my wide angle 7x35, the edges of the image seem to be jumping up and down. When panning from side to side it is fine of course because that was the intention of the pincushion distortion. But looking at something close to you and then at the horizon can be an annoying experience. Something that doesn't happen with binos that work according to the angular condition. And some do both -- show straight lines and no curvature, like the Komz 7x30. Or no curvature but slight pincushion, like NL Pure from what I gather, never looked through one. Or other "normal" flat field binos in my possesion that do have slight pincushion dist.
The field flattener does just remove (more or less) the field curvature.
Before the introduction of pincushion distortion you had binos that showed perfectly straight lines even at the very edge but a certain amount of curvature (as my vintage Hensoldt 6x24 proves -- perfectly straight lines but slight softening of the edges).
But I might be wrong with the details here. The field flatteners certainly introduce new kinds of abberations. The Komz 7x30 as analyzed by Holger with the help of the Swarovski lab shows different amounts of distortion in different areas of the image. Only slight but they are there. I don't see them when using it. But the amount of rectilinear distortion changes closes to the edges and then reverses. If I understood the analysis of Holger on his homepage correctly. (The formulas go way over my head.)

Case study: The distortion of the BPO 7x30

Case study: The distortion of the BPO 7x30

www.holgermerlitz.de

From my understanding they are different things but one is added to remove the other. The rolling ball effect is counteracted by pincushion distortion, i.e rolling ball is the opposite of pincushion. Rolling ball effectively appears to bring the centre of the image closer and the edges further away (although all can still be sharp with a field flattener) where as pincushion brings the edges closer and the centre further away.

Without a field flattener the effect is that the edges are out of focus due to field curvature so there is very little rolling ball, the blur can be focused away. So some pincushion is added when a field flattener is employed so the edges don't display the rolling ball effect but this comes with other compromises.

 

[Binocollector]

Of course. The pincushion was added to remove rolling ball effect which might not be that obvious with slightly softer edges but all of my super wide angle binos from the 60s and 70s have tons of pincushion despite having also lots of field curvature. The flat field makes the pincushion more apparent so Swarovski has been reducing the pincushion distortion for their flat field models not increasing or adding more of it. At least that's what I gathered from reading all or most the articles on that topic on Holger's website. I don't claim however that I understood them all correctly and the math is way above me

 

[Binocollector]

It seems I already wrote a lot of nonsense. The tangent-condition was the distortion free norm pre WWII. The angular condition led to pincushion.

Distortion of the new Zeiss Victory SF: A paradigmatic shift on the binocular market?