 |
 | 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 / DSLR Cameras / October 2006Internal image stabilization vs long lenses?
Several camera makers have introduced SLR cameras with internal image stabilization (IIS). My question is how well will this work with long (300+mm) lenses? I know that IIS works in point and shoot cameras from the Minolta Di-mage A2 I owned. But the A2 IS system was designed around the single lens. Logic tells me that the longer the lens the more an internal sensor will have to move to provide the same degree of stabilization. At some lens length it has to reach a point of diminishing returns. Has anyone seen any real world testing that shows how well IIS works with telephoto lenses? Thanks, Holley I wasn't testing AntiShake in Sonolta A100 for long time but I can say it actually works. I was able (as far as I remember...) to shoot nearly perfectly sharp photo @ 300mm @ 1/15s. I have to save a bit to afford some VR lens and have similar effect. :( I don't know how do the VR and IS perform, compared to sensor stabilization in Sonolta and Pentax but IIS is surely good enough. But there is a small problem - error message saying, the camera is overheated. :) And I suppose warming the sensor up by IIS increase noise. (what is something veeery nasty, because A100 even with cold sensor gives quite noisy results) > I wasn't testing AntiShake in Sonolta A100 for long time but I can say > it actually works. I was able (as far as I remember...) to shoot nearly > perfectly sharp photo @ 300mm @ 1/15s. Printed at 3.5 x 5, maybe. > I have to save a bit to afford > some VR lens and have similar effect. :( [quoted text clipped - 4 lines] > noise. (what is something veeery nasty, because A100 even with cold > sensor gives quite noisy results) At what ISO?  Signature-- r.p.e.35mm user resource: http://www.aliasimages.com/rpe35mmur.htm -- r.p.d.slr-systems: http://www.aliasimages.com/rpdslrsysur.htm -- [SI] gallery & rulz: http://www.pbase.com/shootin -- e-meil: Remove FreeLunch. > Printed at 3.5 x 5, maybe. Viewed it on laptop screen and of course there was same blur but as for this focal length and exposure time it's a nice result. > At what ISO? ISO <800 is OK and prints were very good. (as far as I remember these were 13x18) ISO 800 was a bit too noisy but still usable and good enough. ISO 1600... well... only small prints. :/ Maybe I'm oversensitive but I shoot a lot in poor light and good high-iso performance is one of the most important factors to me. > Several camera makers have introduced SLR cameras with internal image > stabilization (IIS). My question is how well will this work with long [quoted text clipped - 6 lines] > length it has to reach a point of diminishing returns. Has anyone seen any > real world testing that shows how well IIS works with telephoto lenses? I've used it with the Maxxum 7D and a 300mm f/2.8 lens (+ 1.4 and 2x TC's). It is not terribly effective. At best it might help 1/rule-of-thumb shots. Unless the shutter speed is 1/1000 or higher, I really don't see the point in shooting 1/300 w/o a tripod in any case. OTOH, the IS lenses will give an advantage to about 2 stops, even 3 and bit if you are especially steady. The nice thing about A/S in the Minolta is that I got the immediate advantage of it for my 50, 100, 28-70 and 80-200 lenses. Canon/Nikon users with older non IS/VR lenses have to buy individual lenses with the feature. If you're starting a new professional system, however, I would simply go straight to Canon as they have the best digital system, one of the very best film systems and the best collection of up to date lenses available. If you're an amateur and really don't see going beyond that (and amateur does not exclude very high quality images) then consider Sony and Pentax as well as Canon and Nikon. Olympus too. Cheers, Alan. Signature-- r.p.e.35mm user resource: http://www.aliasimages.com/rpe35mmur.htm -- r.p.d.slr-systems: http://www.aliasimages.com/rpdslrsysur.htm -- [SI] gallery & rulz: http://www.pbase.com/shootin -- e-meil: Remove FreeLunch. >> Logic tells me that the longer the lens the more an internal sensor >> will have to move to provide the same degree of stabilization. At some [quoted text clipped - 8 lines] > Unless the shutter speed is 1/1000 or higher, I really don't see the > point in shooting 1/300 w/o a tripod in any case. doh! ^^^300mm^^^ > Several camera makers have introduced SLR cameras with internal image > stabilization (IIS). My question is how well will this work with long [quoted text clipped - 9 lines] > Thanks, > Holley i'd say that in a case of long lenses in-lens IS is more effective than internal one, since in lens max. movement can be smaller. I can only say that my Canon 70-300 IS USM works perfectly, and hand-shooting at 300mm is not a problem at all. >> Several camera makers have introduced SLR cameras with internal image >> stabilization (IIS). My question is how well will this work with long [quoted text clipped - 12 lines] > that my Canon 70-300 IS USM works perfectly, and hand-shooting at 300mm > is not a problem at all. I agree with you, although it's only from intuition. Have you - or any one else- done any tests to show how many stops you pick up? If so, how were they conducted? SignatureJohn McWilliams >>> Several camera makers have introduced SLR cameras with internal image >>> stabilization (IIS). My question is how well will this work with long [quoted text clipped - 16 lines] > else- done any tests to show how many stops you pick up? If so, how were > they conducted? i didn't exactly test how many stops i gain. But i definitely do gain. Say, shooting at 300 mm from hand with 1/50 is definitelly not any problem. Maybe i'll do that one day... But, sure, to have an honest comparison, a man would need to have both systems for testing. Pentax recomend not using over 800mm on the K10D. > Several camera makers have introduced SLR cameras with internal image > stabilization (IIS). My question is how well will this work with long [quoted text clipped - 9 lines] > Thanks, > Holley > Has anyone seen any real world testing that shows how well IIS works with telephoto lenses? Also, if the camera has built-in stabilization under what circumstances is it a bad idea to use it? Mine has an off switch but I can't think why not to just leave it on all the time. POHB napisał(a): > Also, if the camera has built-in stabilization under what circumstances > is it a bad idea to use it? Mine has an off switch but I can't think > why not to just leave it on all the time. -energy consumption Someday, someone will invent 100.000mAh camera batteries and then we will be able to use anti-shake constantly without worrying about power supply. Maybe it's not a problem if you just walk around in your city and shoot some photos. But on a long trips power saving is crucial. -tripod Stabilization in lens can work impropoperly when camera is mounted on a tripod. I suppose sensor stab. will as well try to reduce shakes, which don't exist. -a lot of low-iso shooting IIS produces heat, which can increase sensor noise. > POHB napisał(a): >> Also, if the camera has built-in stabilization under what circumstances >> is it a bad idea to use it? Mine has an off switch but I can't think >> why not to just leave it on all the time. > > -energy consumption i used canon S1 and later S2 with IS constantly on and still be able to shoot over 500 shots with one charge, so, i'd say it's not that of a problem... > i used canon S1 and later S2 with IS constantly on and still be able to > shoot over 500 shots with one charge, so, i'd say it's not that of a > problem... Yes, EC is a minor problem but unfortunately I often ran out of power at the end of a day (I own S2 too), when on a trip. Was using 2500mAh NiMH. I think turning off IS or setting it to "shoot only" mode could allow me to shoot some more photos. But all in all it's lesser problem than heat production. >> i used canon S1 and later S2 with IS constantly on and still be able to >> shoot over 500 shots with one charge, so, i'd say it's not that of a [quoted text clipped - 5 lines] > shoot some more photos. > But all in all it's lesser problem than heat production. I never really tried with IS off and shoot from full bat to flat... Also never had heat problems, since i guess you shoot a lot more than i did... So, i'd say solution to your first problem would be a spare set of batteries... > I never really tried with IS off and shoot from full bat to flat... > Also never had heat problems, since i guess you shoot a lot more than i > did... Heat problem in S2 doesn't exist, 'cause S2 AFAIK uses optical stab. Pentax & Sonolta use sensor stab. I think this amount of heat can't do anything wrong to lens, but can inrease sensor noise. And this "camera overheated" [direct translation, haven't seen it on english lang. set] error in Sonolta A100... :) > So, i'd say solution to your first problem would be a spare set of > batteries... I had spare set then... and they were recharging. :) Damn, I have to think about faster charger... >> I never really tried with IS off and shoot from full bat to flat... >> Also never had heat problems, since i guess you shoot a lot more than i [quoted text clipped - 4 lines] > overheated" [direct translation, haven't seen it on english lang. set] > error in Sonolta A100... :) that could be the problem, i guess. However, some camera models turns on sensor stab only when half-press shutter, so this would minimize heating. >> So, i'd say solution to your first problem would be a spare set of >> batteries... > > I had spare set then... and they were recharging. :) Damn, I have to think > about faster charger... yep...i know what you mean... > POHB napisal(a): > Stabilization in lens can work impropoperly when camera is mounted on a > tripod. I suppose sensor stab. will as well try to reduce shakes, which > don't exist. Surely not? You reckon in the absence of any real shaking the system will imagine some and start moving the sensor around randomly? I'll have to try an experiment. > Surely not? You reckon in the absence of any real shaking the system > will imagine some and start moving the sensor around randomly? I'll > have to try an experiment. I'm not sure how will sensor stab. act, but it is recommended to turn off IS and VR when camera is mounted on tripod. Since AntiShake uses the same mechanism for detecting shakes, I suppose sensor stab. can also get confused. But if you can - try it. I'm courious about it. > I'm not sure how will sensor stab. act, but it is recommended to turn > off IS and VR when camera is mounted on tripod. Actually, at least newer IS detects the tripod. And then not every tripod is 100% unmoving (think stronger wind). -Wolfgang > > Surely not? You reckon in the absence of any real shaking the system > > will imagine some and start moving the sensor around randomly? I'll [quoted text clipped - 5 lines] > confused. > But if you can - try it. I'm courious about it. OK I just tried it. Here's the results: http://www.flickr.com/photos/96236359@N00/sets/72157594334734703/detail/ They're all taken with a 1/2 second exposure and a 300mm lens to get a good ol shake going. As expected there's a huge difference with stab. on and off when handheld I guess my tripod isn't so great because there's still a lot of shake with the camera on it if I just press the shutter and the stab. still makes a big difference. Even if I use the 2 second delay mode that locks the mirror up and lets it stop bouncing around there's still a fair amount of shake with the stab. turned off. Finally if I use the 10s self-timer, despite not having lock-up the stab. doesn't seem to have much effect. So my conclusion? I'll leave the stab. on when using a tripod, and think about getting a remote shutter release. >> > Surely not? You reckon in the absence of any real shaking the system >> > will imagine some and start moving the sensor around randomly? I'll [quoted text clipped - 9 lines] >Here's the results: >http://www.flickr.com/photos/96236359@N00/sets/72157594334734703/detail/ I turn the IS off on my cannon lenses when on the tripod and using the remote shutter release. I've been told on the tripod it can get confused and the image can end up a very slight amount soft. That information might be wrong but I figure if I'm on a tripod I'm there looking for a sharp as a tack shot so why take the chance. And the IS is certainly not helping in that situation. But I have to admit on the 70-200 f2.8 IS USM the IS does make a difference for me. I can see it vancel out some of the shake of my daughter pulling my shirt to get my attention as I try and get a shot. >They're all taken with a 1/2 second exposure and a 300mm lens to get a >good ol shake going. [quoted text clipped - 11 lines] >So my conclusion? I'll leave the stab. on when using a tripod, and >think about getting a remote shutter release. > I turn the IS off on my cannon lenses when on the tripod and using the > remote shutter release. I've been told on the tripod it can get > confused and the image can end up a very slight amount soft. > That information might be wrong but I figure if I'm on a tripod I'm > there looking for a sharp as a tack shot so why take the chance. > And the IS is certainly not helping in that situation. That's only true with the older IS versions, i.e. 75-300 f4.5-5.6, 28-135 f3.5-5.6 and 100-400 f4.5-5.6. Later ones, particularly the 70-200, compensate for the 'pod. I only turn my IS off when I'm in the studio. A gust of wind can move a camera enough to throw off the shot with a long lens, and I shoot from bridges and docks alot, and they ain't the most stable platform in the world. SignatureSkip Middleton www.shadowcatcherimagery.com www.pbase.com/skipm >>>Surely not? You reckon in the absence of any real shaking the system >>>will imagine some and start moving the sensor around randomly? I'll [quoted text clipped - 9 lines] > Here's the results: > http://www.flickr.com/photos/96236359@N00/sets/72157594334734703/detail/ What sort of tripod? 1/2 second & 300mm is long, thanks for the demonstration. > They're all taken with a 1/2 second exposure and a 300mm lens to get a > good ol shake going. [quoted text clipped - 11 lines] > So my conclusion? I'll leave the stab. on when using a tripod, and > think about getting a remote shutter release. SignaturePaul Furman http://www.edgehill.net/1 Bay Natives http://www.baynatives.com > What sort of tripod? > 1/2 second & 300mm is long, thanks for the demonstration. Can't remember the brand and don't have it in front of me right now but it's a lightweight model I bought about 12 years ago. I think the legs are pretty stable but then you crank up a "stalk" at the top (technical term?) and that is a bit springy. I reckon it'd be better with the stalk wound right down so the camera was at the top of the triangle formed by the legs, but then it wouldn't be at a very convenient height for the viewfinder. By the way, in case it wasn't obvious, those images were 1:1 crops of a teeny bit of the picture. > it's a lightweight model I bought about 12 years ago. > I think the legs are pretty stable but then you crank up a "stalk" at > the top (technical term?) and that is a bit springy. A bit ... a rather strong bit, seeing your test images. > I reckon it'd be > better with the stalk wound right down so the camera was at the top of > the triangle formed by the legs, but then it wouldn't be at a very > convenient height for the viewfinder. Buy a larger, heavier, more stable (dare I say carbon) model. :-) Of course, that sort of thing costs quite a sum. On the other hand: you can always bow down. -Wolfgang >Also, if the camera has built-in stabilization under what circumstances >is it a bad idea to use it? Mine has an off switch but I can't think >why not to just leave it on all the time. When the camera has no information (or bad information) about the focal length. This could be because you've fitted a lens that doesn't "talk" to the camera, or maybe because an auxiliary lens is attached. >> Has anyone seen any real world testing that shows how well IIS works with telephoto lenses? > > Also, if the camera has built-in stabilization under what circumstances > is it a bad idea to use it? Mine has an off switch but I can't think > why not to just leave it on all the time. > ... Stabilization algorithms used in these cameras are normally optimized for taking steady motionless shots. Once you need to take shots while _tracking_ your subject with the lens (i.e. when you have to move your camera intentionally) the algorithm might get confused. Especially if the subject's motion is... well, tricky, i.e. its angular speed or direction is changing. For this reason Canon lenses, for example, can switch from steady IS mode to tracking IS mode. I don't know whether your camera has something like this, but if it doesn't, then turning IS off for tracking shots might prove to be a good idea. -- Best regards, Andrey Tarasevich A colleague sold a Nikon 300/2.8 to us and bought same with IIS. He lately said that he had rather kept the old one. -Michael >A colleague sold a Nikon 300/2.8 to us and bought same with IIS. He lately >said that he had rather kept the old one. > > -Michael did he say why? Above statement clearly says that IS wasn't the problem, since if it was, he would simply shut it off. SO, what was the problem? > did he say why? Above statement clearly says that IS wasn't the problem, > since if it was, he would simply shut it off. IS means more glass in the lens. More surfaces for flare to develop. Less sharpness, possibly (again: more surfaces, more elements ...) More weight, as well. All that cannot be simply shut off. (Though I think a good IS is worth lots of money.) -Wolfgang [] > IS means more glass in the lens. More surfaces for flare > to develop. Less sharpness, possibly (again: more surfaces, [quoted text clipped - 5 lines] > > -Wolfgang Why does IS mean more glass in the lens? David > [] >> IS means more glass in the lens. More surfaces for flare [quoted text clipped - 10 lines] > > David I doubt that it means that. I see no reason why it wouldn't be possible to move just one lens already present in the lens. In theory, any lens you move you move image, so really doesn't matter which one, i guess the one which needs to be moved minimum... > "David J Taylor" > <david-taylor@blueyonder.co.not-this-bit.nor-this-part.uk> >> Why does IS mean more glass in the lens? >> [quoted text clipped - 4 lines] > move you move image, so really doesn't matter which one, i guess the one > which needs to be moved minimum... At least as far as Canon's IS goes, it does mean an extra element. And I would venture a guess that if you were to move an element already present in a non IS version, you would alter the optical features in a probably undesirable way... SignatureSkip Middleton www.shadowcatcherimagery.com www.pbase.com/skipm [] >> I doubt that it means that. I see no reason why it wouldn't be >> possible to move just one lens already present in the lens. In >> theory, any lens you move you move image, so really doesn't matter >> which one, i guess the one which needs to be moved minimum... > At least as far as Canon's IS goes, it does mean an extra element. And I > would venture a guess that if you were to move an element > already present in a non IS version, you would alter the optical > features in a probably undesirable way... I venture to suggest that, rather than it being essential just to the IS function, Canon chose to add the extra element or re-organise the groups to produce a little better optical performance so that they could justify marketing the lens at an ever higher price... www.pbase.com/skipm > [] >>> I doubt that it means that. I see no reason why it wouldn't be [quoted text clipped - 11 lines] > to produce a little better optical performance so that they could justify > marketing the lens at an ever higher price... Except that, as far as I know, no IS version of any lens actually outperforms its non IS counterpart. And, since more elements add to the complication, I don't see why Canon would add an element suspended by gyros just to "produce a little better optical performance," especially when they did not achieve that result. SignatureSkip Middleton www.shadowcatcherimagery.com [] > Except that, as far as I know, no IS version of any lens actually > outperforms its non IS counterpart. And, since more elements add to > the complication, I don't see why Canon would add an element > suspended by gyros just to "produce a little better optical > performance," especially when they did not achieve that result. Fair enough, I haven't had occasion to look into the particular manufacturer's range, and compare the IS and non-IS variants. David On 20 Oct 2006 15:05:46 +0200, "Protoncek \(ex. SleeperMan\)" <protoHAHAHncek@bonbon.net> wrote: >> Why does IS mean more glass in the lens? Surely it means a DIFFERENT lens design. Not necessarily more or fewer elements. > Why does IS mean more glass in the lens? Which element of a lens would you move how much to correct a left-right or up-down vibration, so that every point of the image stays on the same point of the sensor? Remember, you have to do that _fast_ and for _every_ point. Currently we have at least the following 4 techniques: - use a sturdy tripod or similar - use some (large, fast spinning) gyroscopes to increase the inertial resistance against tilting (you'll find that sort of technique in some ocean liners, to stop them from rolling) - move the sensor and accept that you neither stabilize the viewfinder view (and that tilting will shrink the image dimensions a bit) - move a corrective element in the lens (as Canon and Nikon do) to get the same image on the sensor. OK, putting the lens (and attached camera) onto an anti-grav carrier and making it's apparent inertia near infinite would be a nice technical feat. :-) -Wolfgang > David J Taylor > <david-taylor@blueyonder.co.not-this-bit.nor-this-part.uk> wrote: [quoted text clipped - 23 lines] > > -Wolfgang Wolfgang, You seem to be assuming that you could not move some element or group of an existing system. OK, I guess it depends on the exact design of the optics, but I don't see why it should not be possible. Moving the appropriate element or group laterally would laterally displace the focal plane image - a simple X-Y movement suffices. Lateral movement would be unlikely to affect the magnification (and a small magnification change would likely be far more acceptable than the image shift with no IS. I remain unconvinced that extra glass is required in a long lens. David [on IS] > Wolfgang, > You seem to be assuming that you could not move some element or group of > an existing system. I do assume that no yet known optical formula can do that, I assume that you may be able to add elements to such a formula to provide for special, moving correction elements. > OK, I guess it depends on the exact design of the > optics, but I don't see why it should not be possible. You could certainly try and reduce the number of lenses and/or elements. After all, having just a few elements was the way of the world before anti-reflective coating was available. Using more elements and more glass does provide advantages, though, that's why we have so many of them nowadays. > Moving the appropriate element or group laterally would laterally displace > the focal plane image - a simple X-Y movement suffices. Lateral movement > would be unlikely to affect the magnification (and a small magnification > change would likely be far more acceptable than the image shift with no > IS. Obviously, today's IS only dampens vibrations orthogonal to the lens axis, not movement. Moving many lens groups is much more complicated than moving just one (and possibly not commercial viable today), and whichever lens group you move _is_ the IS group. Adding another funtion to that group is possible, but would be yet another compromise. Reducing the overall number of elements is also possible, but would again compromise IQ compared to a non-IS version with the same amounts of elements. > I remain unconvinced that extra glass is required in a long lens. I remain convinced that while there may be much better (and maybe even in the end cheaper) ways for IS, they are *not* here today (outside the military & co), and probably are not even commercial viable today. Think of the IQ from the Mars rovers --- and ask if you can sell a mere megapixel chip (without even a body) to the general for, well, astronomical prices, above Canon's most expensive (35mm) camera bodies ... -Wolfgang >> You seem to be assuming that you could not move some element or group of >> an existing system. > > I do assume that no yet known optical formula can do that, I > assume that you may be able to add elements to such a formula > to provide for special, moving correction elements. You don't need to have a special element. Any element with overall power will do. By overall power, I mean it has to either magnify or minify ... not just introduce aberration without power. The major question is whether moving an element of an existing design will do the image motion without introducing a very bad amount of off-axis aberration. Moving an element with power to shift the image causes astigmatism. But remember ... any aberration is better than having the image blurred. Doug McDonald >>> You seem to be assuming that you could not move some element or group of >>> an existing system. >> I do assume that no yet known optical formula can do that, I >> assume that you may be able to add elements to such a formula >> to provide for special, moving correction elements. > You don't need to have a special element. Please reread. You designate an element for IS == you make that element movable orthogonally to the lens axis or tiltable or whatever == that element is special _because_ it can move that way. > But remember ... any aberration is better than having the > image blurred. Do _you_ get aberrations when using IS? -Wolfgang >> You don't need to have a special element. > > Please reread. You designate an element for IS == you make > that element movable orthogonally to the lens axis or > tiltable or whatever == that > element is special _because_ it can move that way. Ah ... I see what you mean. Yes, it is "the" special element for that reason. But optically, it need not be special or extra. Doug McDonald > David J Taylor > <david-taylor@blueyonder.co.not-this-bit.nor-this-part.uk> wrote: [quoted text clipped - 47 lines] > > -Wolfgang I am /not/ talking about Mars rovers, military systems, or anything that complicated! Simply that to achieve basic IS you don't need to add extra elements. I would accept that having off-axis elements may mean that you need to adjust the lens design to keep aberrations under control, though. Clearly, that would depend on the exact lens design. David > I am /not/ talking about Mars rovers, military systems, or anything that > complicated! Simply that to achieve basic IS you don't need to add extra > elements. Sure, you can do that. Just use a lens baby and some gyros to control it's anti-shake movements. However, you will _not_ get image quality. > I would accept that having off-axis elements may mean that you > need to adjust the lens design to keep aberrations under control, though. > Clearly, that would depend on the exact lens design. I would really like to see that lens design of yours. -Wolfgang > David J Taylor > <david-taylor@blueyonder.co.not-this-bit.nor-this-part.uk> wrote: [quoted text clipped - 14 lines] > > -Wolfgang Wolfgang, You are blowing this out of all proportion. Whether or not you need to add extra elements depends on the lens design, as others have confirmed. I accept that you /may/ need to add extra elements. David > ... > OK, putting the lens (and attached camera) onto an anti-grav > carrier and making it's apparent inertia near infinite would be a nice > technical feat. :-) > ... That's not as easy as it seems at the first sight :) How would you focus a lens or press a shutter release button when they have near infinite inertia? -- Best regards, Andrey Tarasevich >> ... >> OK, putting the lens (and attached camera) onto an anti-grav >> carrier and making it's apparent inertia near infinite would be a nice >> technical feat. :-) >> ... > That's not as easy as it seems at the first sight :) How would you focus a lens > or press a shutter release button when they have near infinite inertia? You would have to be selective there; and it certainly is enough to raise the inertia of the lens barrel (on non-extending lenses). The camera and the internal working of the lens are just hanging there. Of course, if you are ambitious, just put the inertia into a tripod plate. So to say, a portable tripod. Of course you'd need to rely on the ruggedness of camera and lens ... just like today. But these are all tiny details, which can easily be overcome. -Wolf'though probably not in my lifetime'gang > David J Taylor <david-taylor@blueyonder.co.not-this-bit.nor-this-part.uk> > wrote: [quoted text clipped - 6 lines] > > Remember, you have to do that _fast_ and for _every_ point. true. But you remember one other thing...IS costs. Not a few, but rather a lot. Your question explains WHY this thing costs so damn much. Because very precise calculations are needed. >> David J Taylor >> <david-taylor@blueyonder.co.not-this-bit.nor-this-part.uk> wrote: [quoted text clipped - 10 lines] > rather a lot. Your question explains WHY this thing costs so damn > much. Because very precise calculations are needed. It doesn't cost that much in non-SLR lenses. Ever felt you are paying too much? David > It doesn't cost that much in non-SLR lenses. > Ever felt you are paying too much? > > David Remember, you are 1) moving a much smaller mass, 2) moving it a much smaller distance, and 3) possibly moving it for a less discerning audience. SignatureSkip Middleton www.shadowcatcherimagery.com www.pbase.com/skipm >> It doesn't cost that much in non-SLR lenses. >> Ever felt you are paying too much? [quoted text clipped - 4 lines] > smaller distance, and 3) possibly moving it for a less discerning > audience. Is the extra cost of a larger motor all that much greater? And in the non-DSLR lens you have to be much more precise. I accept that some manufacturers may feel that (3) is justified. But perhaps the real difference is in the quantities produced? David >>> It doesn't cost that much in non-SLR lenses. >>> Ever felt you are paying too much? [quoted text clipped - 11 lines] > > David "But perhaps the real difference is in the quantities produced?" would be 4). (grin) But it's not just the cost of the motor, although that would be a factor. It would also be the cost of the additional element(s). And, since the element(s) are larger, that's actually the one that needs to be more precise. A small variation in a small movement would be far less noticeable than the same movement, in terms of degrees, in a larger element. SignatureSkip Middleton www.shadowcatcherimagery.com www.pbase.com/skipm > - use some (large, fast spinning) gyroscopes to increase the > inertial resistance against tilting (you'll find that sort of > technique in some ocean liners, to stop them from rolling) No, no, the gyroscopes never have to be "large". They are there just to provide a reference direction (or all three!) and not to provide stability directly. The precession effect means that would not be a sensible thing to do, with liners or with cameras. In the case of liners I believe that the force to oppose the action of the waves comes from vanes like those used in submarines, and the vanes are moved by actuators that are controlled by movement relative to the gyroscopes. This is analogous the the movement of optical components or sensor in a camera. I don't think that anyone has mentioned that in the case of a camera you will still want to be able to pan the camera to follow a moving subject (and perhaps get an intentional blur on the background). The camera may attempt to counteract this unless it is smart enough to tell the difference between panning and twitching! The same can happen with IS binoculars, which are perhaps even more likely to be panned. Mike. SignatureIf reply address = connectfee, add an r because it is free not fee. >> - use some (large, fast spinning) gyroscopes to increase the >> inertial resistance against tilting (you'll find that sort of >> technique in some ocean liners, to stop them from rolling) > No, no, the gyroscopes never have to be "large". They are there just to > provide a reference direction You mis-understood me. I was _not_ thinking of reference data like in inertial navigation systems. I was thinking of usung the brute force of large gyroscopes to, well, 'spin-stabilize' your lens. > In the case of liners I believe that the > force to oppose the action of the waves comes from vanes like those used in > submarines, and the vanes are moved by actuators that are controlled by > movement relative to the gyroscopes. Yes, that technique is also used, as well as using ballast water movement. However, ships tend to roll much worse from side to side than they pitch the nose/stern up and down. By using a vertical spin axis, a spinning gyro couples rolling and pitching, and using the ships resistance to pitching to reduce rolling. Best of all: it needs no computers, no special care, just a bit of room, a few sturdy attachment to the hull and some energy for spinning up ... it's a completely passive device, otherwise. > I don't think that anyone has mentioned that in the case of a camera you > will still want to be able to pan the camera to follow a moving subject (and > perhaps get an intentional blur on the background). Works with quite a few ISses ... -Wolfgang >> - use some (large, fast spinning) gyroscopes to increase the >> inertial resistance against tilting (you'll find that sort of >> technique in some ocean liners, to stop them from rolling) > No, no, the gyroscopes never have to be "large". They are there just to > provide a reference direction (or all three!) and not to provide stability > directly. The precession effect means that would not be a sensible thing to > do, with liners or with cameras. Here is what I meant, note that the gyroscopes are large and precession does not seem to be a problem at all: http://www.ken-lab.com/stabilizers.html -Wolfgang > [] > [quoted text clipped - 9 lines] > > Why does IS mean more glass in the lens? IS/VR lenses have a shiftable glass element added to effect motion correction. Signature-- r.p.e.35mm user resource: http://www.aliasimages.com/rpe35mmur.htm -- r.p.d.slr-systems: http://www.aliasimages.com/rpdslrsysur.htm -- [SI] gallery & rulz: http://www.pbase.com/shootin -- e-meil: Remove FreeLunch. >> [] >> [quoted text clipped - 12 lines] > IS/VR lenses have a shiftable glass element added to effect motion > correction. But you could shift an /existing/ element to achieve the IS effect, I don't think you /need/ to add one. David >> IS/VR lenses have a shiftable glass element added to effect motion >> correction. > But you could shift an /existing/ element to achieve the IS effect, I > don't think you /need/ to add one. So why is nobody doing that? -Wolfgang >>> IS/VR lenses have a shiftable glass element added to effect motion >>> correction. [quoted text clipped - 3 lines] > > So why is nobody doing that? I think both those alternatives are naively over-simplified. They don't take a non-IS lens and add or mobilize some element. Signature--Bryan >>IS/VR lenses have a shiftable glass element added to effect motion >>correction. > > But you could shift an /existing/ element to achieve the IS effect, I > don't think you /need/ to add one. If you dig around the Canon site I think you find an illustration and you will see that it is in fact a double concave lens element that is added. It does not tilt but shift in x,y where z is the lens axis. THe Nikon solution is the same, I believe. If you shift other /usual/ elements, then I believe you'd need also to tilt the film/sensor plane at the same time to keep the image flat. Cheers, Alan Signature-- r.p.e.35mm user resource: http://www.aliasimages.com/rpe35mmur.htm -- r.p.d.slr-systems: http://www.aliasimages.com/rpdslrsysur.htm -- [SI] gallery & rulz: http://www.pbase.com/shootin -- e-meil: Remove FreeLunch. >>> IS/VR lenses have a shiftable glass element added to effect motion >>> correction. [quoted text clipped - 12 lines] > Cheers, > Alan Thanks, Alan. I haven't seen those illustrations, and they obviously reflect what is required in a practical design to maintain the highest image quality, even though the IS effect /alone/ doesn't require them. Cheers, David > IS means more glass in the lens. More surfaces for flare > to develop. Less sharpness, possibly (again: more surfaces, > more elements ...) Not necessarily. There is no need for a separate element to do the stabilization with. Certain preexisting designs may not have a useful element, of course. But to assume that after a full redesign any optical quality need be lost is silly. It MAY, but except for certain designs (for example, an f/1.2 simple 6 element double Gauss normal lens) it probably won't. Doug McDonald >> IS means more glass in the lens. More surfaces for flare >> to develop. Less sharpness, possibly (again: more surfaces, >> more elements ...) > Not necessarily. There is no need for a separate element to > do the stabilization with. I ... disagree. You simply _have_ to asign one or more lens groups to move, and that are the IS element(s), by definition. Giving them also other functions is ... at least going to be very complicated and, of course, will be a compromise again. > Certain preexisting designs may > not have a useful element, of course. But to assume that > after a full redesign any optical quality need be lost is > silly. The more lenses, the more reflections, coatings help, but do not remove the basic problem. The more lenses, the more errors due to physical facts like tolerances of grinding machines. IS-groups do in no way improve the image as such --- they only _reduce_ the effect of camera vibrations on IQ. > It MAY, but except for certain designs (for example, > an f/1.2 simple 6 element double Gauss normal lens) it > probably won't. If you are that good a lens designer, Canon, Nikon, Sigma, Tamron, Tokina, Zeiss, Leica and many more lens producing companies will be paying you big bucks. They certainly would want less, easier elements and assembly (i.e. much cheaper, lighter lenses) with superior image quality. -Wolfgang >>> IS means more glass in the lens. More surfaces for flare >>> to develop. Less sharpness, possibly (again: more surfaces, [quoted text clipped - 4 lines] > > I ... disagree. YOU ARE WRONG You simply _have_ to asign one or more lens > groups to move, and that are the IS element(s), by > definition. That is of course perfectly correct. > Giving them also other functions is ... at least > going to be very complicated and, of course, will be a > compromise again. The moving lens HAS to have another function, unless you use prisms at a point where the internal beam is focused at infinity. BUT, and you seem incapable of understanding this, it need not result in any other compromise. Especially in long teles moving just about any element will do, optically. Sure, you so want a light weight one, and that is the constraint you may be thinking of. >> Certain preexisting designs may >> not have a useful element, of course. But to assume that [quoted text clipped - 3 lines] > The more lenses, the more reflections, coatings help, but do > not remove the basic problem. But you may not need another element. If you do, and choose the design well, it WILL reduce aberration. > The more lenses, the more errors due to physical facts like > tolerances of grinding machines. IS-groups do in no way improve > the image as such --- they only _reduce_ the effect of camera > vibrations on IQ. The last statement may well be false. If designed properly, the element use for IS **WILL** reduce aberration. That's simply the way lens design works. The only optical downside of using a new element is one you do recognize: possible flare. > If you are that good a lens designer, Canon, Nikon, Sigma, Tamron, > Tokina, Zeiss, Leica and many more lens producing companies > will be paying you big bucks. They certainly would want less, > easier elements and assembly (i.e. much cheaper, lighter lenses) > with superior image quality. They have VERY good designers, and you can be sure that they will get an optical aberration improvement out of adding an extra element. Doug McDonald >>>> IS means more glass in the lens. More surfaces for flare >>>> to develop. Less sharpness, possibly (again: more surfaces, >>>> more elements ...) >>> Not necessarily. There is no need for a separate element to >>> do the stabilization with. >> I ... disagree. > YOU ARE WRONG If you say so. >> You simply _have_ to asign one or more lens >> groups to move, and that are the IS element(s), by >> definition. > That is of course perfectly correct. See? >> Giving them also other functions is ... at least >> going to be very complicated and, of course, will be a >> compromise again. > The moving lens HAS to have another function, unless > you use prisms at a point where the internal beam is > focused at infinity. You know that Canon uses VAP (Vari-Angle Prisms) for it's CamCorders? http://www.canon.com/technology/canon_tech/explanation/hvision_lens.html However, their EOS-lenses use a moving lens group. So, pray tell, which "other function" do they have? > BUT, and you seem incapable of understanding this, > it need not result in any other compromise. And you seem not to be getting the point that _everything_ in photography is compromise. ISO 100, 400 or 1600 --- compromise. Zoom versus non-zoom: compromise. IS versus non-IS: compromise (and even ignoring the question of image quality, the price for IS lenses _is_ higher, and current ones are heavier). > Especially > in long teles moving just about any element will do, > optically. Sure, you so want a light weight one, and > that is the constraint you may be thinking of. Ok, let's see your optical formula(s) where you can move almost _any_ elements. >>> Certain preexisting designs may >>> not have a useful element, of course. But to assume that >>> after a full redesign any optical quality need be lost is >>> silly. >> The more lenses, the more reflections, coatings help, but do >> not remove the basic problem. > But you may not need another element. I hear you saying that a lot. Why, do you suggest, is not Nikon or Sigma or someone else delivering cheaper, yet as good IS variants, undercutting Canon (who do it the hard, expensive way)? > If you do, and choose > the design well, it WILL reduce aberration. Will it? Even when moving? And what about the reflections _I_ mentioned and _you_ ignored? >> The more lenses, the more errors due to physical facts like >> tolerances of grinding machines. IS-groups do in no way improve >> the image as such --- they only _reduce_ the effect of camera >> vibrations on IQ. > The last statement may well be false. A million dollars may well be mine by tomorrow. Please compare current IS and non-IS versions of the same lens. Show me where the IS version on a stable tripod delivers higher IQ. > If designed properly, > the element use for IS **WILL** reduce aberration. That's > simply the way lens design works. The more lenses, the less aberrations? Somehow I don't buy that. > The only optical downside > of using a new element is one you do recognize: possible flare. Oh, so there cannot be hazing due to stray light and 100% of the light of the scene entering a lens will come out on the other side? Quick, someone tell the film people to stop messing around with t-stops!! >> If you are that good a lens designer, Canon, Nikon, Sigma, Tamron, >> Tokina, Zeiss, Leica and many more lens producing companies >> will be paying you big bucks. They certainly would want less, >> easier elements and assembly (i.e. much cheaper, lighter lenses) >> with superior image quality. > They have VERY good designers, and you can be sure that they > will get an optical aberration improvement out of adding an > extra element. So, what is your claim? You need no extra element for IS; move just any old element you like, and use tons and tons of extra elements as they reduce optical aberrations? -Wolfgang > Ok, let's see your optical formula(s) where you can move almost > _any_ elements. Tessar or Cooke triplet used as a long (telephoto) lens. >> If you do, and choose >> the design well, it WILL reduce aberration. > > Will it? yes, if designed by a competant designer. > Even when moving? no, when moved off center it will cause astignmatism, but, as I said before, this will be very small, and will be much smaller then the shake that is being corrected. > And what about the reflections > _I_ mentioned and _you_ ignored? If the designer is forced to insert an extra element, yes, you are perfectly correct about flare, etc. >> If designed properly, >> the element use for IS **WILL** reduce aberration. That's >> simply the way lens design works. > > The more lenses, the less aberrations? Somehow I don't buy that. Yes, if designed by a competant designed, yes, once again, it really IS true that more elements will redude aberrations. > So, what is your claim? You need no extra element for IS; move > just any old element you like, and use tons and tons of extra > elements as they reduce optical aberrations? Not every preexisting design will have a suitable movable element, especially things like 50 mm f 1.4 prime lenses. Most zooms will. And yes, tons and tons of extra elements will reduce aberrations, given a compentant designer. As you also claim, correctly, they will also cause flare problems. Doug >>did he say why? Above statement clearly says that IS wasn't the problem, >>since if it was, he would simply shut it off. [quoted text clipped - 3 lines] > more elements ...) > More weight, as well. MTF tests of IS lenses show a minor degredation in sharpness v. similar non IS designs. However, I would bet that any extra flare is not detectable in an ordinary image with obvious flare in it. Likewise contrast. Signature-- r.p.e.35mm user resource: http://www.aliasimages.com/rpe35mmur.htm -- r.p.d.slr-systems: http://www.aliasimages.com/rpdslrsysur.htm -- [SI] gallery & rulz: http://www.pbase.com/shootin -- e-meil: Remove FreeLunch. On 19 Oct 2006 21:55:00 +0200, in rec.photo.digital.slr-systems "Protoncek \(ex. SleeperMan\)" <protoHAHAHncek@bonbon.net> wrote: >>A colleague sold a Nikon 300/2.8 to us and bought same with IIS. He lately >>said that he had rather kept the old one. [quoted text clipped - 3 lines] >did he say why? Above statement clearly says that IS wasn't the problem, >since if it was, he would simply shut it off. SO, what was the problem? Sure he wasn't talking about the 300 f/4 vs the 300 f/2.8 VR. The later will be a fair bit larger, heavier etc. So simply turning VR off doesn't give you the same as the other lens. SignatureEd Ruf (Usenet2@EdwardG.Ruf.com) http://edwardgruf.com/Digital_Photography/General/index.html > On 19 Oct 2006 21:55:00 +0200, in rec.photo.digital.slr-systems "Protoncek > \(ex. SleeperMan\)" <protoHAHAHncek@bonbon.net> wrote: [quoted text clipped - 11 lines] > will be a fair bit larger, heavier etc. So simply turning VR off doesn't > give you the same as the other lens. those two would be uncompareble, since they are not the same... lens to compare would be, say canon 70-300 IS USM and 70-300 USM, then 70-200 f4 L IS and 70-200 f4 L...etc... The stabilization seems to do more damage than good. -Michael > The stabilization seems to do more damage than good. > > -Michael another stupid comment. Would you reply such statements with some facts? I'd say you have non-IS lenses... > Would you reply such statements with some facts? Sorry I can't He just said it does not work as expected. -Michael >> Would you reply such statements with some facts? > > Sorry I can't He just said it does not work as expected. > > -Michael that could mean a lot of thigs. Like, say, he expected that moving objects will become frozen. Wrong... IS works just fine if you KNOW what it meant for. And, sure, if you don't expect too much. Like thinking that setting IS to on and exposure to 1 sec and expect that handheld picture will be sharp... And there's more... I know what to expect and for this works even better as expected. > The stabilization seems to do more damage than good. > > -Michael Patently untrue. SignatureSkip Middleton www.shadowcatcherimagery.com www.pbase.com/skipm > The stabilization seems to do more damage than good. Based on the sample of one friend who said he preferred a lens he had without it?? Do you have more than anecdotal evidence? SignatureJohn McWilliams > Do you have more than anecdotal evidence? As I never tried any IS lens myself and (as reported) got nothing more than this short comment by this man, sorry, but no. -Michael >> Do you have more than anecdotal evidence? > > As I never tried any IS lens myself ... That's what i thought... if you don't have any experience with it, then don't comment. Or at least say "i've heard that IS is a piece of sh.t". See? > if you don't have any experience with it, then don't comment. Or at > least say "i've heard that IS is a piece of sh.t". > See? That is what I did in my initial comment: "A colleague sold a Nikon 300/2.8 to us and bought same with IIS. He lately said that he had rather kept the old one. " -Michael >> if you don't have any experience with it, then don't comment. Or at least >> say "i've heard that IS is a piece of sh.t". [quoted text clipped - 6 lines] > > -Michael Nope... you said exactly this: The stabilization seems to do more damage than good. -Michael That's all you wrote. Signaturehttp://www.protoncek.com > The stabilization seems to do more damage than good. That was an interpretation of what I mentioned in the previous post in that thread > That's all you wrote. Wrong. Please go back in the thread and take a look at what I wrote at 10/19/06. -Michael >> The stabilization seems to do more damage than good. > [quoted text clipped - 7 lines] > > -Michael hm... when i go back, i see that i replied with funny comment on the thread when you wrote above statement. i know about your other thread and i replied that one with totally decent "did he say why? Above statement clearly says that IS wasn't the problem, since if it was, he would simply shut it off. SO, what was the problem?" ...and never got reply about WHY he said that... OK; never mind anyway... Signaturehttp://www.protoncek.com > ...and never got reply about WHY he said that... I did reply: "Sorry I can't He just said it does not work as expected." ------- Sorry that I _still_ can't supply any valuable information. BTW.: in my newsreader all this is in a single thread. -Michael >> ...and never got reply about WHY he said that... > > I did reply: > > "Sorry I can't He just said it does not work as expected." hm....that's not an answer, to be exact. Like i said before, maybe he expected that IS will compensate object moving, for one...it's just you can't blaim IS until you're 100% sure. I've had 3 IS cameras and everwhere IS worked excellent. IF (and i say again, IF) you KNOW what IS is ment for and what to expect - definitely NOT miracles, since nothing will make moving object still, and IS is not made for it either. It's made for small hand shake compensations. And for that, it rocks. When you'll have a chance, try out by yourself, bearing in mind what i said and you'll change your opinion. But possibly not at your friends... :-) If possible, take some long zoom lens (like 70-300 IS), zoom all the way and half-press the shutter...holding camera in your hand, not on tripod, sure...and you'll be amazed. It can't be that hard to ask someone to quick peek through his/her camera, can it? > ------- > [quoted text clipped - 3 lines] > > -Michael Signaturehttp://www.protoncek.com > The stabilization seems to do more damage than good. You obviously have never used any IS system. -Wolfgang > The stabilization seems to do more damage than good. This thread is filled with more than the usual amount of bullshit. I have 4 IS lenses ranging from 28mm to 500 mm f/4 L IS. In the long lens category, I can state that if there is a measurable difference between the 500 f/4 L IS versus non IS it really doesn't matter; same with the 300 f/4 L IS. Stars get focused to a level that most of the brightness is in a single pixel on a Canon 1D Mark II, and the remaining light due to the blur filter. IS allows hand held shots that are essentially impossible without it. E.g.: http://www.clarkvision.com/galleries/gallery.bird/web/road.runner.c11.29.2005.JZ 3F5598.b-700.html The above image was shot beside a friend who was using a 300 mm f/2.8 without IS, and he got no sharp images, despite using a faster shutter speed and shorter focal length. There have been anecdotal reports of a newer IS lens being not as sharp as its non-is version, but I have never seen a side- by-side comparison that proves it. Given a choice of two lenses, 1) IS with a few percent reduction in sharpness, or 2) non-IS a few percent sharper, I would always choose the IS. IS helps in so many situations that allows me to get a higher percentage of sharp photos that otherwise I would just throw away, that it more than makes up for perceived loss in specifications on sharpness. I'm not alone in this opinion. Watch the next football game and notice how many beige Canon super telephotos are on the sidelines, and pretty much all are IS. Professional wildlife photographers have been switching from Nikon to Canon (even before digital) for the IS on super telephotos. Personally I hope Nikon comes out with VR on super telephotos soon to give Canon some competition. I migrated up from other lenses and telescopes (I have up to a 2000 mm f/6.3 system) without IS. I will never buy another telephoto lens or any other lens without IS, unless I need the lens and there is no other choice. Roger Clark More photos at: http://www.clarkvision.com > There have been anecdotal reports of a newer IS lens being not as > sharp as its non-is version, but I have never seen a side- > by-side comparison that proves it. On the old photodo site you could compare IS with non IS lenses (of otherwise the same type) and the MTF was slightly worse in the IS lens. How discernible that would be in an actual image is another matter. Fine detail areas perhaps, well magnified. Cheers, Alan Signature-- r.p.e.35mm user resource: http://www.aliasimages.com/rpe35mmur.htm -- r.p.d.slr-systems: http://www.aliasimages.com/rpdslrsysur.htm -- [SI] gallery & rulz: http://www.pbase.com/shootin -- e-meil: Remove FreeLunch. Some useful info on IS: IS - does it work? Is it worth it? http://www.bobatkins.com/photography/tutorials/is2.html The Third Revolution in Lens Design: Stabilization http://www.digitaljournalist.org/issue9805/cameracorner.htm http://en.wikipedia.org/wiki/Image_stabilization http://ca.konicaminolta.com/products/consumer/digital_camera/slr/dynax-7d/02.html http://www.canon.com/bctv/faq/optis.html |
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
|
©2008 Advenet LLC