Are faster lenses really that hard to make?

[Shakespeare]

Often, it seems that the faster a lens gets, the more expensive it gets. For instance, if you go from the 70-200 f/4 to the f/2.8, it doubles the price.

But the 50mm f/1.8 is an $80 lens, and many people like this lense very much.

So how does speed factor into price? Are faster lenses that much harder to build, and if so how does the 50mm figure in?

Just a curious newbie...

S

[neo_xeno]

i think it's all in the glass. we still haven't found a cheap way to make good optics so the price is expensive. so the faster the lens the bigger the glass has to be to let in more light.

[cmM]

faster means larger in diameter normally. Larger elements are more expensive to make than smaller ones, obviously. But there are many other factors involved, like how many elements - the more elements the more money, build - the 50mm is all plastic, including mount, and many other factors that I don't know about :D

[CyberDyneSystems]

It has to do with the field of view..

A 50mm lens is "normal" (well acutally I think it's like 54mm) or exactly what the eye sees. It does not imagnify or reduce,.. it is not telephoto or wide,.. It is not trying to change the image at all.

This means that the glass is significanlty easier and less costly to produce.. fewer elements etc..


This also means that to get a good image at this focal length.. less elements and engineering are required to overcome the effects that wide and telephotos have to cope with.

It means that to get boatloads of light into it.. you don't need as large a front element as other lenses do to maintain the light level as the image is mashed through all the other elements that for instance a telephoto would need.


You'll note that the closer to 50mm a lens is in focal lenght.. the smaller it tends to be.. and the farther from 50mm a lenses focal lenght (think 600mm) the larger the lens has to be.

So the 50mm 1.8 is no marvel of Canon engineering.. (well it is.. but it it's not uniquley Canon) ...it is merely the way that the optics work that lends itself to being both dirt cheap and super sharp. Nikon has one too..

:)

[karusel]

As it really is not possible to logically link this:

A 50mm lens is "normal" (well acutally I think it's like 54mm) or exactly what the eye sees. It does not imagnify or reduce,.. it is not telephoto or wide,.. It is not trying to change the image at all.

This means that the glass is significanlty easier and less costly to produce.. fewer elements etc..

it raises a few questions. For instance, is the human eye, or should I say, the 'physics' of the eye (like angle of view), ideal? And I'm sort of unsure if the 50mm lens is easiest to make, if it was, I would die for an answer to - why.

Also, don't forget, that the film or sensor is flat, while biological light receptors lie on a curved plane, an inward sphere, so that every ray of light hits the receptor at precisely 90°, whereas sensor is not that lucky which naturally causes distortions and I have yet to learn how engineers are able to overcome that.

[samdring]

All this will alter, I suppose, when ceramic lenses complete their development which is said to be imminent.

[PaulB]

The 'ideal' standard lens on any camera has a focal length equal to the diagonal of the negative/sensor size.
On a 35mm film or FF DSLR this equates to 43.266615305567871517430655209646mm focal length - call it 45mm for ease of use.
The 50mm 'standard' lens came about when Oskar Barnard was thinking up the original Leica design. For the compactness Barnardwanted the obvious film stock to use was existing 35mm cine film - which actually gave an image of 24mm x 18mm (what was to become 1/2 frame in later years in the stills world!) but this image size was too small given the poor emulsions at the time and the wrong way up so the obvious was to use twice the area = 36mm x 24mm.
The original Leica (the first 35mm camera) used a 50mm focal length lens because it was close to the optimum of 45mm, designs existed which could be easilly adapted and were relatively cheap to manufacture; which still applies today which is why most 50mm 'Standard' lenses of modest aperture are cheap and often stunning performers. It is much more difficult and costly to design and make a good wider-angle than 50mm although some manufacturers did (do - Pentax) offer 45 or 48mm 'Standard' lenses.
At the telephoto end the 135mm proved to be the easiest to make a good performer and for many years was the longest lens you could buy for Leica/Contax etc without resorting to add-on reflex finders. This is one reason why the Canon 135/2L is such a stunning performer.

[gcogger]

It has to do with the field of view..

A 50mm lens is "normal" (well acutally I think it's like 54mm) or exactly what the eye sees. It does not imagnify or reduce,.. it is not telephoto or wide,.. It is not trying to change the image at all.

This means that the glass is significanlty easier and less costly to produce.. fewer elements etc..


This also means that to get a good image at this focal length.. less elements and engineering are required to overcome the effects that wide and telephotos have to cope with.

It means that to get boatloads of light into it.. you don't need as large a front element as other lenses do to maintain the light level as the image is mashed through all the other elements that for instance a telephoto would need.


You'll note that the closer to 50mm a lens is in focal lenght.. the smaller it tends to be.. and the farther from 50mm a lenses focal lenght (think 600mm) the larger the lens has to be.

So the 50mm 1.8 is no marvel of Canon engineering.. (well it is.. but it it's not uniquley Canon) ...it is merely the way that the optics work that lends itself to being both dirt cheap and super sharp. Nikon has one too..

:)

I've heard this stated many times, but it's never made any sense to me :o

A 50mm lens does not have the same FOV as my eye, since that is nearly 180 degrees
It's also not a question of having the 'same pespective', since perspective is not affected by focal length (just the distance to the image).
I don't think that it's to do with what you see through the viewfinder since I believe (correct me if I'm wrong!) that different cameras have different apparent maginification through the viewfinder. But even if we are talking about the view through the viewfinder, that is irrelevant to the final image.

All I can think is that this is one of those things like depth of field - it only relates to a print of a certain size viewed at a certain distance. If that's the case, it means the 50mm focal length is not so special after all!

Any thoughts?

[robertwgross]

...
All I can think is that this is one of those things like depth of field - it only relates to a print of a certain size viewed at a certain distance.
...


That is a novel concept you have there, but it doesn't make any sense at all.

Depth of field really has little to do with prints or print sizes or view distance.

---Bob Gross---

[CoolToolGuy]

Thank you, PaulB - I knew the 'normal' lens for a 35mm frame was less than 50mm, but I could not easily find an explanation on the Web. Thanks for providing that illumination.

[CyberDyneSystems]

Typo.. said "I think it's like 54mm .. meant like 45mm.. either way.. I was off a few mm :wink:

[mdude85]

As it really is not possible to logically link this:

A 50mm lens is "normal" (well acutally I think it's like 54mm) or exactly what the eye sees. It does not imagnify or reduce,.. it is not telephoto or wide,.. It is not trying to change the image at all.

This means that the glass is significanlty easier and less costly to produce.. fewer elements etc..

it raises a few questions. For instance, is the human eye, or should I say, the 'physics' of the eye (like angle of view), ideal? And I'm sort of unsure if the 50mm lens is easiest to make, if it was, I would die for an answer to - why.

Also, don't forget, that the film or sensor is flat, while biological light receptors lie on a curved plane, an inward sphere, so that every ray of light hits the receptor at precisely 90°, whereas sensor is not that lucky which naturally causes distortions and I have yet to learn how engineers are able to overcome that.

I believe they overcome that partly by developing curved glass.

[iwatkins]

Of course, a zoom lens is also much more difficult to make over the 50mm fixed focal length.

Cheers

Ian

[gcogger]

...
All I can think is that this is one of those things like depth of field - it only relates to a print of a certain size viewed at a certain distance.
...


That is a novel concept you have there, but it doesn't make any sense at all.

Depth of field really has little to do with prints or print sizes or view distance.

---Bob Gross---

It has everything to do with them! Normal depth of field calculations are based on the visibility of the slight blurring of out of focus areas when viewing a 10x8 print at some specified viewing distance (I forget the actual distance!).

It's not a novel concept - it's not even my concept. It's simply how the calculation is performed.

References:
http://www.photo.net/learn/optics/dofdigital/
http://www.lightandland.co.uk/articles_details.asp?TopicID=6&ID=60
http://www.nikonlinks.com/unklbil/dof.htm

[gcogger]

OK - I know it's bad form answering my own post :) But if anyone is interested in what is special about the 50mm lens...

I just noticed a discussion on the dpreview forums as to what constitutes a 'normal' lens. A detailed calculation is shown here:
http://forums.dpreview.com/forums/read.asp?forum=1029&message=10281935

[PacAce]

Often, it seems that the faster a lens gets, the more expensive it gets. For instance, if you go from the 70-200 f/4 to the f/2.8, it doubles the price.

But the 50mm f/1.8 is an $80 lens, and many people like this lense very much.

So how does speed factor into price? Are faster lenses that much harder to build, and if so how does the 50mm figure in?

Just a curious newbie...

S

The answer to your question is quite simple, really, if you understand what a "faster lens" means. A "faster lens", in comparison to another lens (with everything else being equal), is faster because is has a bigger aperture or f/stop. Aperture or f/stop is the ratio of the lens focal length to the optical diameter of the lens. And the bigger the diameter, relative to the focal length, the faster the lens is. So, if the diameter has to get bigger for the lens to become faster, then that logically means that the lens has to have more glass as someone else already pointed out. More glass means more expense. More glass mean that more grinding has to be done to make the lens and that also means more expense.

A faster telephoto lens is much more expensive to make than, say, a normal lens of the same aperture because the focal length of the telephoto is numerically bigger and hence, the diameter has to be proportionately bigger as well.

[robertwgross]

Normal depth of field calculations are based on the visibility of the slight blurring of out of focus areas when viewing a 10x8 print at some specified viewing distance (I forget the actual distance!).


But I can see the depth of field in my images while they are still on the screen, before they ever see a printer.

---Bob Gross---

[Shakespeare]

Another crazy question then...

Is it possible to make an f/1?

And why are there fewer L lenses in the lower mm?

S

[cmM]

Another crazy question then...

Is it possible to make an f/1?


There is a 50mm f/1.0 L

[CyberDyneSystems]

In fact Canon once made an f/.95 :shock: :shock:

[gcogger]

Normal depth of field calculations are based on the visibility of the slight blurring of out of focus areas when viewing a 10x8 print at some specified viewing distance (I forget the actual distance!).


But I can see the depth of field in my images while they are still on the screen, before they ever see a printer.

---Bob Gross---

Really? You mean there's a sudden division where things go from sharp to blurred?

In theory, objects are only 'in focus' at one precise distance and get progressively more blurred the further from that distance you look. Depth of field is about quantifying at what point that blurring becomes noticeable, and that depends on how closely you look at the image. This is specified by print size and viewing distance. I'm sure that the links I supplied explain it better than I can :) If not, I'm sure that Google could supply more articles.

In the digital age, it would be possible to calculate depth of field in a more absolute way - the distance at which the blurring can be resolved by the image pixels. Unfortunately this would lead to tiny depth of field figures, and it would also depend on how many pixels your sensor contained.

[robertwgross]

Really? You mean there's a sudden division where things go from sharp to blurred?

I said nothing of the kind. If you have a very narrow depth of field, then there can be a short range where the subject focus goes from good to bad.

I think what you are trying to say is that the measurement of the depth of field is related to how in-focus it looks in some mythical 8x10 inch print.

I have some shots where the subject was inadvertently left completely out of the depth of field, and it is extremely obvious as viewed on a screen, and it has nothing to do with a print, since they were never printed.

---Bob Gross---

[Jon]

On why faster lenses are more expensive:
A lens has to be designed so that all light rays from a given point on the subject, and striking any part of the lens surface, are focussed on the same point on the sensor. If the light rays from subject point 1 hitting point A on the lens surface and the light rays hitting point B on the lens surface don't come together at the same point on the sensor, you get a not-so-sharp picture (flat field issues). If the different frequencies of light don't all come to the same focus, you get colour fringing (chromatic aberration, not the same as the deadly purple fringe), which becomes more evident as the distance from lens to film becomes greater (CA is a bigger problem with teles than with wides).
Any lens has to be designed to correct for both these problems. When you've got less lens area (slower lens) you're dealing with a less-complex problem than if you have more lens area (and thus more possible paths to correct for). So, building a faster lens, in addition to requiring a larger chunk of glass, will require careful attention to correcting more possible light paths which will be incident on the sensor from a wider range of angles, and hence will have more varying lengths. Sorry, a diagram would help here. But correcting for the different light paths and for the chromatic aberration of these different light paths are principal contributors to this expense. As wide angle lenses have a bigger problem with the varying lengths of the light path through the lens to the sensor, and as telephoto lenses have more problems with CA, they both have corresponding demands on the design/construction which add to the expense of faster lenses. Whether "normal" lenses are really at the "sweet spot" for design, or whether they're cheaper because of volume, present deponent knoweth not.

When looking at FoV, it's been my understanding that "normal" corresponds, not to your peripheral field of view, but to the cone of vision which roughly corresponds to your stereo vision area (what you're actually seeing with both eyes), the area you see with best detail. You don't try to read a book, or do delicate work, held up at the edges of your total field of vision - you move it to the central zone of vision, and that zone is what roughly corresponds to what we call a "normal" lens FoV.

[cmM]

Sorry, a diagram would help here
Here's one of the problems Jon mentioned above and has to be combated when designing a lens. Chromatic aberation:

Longitudinal chromatic abberation (focal planes of different colors do not coincide)
http://www.vanwalree.com/optics/chromatic/long_color.gif

Transverse chromatic aberation (when the size of the image varies from one color to the other)
http://www.vanwalree.com/optics/chromatic/lat_color.gif

http://www.vanwalree.com/optics/chromatic.html - a must read. It will enlighten anyone. It worked on me, I love this article.

And this is just ONE issue that needs to be taken into consideration. There are many many other, and the bigger the lens, the harder it gets. You can only imagine what goes on in Canon's R&D department.

[Cadwell]

You can only imagine what goes on in Canon's R&D department.

Lots of "rubber band and wheely chair" fights I would imagine. That's what went on in most R&D departments I've worked in. :lol:

[gcogger]

Really? You mean there's a sudden division where things go from sharp to blurred?

I said nothing of the kind. If you have a very narrow depth of field, then there can be a short range where the subject focus goes from good to bad.

I think what you are trying to say is that the measurement of the depth of field is related to how in-focus it looks in some mythical 8x10 inch print.

I have some shots where the subject was inadvertently left completely out of the depth of field, and it is extremely obvious as viewed on a screen, and it has nothing to do with a print, since they were never printed.

---Bob Gross---

I think I'd better make this my last post on this subject, as we've gone so far off-topic!

As I said before, everything other than the exact focus distance is out of focus. 'Depth of field' is simply a question of how far out of focus it has to be before you can see the blurring - and this depends on how you view the image. There is no way to talk about DOF without specifying how you look at the image. If you magnify it enough, you will see the blurring easily at depths very close to the focus distance; if you look at, say, a contact print you won't see the blurring at depths very different from the focus distance.

DOF calculations have been around since way before the era of digital, so the standard DOF ideas were based on viewing a 10x8 print. This is what the DOF marking on lens focus scales are based on, for example.

You're saying, I think, that you can see on-screen when objects are beyond the DOF, since they don't appear sharp. That's fine - viewing on-screen is simply another way of specifying the viewing conditions, and every bit as valid as using a 10x8 print. This won't, however, correlate to what has traditionally been understood as the depth of field for a given aperture and distance. A 100% screen image is much more magnified than a 10x8 print, so you will be able to see slight blurring of parts of the image that would traditionally be regarded as within the DOF.

[robertwgross]

You're saying, I think, that you can see on-screen when objects are beyond the DOF, since they don't appear sharp. That's fine - viewing on-screen is simply another way of specifying the viewing conditions, and every bit as valid as using a 10x8 print. This won't, however, correlate to what has traditionally been understood as the depth of field for a given aperture and distance. A 100% screen image is much more magnified than a 10x8 print, so you will be able to see slight blurring of parts of the image that would traditionally be regarded as within the DOF.

Again, you jump to conclusions. I said nothing about a 100% screen image.

---Bob Gross---

[kraterz]

The 50mm lens is not cheap because it is "normal" (whatever that means) or has the same FOV as the eye or matches the diagonal on the 35mm camera (which is incidentally closer to 43mm).

50mm lenses are cheap to make because they have been manufactured for nearly three quarters of a century. Over these decades the design has changed very little. Most 50mm lenses (f/1.4, f/1.8) are simple designs and use very few elements. The 50/1.8 only uses 4 elements as opposed to 17 or 21 elements in some zooms. This makes it cheaper to get the end product within tolerance limits.