Lens physics? eg - Pana FZ200 f2.8 @ 60mm?

I struggled with where to post this, its a question about lens physics so here I am at Open Talk but I apologize if it should be in the Pana talk forum.

Anyway......

Can someone describe - feel free to dumb down - the lens physics that makes it so newsworthy that the FZ200 has a consistent max aperture from 24 to 600mm.

While I understand its a teenyweeny sensor that enables superzooms I don't understand why 1 camera can have a f2.8 all the way out vs the rest? Is there a downside to this (like the downside of a teenyweeny sensor)?

Can't anybody (Nikanon?) do this? Surely whatever it is, is reverse engineerable. Patents?

Any lens engineers our there?

Jambsi said:

I struggled with where to post this, its a question about lens physics so here I am at Open Talk but I apologize if it should be in the Pana talk forum.

Click to expand...

Well, it's better than putting it in News Discussion

Jambsi said:

Anyway......

Can someone describe - feel free to dumb down - the lens physics that makes it so newsworthy that the FZ200 has a consistent max aperture from 24 to 600mm.

While I understand its a teenyweeny sensor that enables superzooms I don't understand why 1 camera can have a f2.8 all the way out vs the rest? Is there a downside to this (like the downside of a teenyweeny sensor)?

Can't anybody (Nikanon?) do this? Surely whatever it is, is reverse engineerable. Patents?

Any lens engineers our there?

Click to expand...

I'm not a lens engineer, but one downside will be the size & weight. In order to keep the f/2.8 at 600mm, the diameter of the lens is much larger than other zooms that are not a constant aperture. The other downside is the smallish sensor rather than something larger; a bigger sensor would make the whole package much larger, heavier, and far more expensive.

Compare the FZ200 to, say the Canon SX50IS. The weight is very similar (both approx 1.3 lb) despite the Canon lens being twice as long on the far end. Then consider something with a similar zoom range but not constant aperture, like the Olympus SZ-31MR. It starts at f/3.0 on the wide end but stops down to f/6.9 at 600mm; but the weight of the camera is a mere half a pound.

For starters, look at the actual focal length of the lens. It is 24-600mm "equivalent" focal length, but the actual focal length is 4.5-108mm. So that's still a pretty impressive range and f2.8, but not anywhere like a real actual 600mm f2.8 lens.

Now the sensor is small, and the lens is attached so it can be designed to sit right at the sensor. That prohibits challenging problems like retrofocal elements that DSLR cameras have to deal with on standard/wide angle zooms. Furthermore, the tiny sensor means they don't have to make the elements any larger than the aperture to combat vignetting and edge softness, so that makes the design cheaper and smaller.

Jambsi said:

Can someone describe - feel free to dumb down - the lens physics that makes it so newsworthy that the FZ200 has a consistent max aperture from 24 to 600mm.

Click to expand...

Couple things:

1. It doesn't have a 24-600 zoom, it has a 4.5-108mm zoom. That's important.

2. It doesn't have a constant max aperture in that zoom range, it has a constant max f-number. That's also important -- if it had a constant max aperture, the f-number at the long end of the zoom would be much higher -- like most superzooms.

The f-number is given by the ratio of the focal length and aperture (technically the entrance pupil). Working back the entrance pupil of the lens at 108mm is a bit over 38mm. At the short end it's much smaller than that (at 4.5mm the entrance pupil is only 1.6mm), but that's easily managed.

Jambsi said:

Can't anybody (Nikanon?) do this? Surely whatever it is, is reverse engineerable. Patents?

Click to expand...

Sure. At 600mm and f/2.8 a zoom would require an entrance pupil of about 214mm. Not a problem if you keep a pet rhino with you to carry it around (keep in mind that mass goes as the 3rd power of linear dimension). A little tricky to fit into a shoulder bag -- but then, so's a rhino.

wfektar said:

1. It doesn't have a 24-600 zoom, it has a 4.5-108mm zoom. That's important.

Click to expand...

I get that. If I quote the actual length I get corrected with the equivalents. If I quote the equivalents I get corrected with the actuals. I get it.

wfektar said:

2. It doesn't have a constant max aperture in that zoom range, it has a constant max f-number. That's also important -- if it had a constant max aperture, the f-number at the long end of the zoom would be much higher -- like most superzooms.

Click to expand...

wfektar said:

The f-number is given by the ratio of the focal length and aperture (technically the entrance pupil). Working back the entrance pupil of the lens at 108mm is a bit over 38mm. At the short end it's much smaller than that (at 4.5mm the entrance pupil is only 1.6mm), but that's easily managed.

Click to expand...

You lost me here. You're saying that f2.8 at the long end is less light than f2.8 at the short end? If that were true wouldn't (let's say the camera is in aperture priority) the meter in the camera be reducing the shutter speed as the lens zoomed? But I see the same aperture. shutter and ISO throughout the zoom range. So, sorry but I don't get it. I would appreciate your time in taking another crack at that explanation - maybe I'm misunderstanding something really big.

wfektar said:

wfektar said:

Can't anybody (Nikanon?) do this? Surely whatever it is, is reverse engineerable. Patents?

Click to expand...

Click to expand...

Here's the crux of the original question. ALL the reviewers of the FZ200 lens say its groundbreaking. DPR used the terms "veritable engineering coup" and "defy the laws of physics". It certainly doesn't seem to be just a 'make it a little bigger' thing as compared to others in its class. This is what I'm trying to understand.

wfektar said:

Sure. At 600mm and f/2.8 a zoom would require an entrance pupil of about 214mm. Not a problem if you keep a pet rhino with you to carry it around (keep in mind that mass goes as the 3rd power of linear dimension). A little tricky to fit into a shoulder bag -- but then, so's a rhino.

Click to expand...

I get the issue of big(ger) sensor = much bigger lens for same zoom capability. This is why I own a superzoom. Is it also true that for a DSLR zoom lens the faster it is, the bigger it is? Is there such a thing as a DSLR lens with a constant f across the zoom range? I know there's no other superzooms that do that. Why not?

Thanks for sharing your knowledge and observations, I really appreciate it.

Think of it this way:

The physical aperture, is always a single size (let's say 10 or 20mm across), but optically it needs to appear as an entrance pupil to shift from 1.6mm to 38mm! In other words they had to design an physical and optical system that changed the apparent size of that aperture by 24 times (when most constant max aperture lenses might do that across a 5x range).

DSLR lenses can't because as elements scale up in size their curvature leaves less room for the kinds of movement that allows zooming at large apertures. If you want to get a better understanding, you could take apart a cheap zoom, or there was once a very nice flash based site (in french) that showed the lens diagram of the Nikkor 70-200 f/2.8 and how the elements moved to zoom and focus. If someone else remembers it, and posts it I'll bookmark it for future reference.

Jambsi said:

wfektar said:

The f-number is given by the ratio of the focal length and aperture (technically the entrance pupil). Working back the entrance pupil of the lens at 108mm is a bit over 38mm. At the short end it's much smaller than that (at 4.5mm the entrance pupil is only 1.6mm), but that's easily managed.

Click to expand...

You lost me here. You're saying that f2.8 at the long end is less light than f2.8 at the short end?

Click to expand...

f/2.8 literally stands for " the focal length of the lens divided by 2.8". So it will be much higher at the telephoto end of the zoom range than at the wide angle end.

The reason we use the f/ number notation is that the f-numbers are directly comparable across sensors of different sizes. For a given source of illumination, f/2.8 admits the same amount of light per square millimeter regardless of whether you are talking about medium format film, 35mm film, a full-frame DSLR, an APS-C DSLR, a superzoom, or a cell phone camera. But these cameras have very different sensor sizes. So the total light collected by the medium format film negative will be MUCH greater than the total light collected by the cell phone. (And assuming comparable fields of view, the medium format camera will require a MUCH longer and heavier lens, with a much wider entrance pupil.)

Jambsi said:

I get the issue of big(ger) sensor = much bigger lens for same zoom capability. This is why I own a superzoom. Is it also true that for a DSLR zoom lens the faster it is, the bigger it is?

Click to expand...

Yes. The minimum size is related to the focal length and the maximum aperture. Now if you were starting with a design that was sub-optimal (larger than it had to be) in the first place, you might be able to improve it without making it bigger ... but you get the point.

Jambsi said:

Is there such a thing as a DSLR lens with a constant f across the zoom range?

Click to expand...

Yes. Such lenses are usually heavier and more expensive than variable-maximum-aperture counterparts, but they do exist. For instance, several companies make constant f/2.8 "standard" zooms in the 17-55mm (plus or minus) range; a third-party one might cost $450+; one from Canon or Nikon, $1000+.

The reason the actual focal length is important is because it dictates the size of the aperture (and thus the glass). With the lens being 108mm f2.8, that means the aperture will be 108/2.8 = 38.6mm in diameter. That's a pretty good sized lens for a compact. Very impressive, really.

Now if you had 600mm zoom on an aps-c DSLR, that is about 400mm in actual focal length. 400mm f2.8 would be a 142mm aperture (almost 6") across. That is the minimum size of the glass, too. So you start with a salad plate for a front element and there is no way it is going to end up compact in any fashion.

Because it's not a "teenyweeny" sensor, it's a 1/2.3", which though it's much smaller than your average dSLR is still considerably larger than that of a cellphone or camcorder, and large enough that making long, fast lenses for them has traditionally been large and unwieldy. Well, that and the fact that it apparently doesn't suck like 99.9% of superzooms out there, SLR lenses included.

Look at the FZ200's grandpa for instance, the FZ50, then consider that in spite of its lens' significant size it's both shorter and slower at full telephoto. Then consider that back then the FZ50's lens was considered a small miracle in terms of reach, aperture and resolution when compared to the lenses seen until then.

Technology, 'tis such a lovely thing.
--
http://www.flickr.com/photos/Draek

Jambsi said:

If I quote the actual length I get corrected with the equivalents. If I quote the equivalents I get corrected with the actuals. I get it.

Click to expand...

Can't win, I know. So try going by context: if you're talking FOV, it might make sense to use 35mm equivalent. For this purpose, it doesn't.

Jambsi said:

Jambsi said:

2. It doesn't have a constant max aperture in that zoom range, it has a constant max f-number. That's also important -- if it had a constant max aperture, the f-number at the long end of the zoom would be much higher -- like most superzooms.

Click to expand...

Jambsi said:

The f-number is given by the ratio of the focal length and aperture (technically the entrance pupil). Working back the entrance pupil of the lens at 108mm is a bit over 38mm. At the short end it's much smaller than that (at 4.5mm the entrance pupil is only 1.6mm), but that's easily managed.

Click to expand...

You lost me here. You're saying that f2.8 at the long end is less light than f2.8 at the short end? If that were true wouldn't (let's say the camera is in aperture priority) the meter in the camera be reducing the shutter speed as the lens zoomed? But I see the same aperture. shutter and ISO throughout the zoom range. So, sorry but I don't get it. I would appreciate your time in taking another crack at that explanation - maybe I'm misunderstanding something really big.

Click to expand...

You lost me here. I'm saying that an entrance pupil that gives f/2.8 at 108mmis a lot larger than one that gives f/2.8 at 4.5mm. The ratio (the f-number) is the same in both cases.

Jambsi said:

Here's the crux of the original question. ALL the reviewers of the FZ200 lens say its groundbreaking. DPR used the terms "veritable engineering coup" and "defy the laws of physics". It certainly doesn't seem to be just a 'make it a little bigger' thing as compared to others in its class. This is what I'm trying to understand.

Click to expand...

AIUI this is the only constant f/2.8 25x superzoom. If that's true it would definitely be unprecedented and "engineering coup" might well be a fair way to put it. As for "defy the laws of physics" -- maybe it would be better if the DPR reviewer explained that to you. Because I can't.

Jambsi said:

I get the issue of big(ger) sensor = much bigger lens for same zoom capability. This is why I own a superzoom. Is it also true that for a DSLR zoom lens the faster it is, the bigger it is? Is there such a thing as a DSLR lens with a constant f across the zoom range? I know there's no other superzooms that do that. Why not?

Click to expand...

Sure, but they're over much smaller zoom ranges (70-200, 17-50, etc). I can't think of anything like a 25x superzoom in this form factor, constant f-number or not. The only thing I can think of to approach that would be the various 18-270s. Those aren't constant f-number though.

Why no superzooms comes down to size, whether constant f-number or not. It's basically the same reason why most people don't use medium format for wildlife or birding: to get a sufficiently narrow FOV the lenses would be immense.

Here's a concrete example: to get 600mm at f/2.8 you need an entrance pupil of 214mm. The only 35mm lens I know with a refractive element that large is the Nikkor 1200-1700 zoom from the 90s. That one is 88cm long and weighs 16kg. You can see it in action here: http://blogs.reuters.com/photographers-blog/2007/07/18/unleashing-the-beast/

Details about the lens can be found here: http://www.mir.com.my/rb/photography/companies/nikon/nikkoresources/zoomsMF/12001700mm.htm

Now maybe current technology can take a couple kg and a few cm off that today, but 1) it's a 1.4x zoom, not a 25x; and 2) it's not constant f-number. A constant f-number (if it's reasonably fast) 25x superzoom would be ... well, I thought I was only kidding about that rhino...

BTW, there is a fast MF supertelephoto, the Zeiss 1700mm f/4. A mere 256 kg in weight, it dwarfs the big Nikkor. But it's not even a 1.4x zoom! See it here: http://www.dpreview.com/news/2006/10/1/zeiss1700f4

I feel like I just learned a bucketful. Lens physics is complicated. Thanks all for your sharing & time.

1) Think about when it makes sense to use equivalent or actual focal length. When giving the equivalent focal length, make this clear - do not just give the numbers. This way you avoid most, if not all complaints.

2) The f number is a ratio. It indicates the diameter (size) of the entrance pupil compared to focal length. Therefor longer focal lengths will have larger entrance pupils for the same f number.

3) Because of the above, making a fixed f-number zoom lens is "harder" than making a variable f-number lens because the lens must be bigger (bigger entrance pupil) and because more advanced optical designs must be used.

3)

• Making a quality fast (small f-number) lens is harder than making a quality slow lens.
• Making a quality zoom lens is harder than making a quality fixed focal length lens.
• Making a quality zoom lens that covers a very big zoom range is even harder

Combining all three of the above, while keeping size and weight small, requires very good engineering.

If you go back and look at the specifications for early zoom lenses in the late 60's; you'll find a constant aperture was common. As time went on zoom lenses began to have variable apertures to get smaller, lighter, and easier to design and manufacture lenses. This trend has continued to the present, particularly for DSLR's

In the 60's and early 70's a fixed aperture lens made life a lot easier for the photographer as the cameras didn't yet have program exposure mode or TTL flash; both of which compensate for the change in aperture. The older aperture priority cameras and sensor flash units both work a lot easier if the lens aperture doesn't change when you zoom.

The emergence of Program auto cameras and TTL flash freed the lens designers and made variable aperture zooms a lot more palatable for the photographer - I know it did for me.

--
Jerry