 |
 | Home | Contact Us | FAQ | Search & Site Map | Link to Us |
 | Sign In | Join | Other 45 Sites in Network |
| PhotoKB HomeDiscussion Groups Digital Photography Digital PhotoDSLR CamerasZLR CamerasPoint & Shoot CamerasFilm Photography 35 mmLarge FormatMedium formatDarkroomFilm and LabsOther EquipmentPhoto Technique Nature PhotographyPeople PhotographyTechnique GeneralGeneral Photo Topics General TopicsAustralian PhotographyUK PhotographyDirectoryPhoto Clubs |
Photo Forum / Digital Photography / Digital Photo / October 2006What make a lens normal ?
I'm back with more silly / newbie questions... In the 24 x 36 mm world, a normal lens is 50mm, normal because it offers the same angle of vision as the human eye. Now, this might be an error, my understanding is that at the equivalent quality, a normal lens is cheaper to make than either a wide angle and/or a telephoto lens. At the same time, it is possible to make normal lenses faster than any other type of lenses. My assumption here is because with the normal lenses you don't have to bend the light at weird angles the way it's done with wide an tele lenses. True / false ? Now here is the real question. With the 15 x 23 mm (aps-c) sensors, normal is 33mm. Right ? So, are we going to see cheap 33 mm lens, and faster than what we see today ? Thanks. > I'm back with more silly / newbie questions... > [quoted text clipped - 17 lines] > > Thanks. No. Nearly all lenses today for digital cameras are zooms. That is what the public wants and buys. You may start to find prime single focal length lenses in the normal lens for digitals, and some faster, but chances are the price will be more because the volume will be much less. The real reason that the "normal" lenses were cheap was because they made so many of them.  SignatureJoseph Meehan Dia duit > Now here is the real question. With the 15 x 23 mm (aps-c) sensors, normal > is 33mm. Right ? > > So, are we going to see cheap 33 mm lens, and faster than what we see > today ? I think it's difficult because the lens mount to film plane distance is still the same as with a full frame SLR. So the 33mm lens has to be built something like a full frame SLR wideangle and not like a normal lens. I'd also say the absence of a 33mm-ish macro lens is even more annoying than the situation with normal lenses. There are good 35/2 and 35/1.4 lenses available if you want to pay for them. But there's no good flat-field macro lens in that focal length. IMO, these APS-C DSLR's using 35mm SLR lenses are something of a stopgap because full frame sensors are still quite expensive. But I think that will improve over the next year or two. After rebates an EOS-5D is just a little over $2K, which is not much over what the Nikon D100 (maybe the first more-or-less consumer DSLR) cost when it came out. Remember that Nikon's early DSLR's, the E2 and E3, cost over $10K each and used 2/3", 1.3 megapixel sensors or thereabouts. I hope Nikon will release some full frame models at PMA 2007, initially including a D300 at today's D200 price level. Once there is competition for the 5D, these cameras should soon reach the D80 price level. Then you'll be able to use your 50mm normal and macro lenses. >> Now here is the real question. With the 15 x 23 mm (aps-c) sensors, >> normal [quoted text clipped - 24 lines] > 5D, these cameras should soon reach the D80 price level. Then you'll > be able to use your 50mm normal and macro lenses. Nice dream but unless they improve the yield for the lager sensors you will be waiting a bit longer. > Nice dream but unless they improve the yield for the lager sensors you will > be waiting a bit longer. They have already improved the yield tremendously for sensors of all sizes. 36x48mm sensors (found in medium format backs) were unobtanium a year or so ago. The D1 was $5000 when it came out. Its predecessor the E3 was over $10000 and had a much smaller sensor than the D1. The EOS-1DS was around $8000 and now the 5D is just over $2000. There's no reason to expect this process not to continue. >> Nice dream but unless they improve the yield for the lager sensors you >> will [quoted text clipped - 6 lines] > EOS-1DS was around $8000 and now the 5D is just over $2000. There's > no reason to expect this process not to continue. unobtanium Is that a new space age material? Ignore me, just having fun...; ) John > unobtanium > Is that a new space age material? http://en.wikipedia.org/wiki/Unobtanium > Nice dream but unless they improve the yield for the lager sensors > you will be waiting a bit longer. They have, and how! The big Fairchild, they fabed 22, 17 worked, and 15 passed QC! That is for 100+ Mp CCD for survey work. SignaturePaul Repacholi 1 Crescent Rd., +61 (08) 9257-1001 Kalamunda. West Australia 6076 comp.os.vms,- The Older, Grumpier Slashdot Raw, Cooked or Well-done, it's all half baked. EPIC, The Architecture of the future, always has been, always will be. > > Nice dream but unless they improve the yield for the lager sensors > > you will be waiting a bit longer. [quoted text clipped - 11 lines] > Raw, Cooked or Well-done, it's all half baked. > EPIC, The Architecture of the future, always has been, always will be. Hello, Paul: What is a "big Fairchild," anyhow? Cordially, John Turco <jtur@concentric.net> > I'm back with more silly / newbie questions... > [quoted text clipped - 7 lines] > the light at weird angles the way it's done with wide an tele lenses. True / > false ? Sort of, less bending then a wide angle lens and smaller elements then a telephoto lens. There is a sweet spot when lenses are the cheapest to make and still be fast and that would seem to be right around 50mm for a FF camera. For those how have not had the joy of trying to design a wide angle lens you can't make one without some pretty short radius surfaces and this dictates some pretty tight tolerances in alignment between the elements. You also need more elements for a wide angle lens. > Now here is the real question. With the 15 x 23 mm (aps-c) sensors, normal > is 33mm. Right ? > > So, are we going to see cheap 33 mm lens, and faster than what we see > today ? Well getting any lens faster then about 1.2 if pretty hard no matter how short the FL is. But I would like to see some f/1.4 shorter lenses that only work with the smaller sensors. If you only have an image circle large enough for the small sensors the lens should be cheap to produce. Scott >> I'm back with more silly / newbie questions... >> [quoted text clipped - 29 lines] > >Scott I wondered about the 135 MM lenses for full frame 35 mm film, it used to seem to me they were quite inexpensive relative to other focal lengths. > I wondered about the 135 MM lenses for full frame 35 mm film, it used > to seem to me they were quite inexpensive relative to other focal > lengths. You mean like the Canon 135/2.0? (List price in Japan is about US$1100, US street price is over US$800.) David J. Littleboy Tokyo, Japan > You mean like the Canon 135/2.0? (List price in Japan is about US$1100, US > street price is over US$800.) Back in the day, the situation wasn't anything like that. A top quality 135/2.8 was around $150. A not-so-top quality 135/1.8 was under $200. I got mine for $60 used at a camera store. Today the 30/1.4 Sigma cost about as much as a full frame 35/1.4 Nikkor used to cost. The Nikkor is now about $700 new, but that's because all the MF Nikkors are now quite expensive given the rarefied demand for them. I got mine used for around $150 on ebay a couple years ago. >> I wondered about the 135 MM lenses for full frame 35 mm film, it used >> to seem to me they were quite inexpensive relative to other focal [quoted text clipped - 5 lines] >David J. Littleboy >Tokyo, Japan I was remembering from back in the late 1970s, back when I thought I knew something about the subject. I haven't priced that range of lens lately, obviously things have changed. >>> I wondered about the 135 MM lenses for full frame 35 mm film, it used >>> to seem to me they were quite inexpensive relative to other focal [quoted text clipped - 6 lines] > knew something about the subject. I haven't priced that range of lens > lately, obviously things have changed. Sorry about the cheap shot. The 135/2.0 is a great lens, but it costs like a great lens. Seriously, though, you are quite right: there are a _lot_ of fun and affordable fixed-focus lenses for FF 35mm, both new and used. I have a friend who owns 17 or so 50mm lenses for his Leica and insists they all have different "characters". David J. Littleboy Tokyo, Japan > I wondered about the 135 MM lenses for full frame 35 mm film, it used > to seem to me they were quite inexpensive relative to other focal > lengths. You can make a really cheap 135mm lens, but not if it is also fast. A 135mm lens at f/1.4 would be very costly. Scott > Sort of, less bending then a wide angle lens and smaller elements then > a telephoto lens. There is a sweet spot when lenses are the cheapest > to make and still be fast and that would seem to be right around 50mm > for a FF camera. So, what is the sweet spot due to ? Something magical about 50mm (physics ?), or a connection to the size of the sensor/film ? > For those how have not had the joy of trying to design a wide angle > lens you can't make one without some pretty short radius surfaces and > this dictates some pretty tight tolerances in alignment between the > elements. You also need more elements for a wide angle lens. So again, wide angle as an absolute value, or compared to the sensor/film size ? Bye James E. Akiyama <james_akiyama@prodigy.net> Tue, 24 Oct 2006 14:06:22 - 0700 Actually a normal lens is described as one which provides a "normal" perspective. Normal perspective, I believe, was somewhat subjectively found. In any case it is roughly equal to a focal length equal to the diagonal of the image sensor. Using this 35mm film is approximately 43mm which was rounded to 50mm. James ----- Original Message ----- From: x@x.com Sent: Mon, 23 Oct 2006 23:37:16 Subject: Re: What make a lens normal ? Scott W <biphoto@hotmail.com> wrote: > Sort of, less bending then a wide angle lens and smaller elements then a > telephoto lens. There is a sweet spot when lenses are the cheapest to > make and still be fast and that would seem to be right around 50mm for a > FF camera. So, what is the sweet spot due to ? Something magical about 50mm (physics ? ), or a connection to the size of the sensor/film ? > For those how have not had the joy of trying to design a wide angle lens > you can't make one without some pretty short radius surfaces and this > dictates some pretty tight tolerances in alignment between the elements. > You also need more elements for a wide angle lens. So again, wide angle as an absolute value, or compared to the sensor/film size ? > Bye > James E. Akiyama <james_akiyama@prodigy.net> Tue, 24 Oct 2006 14:06:22 - [quoted text clipped - 5 lines] > diagonal of the image sensor. Using this 35mm film is approximately 43mm > which was rounded to 50mm. I understood a normal lens to be one that produces a perspective equivalent to the human eye. For example, you have two people, of equal height, one 100 feet away and the second 200 feet away. If you look at them through a normal lens or with the Mark I eyeball the farther person will look half as tall. If you look at them through a telephoto they will look closer in height. And if you look at them through a wide angle the far person will look even less than half. SignatureClark Martin Redwood City, CA, USA Macintosh / Internet Consulting "I'm a designated driver on the Information Super Highway" >I understood a normal lens to be one that produces a perspective >equivalent to the human eye. For example, you have two people, of equal [quoted text clipped - 3 lines] >will look closer in height. And if you look at them through a wide >angle the far person will look even less than half. That's definitely *not* true, as you can verify for yourself by simply trying different lenses in the scenario you describe. Under the conditions you describe, the near person will *always* be twice as tall in the image as the far person, no matter what lens you use. Changing focal length will make both of them smaller or larger at the same time, but the 2:1 ratio between them always remains. Dave >So, what is the sweet spot due to ? >Something magical about 50mm (physics ?), or a connection to the size of >the sensor/film ? There's nothing special about the absolute number "50". It is the angular relationship between the focal length and the size of the image circle that matters. For a 24x36 mm image, the diagonal is about 43 mm. By the usual standards, the "normal" lens for the format ought to be about 43 mm focal length. When focused at infinity, a point in the subject that is imaged in one corner of the frame is about 21.5 mm off the optical axis, which is an angle of arctan(0.5) = 27 degrees for a 43 mm lens. A 50 mm lens would reduce this angle to about 23 degrees, not a large change. And that's what really defines a "normal" lens: angular coverage of around 25 degrees off-axis. If you have a digital sensor with a 10 mm diagonal, a 10 mm lens would be normal for that format. This a "sweet spot" for lens design because things get harder and more expensive on both sides of that. Wide angle lenses have light arriving at much steeper angles to the axis, which makes controlling aberrations and avoiding vignetting more difficult. The wider you get, the lenses tend to have more elements that are larger, and thus more expensive. Telephoto lenses have fewer problems with off-axis aberrations. But their long focal length means that very small amounts of chromatic aberration are magnified into visible colour fringes in the image. So telephotos have to worry more about colour correction, and sometimes use exotic (expensive) glass for that. Also, every time you double the focal length, you need to double the lens element diameters to maintain the same f/number, and doubling diameter means 8 times the weight of glass per element. >So again, wide angle as an absolute value, or compared to the sensor/film >size ? Wide angle is determined by the *angle* as an absolute number, which is also related to the ratio between sensor size and focal length. Dave > There's nothing special about the absolute number "50". It is the > angular relationship between the focal length and the size of the image > circle that matters. > For a 24x36 mm image, the diagonal is about 43 mm. By the usual > standards, the "normal" lens for the format ought to be about 43 mm Thanks, that's what I thought, but needed a confirmation. So, say on a Pentax K100d/K110d, the sensor is 23.5 x 15.7 (according to DP review), that makes a diagonal of 28.26 mm. Therefore on those cameras a normal lense would be aorund 28mm. Right ? > focal length. When focused at infinity, a point in the subject that is > imaged in one corner of the frame is about 21.5 mm off the optical axis, > which is an angle of arctan(0.5) = 27 degrees for a 43 mm lens. aheu... you lost me here. I can't quite picture this. We know the size of the sensor (24mm x 36mm), and we know the distance between the sensor and the bottom of the lens (43mm), but do we know the size of the bottom of the lens ? How do you work out the angle ? > This a "sweet spot" for lens design because things get harder and more > expensive on both sides of that. Wide angle lenses have light arriving > at much steeper angles to the axis, which makes controlling aberrations > and avoiding vignetting more difficult. The wider you get, the lenses > tend to have more elements that are larger, and thus more expensive. So, if I understand you right, the sweet spot for the APS-C type camera should be 28mm ? 28mm should be easy to make but smaller will mean wide angle, therefore more difficult ? >>So again, wide angle as an absolute value, or compared to the sensor/film >>size ? > Wide angle is determined by the *angle* as an absolute number, which is > also related to the ratio between sensor size and focal length. So in the APS-C world, wide angle is < 28mm ? Therefore, we should see "cheap", fast, good quality 28mm lenses coming on the market... ? Yves. : 28mm should be easy to make but smaller will mean wide angle, therefore : more difficult ? You keep going back to cost. And the manufacturing cost (and selling price) of a lens has many many many more variables than the focal length of he lens package (as opposed to an individual hunk of shaped glass or plastic that could also be called a lens or lens element). First the larger the physical size and number of the lens elements that make up a lens package will effect the cost. Also the material the lens elements are made of will effect the cost (plastic, glass, etc). And if the elements are coated to reduce various unwanted lighting effects (such as internal reflections) the type and quality of these coatings can effect the cost. Then there is the size of the sensor that the lens package is designed to properly project on can effect the cost. And then lastly, yes, the focal length of the lens can effect the cost. But there is also a problem that any lens that there is a lot of demand for can allow mass producing the required lens elements and thus reduce the cost. Also if the camera is not supposed to be a high quality camera the number of picky people buying it will be lower and thus the manufacturer may be able to get away with a slightly lower quality (and less expensive) lens. Think of the plastic paper disposable cameras of recent years. A professional photographer could get a good photo from such a camera but if he didn't he wouldn't be suprized is the lens quality is less than perfect. If a small box type camera with a single, one element, plastic lens with no focus, and a set aperture, and a mechanical spring loaded shutter is made, the cost of the whole thing will be rather cheep. The lens alone may cost a few cents. It probably will be at about "normal" length for the image capture device (film, etc) but could be almost any focal length. The reason the lens is inexpensive is because it is made from a chunk of plastic about the size of a dime and was cast into the shape. The amount of individual attention given to this lens is minimal in the making of it. Now if we look at a lens that has the same focal length but is for use on a professional type SLR with very large, very precisely ground, glass lens elements which have been individually ground and tested, and then individually placed in the lens housing and precicely adjusted for the most exacting alignments. Then the individual lens elements are coated with very rare elements that cut out any problems caused by internal reflections and inconsistancys caused by non-visible EM frequency waves that may cause problems with a sensor that can, at least partially, detect these invisible light waves that the photographer can't see. Even tho the focal length is the same, the cost will be WAY WAY higher than the few cents before. Any of the factors in the manufacture of a lens that is taken to an extreme will cost more. A lens that is 4 ft across, that is in all other ways exactly the same, will cost more than one that is 1 inch in diameter. At the same time one that is 1/16th of an inch in diameter will also cost more than the 1 inch one. You notice that the focal length is not in this particular formula. So when you keep trying to equate the focal length that is concidered "normal" for a particular camera with price, you are confused. Unless absolutely every other possible factor is precisely the same the focal length is only a small item in a very long list of factors in the price. True, if more of a particular FL lens are being made and sold the price for that lens (for that particular camera) may go down due to the savings of mass production. But this has little or no direct connection to what is a "normal" lens. If someone sells lots of telephoto lenses for a camera regularly used for sports photography, this does not make this telephoto lens "normal". And because there are more of them made, it is very possible that the cost of the telephoto lens could be less than the cost for a "normal" lens for that camera. Simply because fewer normals are made and thus the production cost is more per unit. Randy ========== Randy Berbaum Champaign, IL >Thanks, that's what I thought, but needed a confirmation. >So, say on a Pentax K100d/K110d, the sensor is 23.5 x 15.7 (according to >DP review), that makes a diagonal of 28.26 mm. Therefore on those cameras >a normal lense would be aorund 28mm. Right ? Yes, that's correct. >> focal length. When focused at infinity, a point in the subject that is >> imaged in one corner of the frame is about 21.5 mm off the optical axis, >> which is an angle of arctan(0.5) = 27 degrees for a 43 mm lens. >aheu... you lost me here. I can't quite picture this. We know the size of the >sensor (24mm x 36mm), and we know the distance between the sensor and the >bottom of the lens (43mm), but do we know the size of the bottom of the lens ? >How do you work out the angle ? 43mm is not the distance from the back of the lens to the sensor. It is the effective focal length of the lens, which means that the magnification is the same as for a pinhole located 43 mm from the sensor. If your camera used a single-element lens, the centre of the lens would be close to 43 mm from the sensor. But the actual rear element of a multi-element complex lens may be either further or closer. The computation is based on the pinhole equivalent of the lens. The corner of the frame is 21.5 mm off the optical axis, and the pinhole is located 43 mm from the sensor. So, for a light ray coming from the subject, going through the pinhole, and hitting the corner of the frame, there is a certain angle between the optical axis (a line from the centre of the sensor through the pinhole) and this ray. We've got a right-angle triangle and we know the length of the sides adjacent to and opposite the angle in question: 43 mm and 21.5 mm respectively. The ratio of these two is the tangent of the angle, so we can use arctan to calculate the angle, and that's what I did. Basic trigonometry. >So, if I understand you right, the sweet spot for the APS-C type camera >should be 28mm ? >28mm should be easy to make but smaller will mean wide angle, therefore more >difficult ? If a manufacturer scaled down the entire camera, including lens mount flange depth, then they could also scale down a normal lens design to get a small light cheap 28 mm normal lens with good performance. However, manufacturer's don't do that - they keep the same lens mount position on the APS-C cameras as they do full-frame cameras, so you can't necessarily use the same simple design. Giving a lens-to-sensor distance that is substantially larger than the focal length requires a special (more complex) lens design. >So in the APS-C world, wide angle is < 28mm ? >Therefore, we should see "cheap", fast, good quality 28mm lenses coming on >the market... ? Probably not, since APS-C DSLRs are not simply scaled-down full-frame DSLRs. This allows you to use the wide array of full-frame lenses available, but there is a price to pay too, and the lack of small cheap fast normal lenses may be part of that price. Dave > 43mm is not the distance from the back of the lens to the sensor. It > is the effective focal length of the lens, which means that the > magnification is the same as for a pinhole located 43 mm from the > sensor. THANK YOU ! This is the first time I read a definition of focal length that make sense to me (and interrestingly enough, I've been playing quite a bit with pinholes lately). > The computation is based on the pinhole equivalent of the lens. The Right, and now it makes sense since you get a triangle. Thanks again. >>So, if I understand you right, the sweet spot for the APS-C type camera >>should be 28mm ? >>28mm should be easy to make but smaller will mean wide angle, therefore more >>difficult ? > If a manufacturer scaled down the entire camera, including lens mount > flange depth, then they could also scale down a normal lens design to > get a small light cheap 28 mm normal lens with good performance. > However, manufacturer's don't do that - they keep the same lens mount Hmmm, interresting. Reading this you could think the four thirds system is the way to go, but their lenses at equivalent speed (I guess quality is somewhat subjective) aren't cheaper than the competition. Thanks again, you seem very knowledgeable, and are able to transfer that knowledge in a language that stays simple. You should write a book :-) Yves. >> 43mm is not the distance from the back of the lens to the sensor. It >> is the effective focal length of the lens, which means that the >> magnification is the same as for a pinhole located 43 mm from the >> sensor. >THANK YOU ! >This is the first time I read a definition of focal length that make sense >to me (and interrestingly enough, I've been playing quite a bit with >pinholes lately). With a pinhole, you can have any "focal length" you want. With a lens, you find the point of best focus, and then its focal length *is defined by the image magnification* near the lens axis. In a sense, that's why the simple geometrical lens formulas work for complex lenses. >> If a manufacturer scaled down the entire camera, including lens mount >> flange depth, then they could also scale down a normal lens design to >> get a small light cheap 28 mm normal lens with good performance. >> However, manufacturer's don't do that - they keep the same lens mount >Hmmm, interresting. Reading this you could think the four thirds system is >the way to go, but their lenses at equivalent speed (I guess quality is >somewhat subjective) aren't cheaper than the competition. Don't they say that they designed the system so that the rays of light reaching the corner of the sensor arrive at a shallower angle than they would with film? This means a large lens mount diameter and probably not reducing the flange depth, then using retrofocus designs like the other DSLR manufacturers would have to. There goes small, light, and inexpensive normal lenses. Dave > With a pinhole, you can have any "focal length" you want. With a lens, > you find the point of best focus, and then its focal length *is defined > by the image magnification* near the lens axis. In a sense, that's why > the simple geometrical lens formulas work for complex lenses. Right, a pinhole is always in or out of focus (or always out of focus by the same amount). So in a lens, what happens physically when one turns the focus ring (I've asked this in a different thread, but not getting any answer) ? >>Hmmm, interresting. Reading this you could think the four thirds system is >>the way to go, but their lenses at equivalent speed (I guess quality is >>somewhat subjective) aren't cheaper than the competition. > Don't they say that they designed the system so that the rays of light > reaching the corner of the sensor arrive at a shallower angle than they > would with film? This means a large lens mount diameter and probably > not reducing the flange depth, then using retrofocus designs like the > other DSLR manufacturers would have to. There goes small, light, and > inexpensive normal lenses. Yes, but I had not realised that they had to make the lens larger in order to achieve this. I did find that the lenses were pretty large, but thought that this was due to the engine for the zoom and focus inside the lenses. That and the fact that they decided to snob north-America with their newer better model made me look at a different brand ! Yves. > Right, a pinhole is always in or out of focus (or always out of focus by the > same amount). So in a lens, what happens physically when one turns the > focus ring (I've asked this in a different thread, but not getting any > answer) ? The following link shows how a pinhole creates an image, and why it's dim and not very sharp. http://i106.photobucket.com/albums/m266/lemel_man/Lens1.jpg A perfect pinhole would allow only one ray through for each point on the object; the image would be perfectly sharp everywhere, but very dim. If you enlarge the pinhole to let more light through, you get a little blurred circle of light for each object point, so the image is brighter, but blurred. The link also shows how a perfect convex lens bends rays of light. When a ray of light passes from one transparent medium to another (air to glass, for example), it will bend unless it is perpendicular to the boundary. A convex lens (or magnifying glass)is designed so that rays parallel to its axis are bent so as to pass through a single point. This single point is called the Focal Point, and the distance of this point from the centre of the lens is called the Focal Length of the lens. This link, http://i106.photobucket.com/albums/m266/lemel_man/Lens2.jpg shows how a convex lens produces an image. Of the rays from the top of the image, one is parallel to the lens axis and gets bent to pass through the focal point as shown, another goes straight through the centre of the lens. A sharp image is formed where they meet. This effect happens for all the rays from every object point that passes through the lens. Think of it as a cone of light coming from each object point to the lens, then getting bent to form another cone on the other side of the lens. The tip of this cone is then an image of the original object point. If you place the image plane at this point, the image will be sharp, but if you move the plane nearer to, or further from, the lens, the rays no longer meet and instead of getting a sharp point you get a little blurred circle and the image is out of focus. When you turn the focus ring on a camera lens you simply move the entire lens closer to or further from the film plane. The amount of blurring is determined by the size of the image circle that is meant to be a point; the bigger the circle the more the image is blurred. The size of the circle is determined by the thickness of the cone and the distance from its tip: the thicker the cone, the more critical the focus. In camera terms, the F-number of a lens is the ratio of the focal length (FL) divided by the diameter of its aperture (D): a 50mm FL lens with a D of 20mm would have an F-number of 2, and a D of 10mm would give an F-number of 4. Big F-numbers mean thinner light cones which imply less critical focussing. SignatureI hope this helps, Gary Wooding (To reply by email, change feet to foot in my address) -- Posted via a free Usenet account from http://www.teranews.com >Right, a pinhole is always in or out of focus (or always out of focus by the >same amount). So in a lens, what happens physically when one turns the >focus ring (I've asked this in a different thread, but not getting any >answer) ? It depends on the lens design. The "classical" way of doing this is mounting all of the lens elements in a rigid barrel with fixed spacing between them, and then use a focusing mount (which just uses a helical thread) to move the whole optical barrel forward and back at the same time. So the lens focal length remains exactly the same; you just change the distance to the film/sensor as the subject gets closer. On the other hand, some lenses have "internal focusing". In these, the optics are arranged in two or more groups that can move independently, and turning the focusing ring operates a cam that moves one or more groups, changing the internal spacing of the lens elements. This might change the actual focal length, or just move the position of the rear principal plane without actually moving the rear element any. It's up to the lens designer. This allows more design flexibility, but requires some of the complexity of manufacture of a zoom lens. Dave > It depends on the lens design. The "classical" way of doing this is > mounting all of the lens elements in a rigid barrel with fixed spacing > between them, and then use a focusing mount (which just uses a helical > thread) to move the whole optical barrel forward and back at the same ../.. > On the other hand, some lenses have "internal focusing". In these, the > optics are arranged in two or more groups that can move independently, So on a zoom, assuming a very simple lens design, the focussing ring moves the lens very slightly, but the zoom ring moves it by a lot more ? On zooms, the size of the image does not change as you move the focusing ring. Yves. >So on a zoom, assuming a very simple lens design, the focussing ring moves >the lens very slightly, but the zoom ring moves it by a lot more ? All you can say with certainty is that the focusing ring moves one or more components of the zoom in such a way that the plane of best focus moves towards or away from the camera. While the zoom ring moves one or more components of the zoom in such a way that the in-focus plane does not move but magnification does change. But the two movements will be different internally, since they have very different effects. >On zooms, the size of the image does not change as you move the focusing ring. Not quite true, since the size of the image does change as you adjust the focusing ring for *any* standard design lens, fixed focal length or zoom. The exception to this is lenses that a telecentric in image space, which are pretty unusual in photography. Dave > I'm back with more silly / newbie questions... > > In the 24 x 36 mm world, a normal lens is 50mm, normal because it offers > the > same angle of vision as the human eye. The "normal" lens offers neither the wide angle capability of the human eye (looking straight ahead, one's peripheral FOV is much closer to what a 20mm lens sees) nor the narrow angle of the central section where our vision is the sharpest. It looks to me that "normal" is a FOV that is easy to create optically (anything wider would be difficult with the three (or fewer) element lenses used in early P&S cameras) and not too long to be inconvenient (anything longer would make the (what is now medium format) P&S cameras that appeared starting in 1900 be bulkier than would be convenient. FWIW, Canon's three simplest lenses are the 50/1.8 (6 elements), 35/2.0 (7 elements), and 90/2.8 TSE (6 elements). > Now, this might be an error, my > understanding is that at the equivalent quality, a normal lens is cheaper > to > make than either a wide angle and/or a telephoto lens. At the same time, > it is possible to make normal lenses faster than any other type of lenses. From looking at MTF curves, my impression is that 50mm lenses fall into the "wide angle" camp. This is because pretty much all lenses longer than normal (e.g. the 100/3.5 Hasselblad lens, the 110/2.8 Mamiya (both of which are only slightly longer than normal), and all the Canon/Nikon primes 85mm and over) retain excellent sharpness right out to the corners, whereas pretty much every lens normal to shorter (including the Hassy 80/2.8) shows significantly reduced contrast at the corners. > My assumption here is because with the normal lenses you don't have to > bend > the light at weird angles the way it's done with wide an tele lenses. True > / > false ? Longer lenses have a much easier time of it. They're physically larger, and thus more expensive. Also, until recently, the normal lenses were produced in vast quantities compared to any other lens, and thus were cheaper. > Now here is the real question. With the 15 x 23 mm (aps-c) sensors, normal > is 33mm. Right ? > > So, are we going to see cheap 33 mm lens, and faster than what we see > today ? The Sigma 30/1.4 is fairly old news. Big and expensive, though. Unlike FF normal lenses, it won't be produced in large quantities. David J. Littleboy Tokyo, Japan > I'm back with more silly / newbie questions... > [quoted text clipped - 19 lines] > > Thanks. Not a silly question at all. That is what you should see...however the camera companies are taking this opportunity to slip in a major price increase by creating APS-C lenses and not doing this. (Just like Danon reducing the weight of their yogurt by 25% while keeping the price the same...the sneaky price increases are the ones that sell you less for the same price and those really irritate me.) Also, if you compare normal lenses for view cameras with normal lenses for medium format with normal lenses for 35mm, you'll also notice that they get cheaper AND faster with smaller coverage area...so, I'd assume that your 33mm lens should be a f/1.0-1.2 and cheaper than a 50mm f/1.4 for a 35mm camera...but, I am not holding my breath for such a thing. George > smaller coverage area...so, I'd assume that your 33mm lens should be a > f/1.0-1.2 and cheaper than a 50mm f/1.4 for a 35mm camera...but, I am not > holding my breath for such a thing. There was a 40/1.4 and 42/1.2 for the Olympus Pen F half-frame 35mm system, whose image area was similar in size to APS-C. But I think that was easier to do because the whole camera was scaled down compared with a 35mm camera, including the mount-to-film distance. The 35mm-descended DSLR system cameras are fundamentally malproportioned. Canon implemented a workaround for that in some of the cheap digital EOS cameras and lenses (EF-S, I think), but I think they didn't dare extend it to the higher end models. Maybe there will be some fast normal primes for the 4/3 system. For 35mm-based systems, I think we should consider APS-C sensors to be a cost-imposed stopgap, and 24x36mm sensors should become the norm as technology improves in the future. > There was a 40/1.4 and 42/1.2 for the Olympus Pen F half-frame 35mm > system, whose image area was similar in size to APS-C. But I think [quoted text clipped - 4 lines] > the cheap digital EOS cameras and lenses (EF-S, I think), but I think > they didn't dare extend it to the higher end models. Heh, that is simply because the higher end models have bigger sensors (1.25 or 1.0 crop), and EF-S is specifically made for 1.6 crop sensors. Firstly, the EF-S's don't cast an image circle big enough to fill the sensors of the 1D/5D series of cameras, and secondly they protrude so far into the camera body that the mirror on a non-crop camera would hit the end of the lens. : I'm back with more silly / newbie questions... : In the 24 x 36 mm world, a normal lens is 50mm, normal because it offers the : same angle of vision as the human eye. Now, this might be an error, my : understanding is that at the equivalent quality, a normal lens is cheaper to : make than either a wide angle and/or a telephoto lens. At the same time, : it is possible to make normal lenses faster than any other type of lenses. Not necissarily. "Normal" has nothing to do with the cost of the lens or even the ease of manufacture. This designation is a somewhat arbitrary name for the lens that gives a field of view (along the long edge of the image) of about 40 deg. Supposedly this was to match the view of the human eye (tho I know my field of view is much wider than 40 deg). : My assumption here is because with the normal lenses you don't have to bend : the light at weird angles the way it's done with wide an tele lenses. True / : false ? A lens always has to bend light or it would be a window. :) : Now here is the real question. With the 15 x 23 mm (aps-c) sensors, normal : is 33mm. Right ? Yes. It isn't exact as if I remember my calculations (some time ago) the 40 deg mark comes at about 32.85 mm. But since Normal is an approximate designation you could count anything from about 31 to 35 as a good approximation of "Normal". : So, are we going to see cheap 33 mm lens, and faster than what we see : today ? The wider the lens the more difficult it is to grind. And to reduce some of the more objectionable edge effects there are more involved optics in the lens barrel. But with the prevalence of Digital cameras with similar sensor dimensions the mass produced price of wider lenses will get cheaper. I can remember back when it was virtually impossible to find a lens wider than 18mm without getting a fisheye lens. And the ones down in that end of the spectrum were major bucks (>$1000). Now I have found several makes of zooms that go down to about 12mm for about $500. So the prices are coming down. Of course all of this is JMHO and so YMMV. :) Randy ========== Randy Berbaum Champaign, IL > I can remember back when it was virtually impossible to find a > lens wider than 18mm without getting a fisheye lens. And the ones down in > that end of the spectrum were major bucks (>$1000). Now I have found > several makes of zooms that go down to about 12mm for about $500. Not any that cover 24x36mm! There is a 12mm lens made by Voigtlander for the Bessa rangefinders but its rear element comes very close to the film (couldn't work on an SLR because no clearance for the reflex mirror), and also it's not a zoom. > > I can remember back when it was virtually impossible to find a > > lens wider than 18mm without getting a fisheye lens. And the ones down in [quoted text clipped - 5 lines] > the film (couldn't work on an SLR because no clearance for the reflex > mirror), and also it's not a zoom. I've noticed that the "wide zoom" 18-50mm Sigma kit lens that came with my Pentax *ist-DS can actualy cover a full 35mm frame between 50mm and about 24mm, wider than that it progressively becomes more vignetted until 18mm, when it gives a circular image on the film. I tested it on my K1000 just for laughs. : > I can remember back when it was virtually impossible to find a : > lens wider than 18mm without getting a fisheye lens. And the ones down in : > that end of the spectrum were major bucks (>$1000). Now I have found : > several makes of zooms that go down to about 12mm for about $500. : Not any that cover 24x36mm! There is a 12mm lens made by Voigtlander : for the Bessa rangefinders but its rear element comes very close to : the film (couldn't work on an SLR because no clearance for the reflex : mirror), and also it's not a zoom. Since we were talking about the smaller APS sensor I was confining my comment to that. And as to SLRs I have seen several makes of zoom lens for my Ist-DS that are (at last check) in that price and zoom range. True fewer lenses are being made for a full frame (35mm film) sensor, but until more full frame bodys are marketed this is probably going to continue. In part the price will fall for those items that there is a larger desire for. Even if the materials cost is equivalent. A brake assembly for a steam driven car will cost more than one for a new GM. Even tho the materials cost would be about the same. :) On the other hand, when full frame sensors become the norm on the majority of digital camera bodies I would expect the cost of FF lenses to fall to some extent. Of course the rising costs of materials, manufacturing and shipping will offset that to an extent. Also few FF cameras will use a simple 1970's lens, and all those additional built in devices will add to the price (few early '70s SLR lenses included a chip for reporting the f-stop, had an internal motor for AF, or included IS technology). But I would expect that whatever size and style of sensor is most prevalent at the time will be at least comparitively less expensive than those not in current wide spread use (speaking of most recent vintage here). JMHO Randy ========== Randy Berbaum Champaign, IL > : > I can remember back when it was virtually impossible to find a > : > lens wider than 18mm without getting a fisheye lens. And the [quoted text clipped - 5 lines] > Since we were talking about the smaller APS sensor I was confining > my comment to that. Huh? Those $1000+ 18mm lenses were for 24x36mm. If you're talking about smaller formats, then of course there were shorter lenses. 6mm was a fairly standard wideangle for 16mm cinema, I think. The Minox still camera had a very good 15mm prime. And there were Beaulieu 8mm movie cameras that had a 6-70/1.4 zoom in the 1970's or so: http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=280037999767 > > I can remember back when it was virtually impossible to find a > > lens wider than 18mm without getting a fisheye lens. And the ones down in > > that end of the spectrum were major bucks (>$1000). Now I have found > > several makes of zooms that go down to about 12mm for about $500. > > Not any that cover 24x36mm! Well, yes. The Sigma 12-24mm is a full frame lens, and still not insanely expensive. ---Ketil > Well, yes. The Sigma 12-24mm is a full frame lens, and still not > insanely expensive. http://www.sigma-photo.com/lenses/lenses_all_details.asp?id=3236 Wow!!! I'd thought that "EX DG" referred to APS-sized coverage but it looks like it is full frame. Still, it's two stops slower than a 20/2.8 AF Nikkor, and costs more. > http://www.sigma-photo.com/lenses/lenses_all_details.asp?id=3236 > > Wow!!! I'd thought that "EX DG" referred to APS-sized coverage but it > looks like it is full frame. Still, it's two stops slower than a > 20/2.8 AF Nikkor, and costs more. "DG" means "digital" coatings, I believe. EX is the "pro" series. If a Sigma lens has "DC" in its alphabet soup it has an APS-sized image circle. Of course it is slower and more expensive than the fixed-focal Nikkor. It's a zoom after all. > I'm back with more silly / newbie questions... > [quoted text clipped - 15 lines] > > Thanks. What is considered a "normal" lens has changed with time. Originally, for 35mm it was considered to be a focal length equal to the diagonal of the format. This actually was about 44 mm. That actually created a bit of wide angle perspective, and most folks actually preferred something closer to 50 mm. There was a period when wide angle became unpopular and the "normal" lens supplied by many mfgs in SLR "kits" became more like 55mm. Supposedly, however, the reason this is a 'norma' lens is because when an 8 x 10 print is viewed at a comfortable and normal distance, a picture taken with a normal lens will yield a fov or angular subtense that approximates the subtense of that print. That is, a "normal" lens yields a 1:1 magnification when an 8 x 10 print is viewed. Note that unlike modern times, 8 x 10 was considered the normal print size, not 4 x 6 :-) >Note that unlike modern times, 8 x 10 was considered >the normal print size, not 4 x 6 :-) I wonder who considered 8x10 normal for 35mm. It strikes me as a rather useless format. It is too small to hang on the wall and it is too big to pass around. SignatureThat was it. Done. The faulty Monk was turned out into the desert where it could believe what it liked, including the idea that it had been hard done by. It was allowed to keep its horse, since horses were so cheap to make. -- Douglas Adams in Dirk Gently's Holistic Detective Agency Philip Homburg skrev: > >Note that unlike modern times, 8 x 10 was considered > >the normal print size, not 4 x 6 :-) > > I wonder who considered 8x10 normal for 35mm. It strikes me as a rather > useless format. It is too small to hang on the wall and it is too big > to pass around. I sort of did, the days I halfway lived in a darkroom :-) I printed almost everything on 24x30cm - must be 9*12 or thereabout. Big enough for the walls of my small room, and the biggest size that would fit my rucksack. And yes, easy enough to pass around. I started that when I realised that it was DKK5.- (in 1990) per sheet - about 80c at todays rate. With a prodiction of 2 per hour in c/neg and 4-6/h in B/W, I wanted something that impressed. /Martin |
Reply to this Message |
 Sign In
 Join
 My Latest Posts My Monitored Threads My Blog My Photo Gallery My Profile My Homepage Start New Thread Enable EMail Alerts Rate this Thread
|
©2010 Advenet LLC Privacy Policy - Terms of Use
This website includes both content owned or controlled by Advenet as well as content owned or controlled by third parties.