Pages: [1] 2 3 ... 5   Go Down

Author Topic: Image stabilization  (Read 30973 times)

gkramer

  • Newbie
  • *
  • Offline Offline
  • Posts: 27
Image stabilization
« on: December 15, 2006, 07:26:10 am »

The García-Oliver-Martín-Osuna piece on the Leica M8 was quite interesting (though as a one-time M3 owner, I am disappointed that the widest the M8 can go without an auxiliary finder is only 32mm [full-frame equivalent]; I'll hold out for the M9, which hopefully will get down to 25mm or so).

But I am baffled by their aside about image stabilization, which is probably as revolutionary a development as was TTL metering in its day (which was fiercely resisted by the working pros, with their incident-light meters, just as tripod-addicted pros nowadays are discounting image stabilization). They remark, in passing,

"Shake is another source of quality losses, but Leica lenses (and the best Canon primes) are not stabilized by means of small optic groups, micromotors and computer calculations. The reason is that stabilization affects negatively to the size, luminosity and performance of the lenses (see this excellent comparative analysis of Leica and Olympus zooms by Valentín Sama)."

My Spanish isn't very good and perhaps I've misread Valentin Sama's "excellent comparative analysis," but it seems to be a pretty straightforward comparison of a few hand-held shots taken by each lens.

This probably tell us more about the steadines of Mr. Samas' hands, and his handholding technique, than it does about the effect of Image Stabilization (Nikon call it Vibration Reduction) on image quality. It is no a simple matter to test the latter--even Phil Askey hasn't been able to figured out a meaningful, stadardized way of doing it. Unlike testing an unstabilized lens, where any sturdy tripod can do the job, it would presumably require an elaborate, standardized "vibrating machine" to hold the camera, whose vibration rates and modes could be adjusted to simulate a variety of field conditions. Perhaps some lens makers have such a machine, but it seeme to be a closely held secret, not available to the general public. Too bad, because, as nearly as I can tell, some VR/IS systems are better than others--but until we have a standardized testing protocol, we are at the mercy of manufacturer's advertising claims and antecdotal blogs.

That said, I am frankly doubtful about the García-Oliver-Martín-Osuna statement that IS/VR "affects negatively to the size, luminosity and performance of the lenses." My experience is with Nikon's 70-200mm VR and 200-400mm VR lenses, both or which, with proper technique (a shoulder stock, preferably with a rest (such as a beanbag, or handy fence rail or car-window sill), and using as fast a shutter speed as the light permits) can produce images as sharp and contrasty as any equivalent tripod-mounted unstabilized lens. Not 100%, but as the shutter speed increases, the percentage of "pin-sharp" images increases, approaching 90+ % at 1/180 sec with the 200-400 @ 400mm (with a rest; the percentage would be somewhat lower without a rest, depending on how unsteady the stance, and would be substantailly lower without a shoulder stock). One of the singular adavantages of digital over film is that one can be bracket freely (for sharpness, in this case), shooting a dozen or so shots of an interesting subject, and later cull out the unsharp ones.

Bjørn Rørslett, after testing the Nikon 200-400mm VR, commented that "This might well be the finest telephoto or zoom lens I've ever tested. The image quality delivered by the 200-400 is absolutely marvellous and should put the legendary predecessor MF 200-400 f/4 Nikkor to a deserved rest." He give it a 5, his highest rating (and though he tested it only on the D70 and D2H, in my experience it is also superb with the D2X). He doesn't seem to use one himself (he's evidently a "tripod man," and fond of his unstabilized Zoom-Nikkor 200-400mm), but I think his review can at least be taken to show that the VR's image quality is as quite as good as any of its unstabilized Nikon (or Canon) predecessors.
Logged

feppe

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2906
  • Oh this shows up in here!
    • Harri Jahkola Photography
Image stabilization
« Reply #1 on: December 15, 2006, 08:00:55 am »

Quote
The García-Oliver-Martín-Osuna piece on the Leica M8 was quite interesting (though as a one-time M3 owner, I am disappointed that the widest the M8 can go without an auxiliary finder is only 32mm [full-frame equivalent]; I'll hold out for the M9, which hopefully will get down to 25mm or so).

But I am baffled by their aside about image stabilization, which is probably as revolutionary a development as was TTL metering in its day (which was fiercely resisted by the working pros, with their incident-light meters, just as tripod-addicted pros nowadays are discounting image stabilization). They remark, in passing,

"Shake is another source of quality losses, but Leica lenses (and the best Canon primes) are not stabilized by means of small optic groups, micromotors and computer calculations. The reason is that stabilization affects negatively to the size, luminosity and performance of the lenses (see this excellent comparative analysis of Leica and Olympus zooms by Valentín Sama)."

Do GOMO (yeah I just made that up) claim that IS/VR lens with image stabilization turned OFF results in inferior pictures? Or do they just claim using IS/VR does it? If it's the former, there is a real issue. If it's the latter, it sounds perfectly plausible - something to do with having and eating a cake.

In any case I think the reduced quality with IS/VR on is a non-issue. Most photographers who can afford IS/VR lenses know how to use them (!) and wouldn't turn it on unless necessary. That means that a photo that was taken with IS/VR on would've been unacceptably blurry otherwise. In this case loss of luminosity and some other minor quality aspects should be ok - and they appear to be minor as this is the first time I've heard about this. Of course it is important to realize that IS/VR does affect image quality but I assumed so even before reading your post.

Well, thinking about this further. As IR/VR is available in increasingly many consumer point-and-shoots and might become standard in SLR lenses in a few years we might end up with people just turning it on by default. So assessing just how much image degradation IS/VR results in would be useful. But I'm sure different solutions - and lenses - result in different levels of degradation.

Jonathan Wienke

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 5829
    • http://visual-vacations.com/
Image stabilization
« Reply #2 on: December 15, 2006, 08:07:01 am »

I have the Canon 70-200/2.8L IS, and while some have stated the non-IS version of the lens is slightly sharper, I discount this for 2 reasons.

1. The claimed differences are smaller than the lens-to-lens variation, and aren't apparent to any but the most hardcore pixel-peeping analysis.

2. The benefits of IS far outweigh the claimed sharpness reduction when any camera shake or vibration is present. This includes situations where the camera is tripod mounted in windy conditions.

The only real disadvantage to using IS is slightly shorter battery life.
Logged

John Camp

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2171
Image stabilization
« Reply #3 on: December 15, 2006, 01:50:48 pm »

Not being one of you techy guys, I've never quite understood why you should turn off image stabilization when the camera is mounted on a tripod; and now Jonathan suggests that it works on a tripod in windy conditions. Does this mean IS should only be used when there is vibration present? If so, this suggests a reason that IS lenses might be considered to be less sharp -- I would presume that NO vibration is always better than some vibration, and if you only use IS when there is at least some vibration, then you are dealing with an inherently less-sharp situation. Is that incorrect thinking, and should I be slapped in the face for thinking it?

JC
Logged

howiesmith

  • Guest
Image stabilization
« Reply #4 on: December 15, 2006, 02:21:28 pm »

Quote
Not being one of you techy guys, I've never quite understood why you should turn off image stabilization when the camera is mounted on a tripod; and now Jonathan suggests that it works on a tripod in windy conditions. Does this mean IS should only be used when there is vibration present? If so, this suggests a reason that IS lenses might be considered to be less sharp -- I would presume that NO vibration is always better than some vibration, and if you only use IS when there is at least some vibration, then you are dealing with an inherently less-sharp situation. Is that incorrect thinking, and should I be slapped in the face for thinking it?

JC
[a href=\"index.php?act=findpost&pid=90729\"][{POST_SNAPBACK}][/a]

Caveat:  I do not own nor have I ever used an IS lens.  I have no camera programming experience.  I do not work, nor have I ever, for any optical company.  My only experience with vibration is earthquake responses of structures - typically rather high amplitude and many frequencies, ranging from very low to very high.

Given that, I understand that camera image stabilizers (IS) react to (try to follow without anticipating) detected motion.  If that is correct, there will be natural system frequencies of vibration that the IS can follow OK with minimum error, that will result in even greater "blur" (error) than without IS (amplified vibration) and frequencies where the IS simply doesn't have time to react before the motion is in the opposite direction and the IS is sent there (smooth operation and might as well be off).

I suspect turning off IS while on a tripod is to prolong battery life and perhaps the amplitude and frequencies of vibrations are not within the range of the real effectiveness of IS.  It seems, without any proof or calculations, that IS would be most effective with relatively low frequency vibrations, such as found with hand holding the camera.  To reduce blur due to mirror slap, IS would have to be rather sensitive to higher frequency and lower amplitude vibrations.

I think the best thing would be to try your system (camera, lens, tripod, etc.) with and without IS on, under your expected shooting conditions and use of the images.  In the real world, vibrations are usually the sum of many frequencies at differing amplitudes and without a lot of knowledge of those frequencies and amplitudes and how the IS works (reacts to that input), determining how a certain sytem will react is difficult, if not impossibe.
« Last Edit: December 15, 2006, 02:32:07 pm by howiesmith »
Logged

Tim Gray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2002
    • http://www.timgrayphotography.com
Image stabilization
« Reply #5 on: December 15, 2006, 02:28:16 pm »

Quote
Not being one of you techy guys, I've never quite understood why you should turn off image stabilization when the camera is mounted on a tripod; and now Jonathan suggests that it works on a tripod in windy conditions. Does this mean IS should only be used when there is vibration present? If so, this suggests a reason that IS lenses might be considered to be less sharp -- I would presume that NO vibration is always better than some vibration, and if you only use IS when there is at least some vibration, then you are dealing with an inherently less-sharp situation. Is that incorrect thinking, and should I be slapped in the face for thinking it?

JC
[a href=\"index.php?act=findpost&pid=90729\"][{POST_SNAPBACK}][/a]

The older IS eg on the 100-400 will "hunt" when stabilized on a tripod - the image visibly drifts in the viewfinder.  The newer ones don't.
Logged

Jonathan Wienke

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 5829
    • http://visual-vacations.com/
Image stabilization
« Reply #6 on: December 15, 2006, 07:20:47 pm »

Quote
Not being one of you techy guys, I've never quite understood why you should turn off image stabilization when the camera is mounted on a tripod; and now Jonathan suggests that it works on a tripod in windy conditions. Does this mean IS should only be used when there is vibration present? If so, this suggests a reason that IS lenses might be considered to be less sharp -- I would presume that NO vibration is always better than some vibration, and if you only use IS when there is at least some vibration, then you are dealing with an inherently less-sharp situation. Is that incorrect thinking, and should I be slapped in the face for thinking it?

The 70-200 uses an IS design that senses when the lens is tripod mounted, and uses different vibration dampening techniques (more focused on low-amplitude, high frequencies like wind turbulence and mirror slap than the high amplitude, low frequency vibrations typically encountered while handholding) when on a tripod than when off. Some older lenses use a less intelligent design which twitches and jumps around when the lens is tripod mounted, and IS on these lenses should be turned off when on a tripod. My experience with IS is that it is well worth the additional weight, cost, and battery power in many, if not most shooting conditions.
Logged

aaykay

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 359
Image stabilization
« Reply #7 on: December 16, 2006, 12:42:38 am »

I think the luminosity/sharpness argument is oriented towards the fact that in an IS lens, additional "floating" lens elements (un-needed in an non-IS lens) have to be introduced into the optical structure, as part of the IS mechanism, which in theory can degrade the image.

In case of Canon, I believe the 70-200L f/2.8IS was designed several years after the original 70-200 f2.8 non-IS, and theoretically had improvements built in, being a later design, and is still rated less sharp than the non-IS variant.

Having said the above, I would say that the sharpness trade-off is something a lot of people are willing to live with, when considering the other huge advantages from an IS lens that would enable the obtainment of shots that would otherwise not have been possible.
Logged

Jonathan Wienke

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 5829
    • http://visual-vacations.com/
Image stabilization
« Reply #8 on: December 16, 2006, 04:49:20 am »

Giving up maybe 3% of theoretical maximum sharpness under ideal conditions (sturdy tripod, no wind, etc) to gain 200-300% in real-world, less-than-ideal conditions sounds like a good deal to me. In any case, the IS version of the 70-200 is one of the sharpest zooms you'll find; better than some primes. There's a good reason it's one of the most popular lenses used by portrait and wedding photographers; it's an excellent lens by any standard.
Logged

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
Image stabilization
« Reply #9 on: December 16, 2006, 09:22:28 am »

I'm in the market for the new Canon EF 70-200L f/4 IS. This seems excellent value, lightweight and fast enough to autofocus with a 1.4x extender (on a 20D or 5D).

But I can't find any reviews. Is this lens not available yet?
Logged

gkramer

  • Newbie
  • *
  • Offline Offline
  • Posts: 27
Image stabilization
« Reply #10 on: December 16, 2006, 09:31:10 am »

"I think the luminosity/sharpness argument is oriented towards the fact that in an IS lens, additional "floating" lens elements (un-needed in an non-IS lens) have to be introduced into the optical structure, as part of the IS mechanism, which in theory can degrade the image."

It's true that, other things equal, adding a couple of image-stabilization elements to a lens results in some theoretical loss in light transmission; but other things aren't equal if the entire optical formula is redesigned. Nikon came late to the image-stabilization game (although they were the first to introduce a stabilized lens; but concluded there was no market for it), and offers far fewer stabilized lenses than Canon; but many of Canon's date from the film-camera era, and (one gathers from various blogs), many older Canon lenses can't really cut it with digital. All of Nikon's stabilized lenses were designed for digital, and are highly corrected (particularly for for chromatic abberations, to which digital is particularly prone). This means lots of lens elements (21 for the 70-200mm f2.8 VR, 24 for the 200-400mm f4 VR, plus a drop-in dummy filter that must be kept in place when no real filter is used); and I doubt that the couple of stabilization elements are much of a factor. (If Nikon wanted to minimize the number of elements, they could redesign the 200-400mm to eliminate the drop-in filter, which adds two glass-to-air surfaces.)

The 200-400mm VR is a formidible lens under most lighting conditions; its only real weakness is with backlit subjects, when, even with a good lens hood in place, and no direct light hitting the front element, it is prone to ghosting; but so are most of Nikon's newest, many-element, highly corrected digital lenses (and, one presumes, Canon's as well), whether image-stabilized or not. A price one has to pay to get chromatic abberations under control.
Logged

marcmccalmont

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1780
Image stabilization
« Reply #11 on: December 16, 2006, 01:42:41 pm »

Quote
Caveat:  I do not own nor have I ever used an IS lens.  I have no camera programming experience.  I do not work, nor have I ever, for any optical company.  My only experience with vibration is earthquake responses of structures - typically rather high amplitude and many frequencies, ranging from very low to very high.

Given that, I understand that camera image stabilizers (IS) react to (try to follow without anticipating) detected motion.  If that is correct, there will be natural system frequencies of vibration that the IS can follow OK with minimum error, that will result in even greater "blur" (error) than without IS (amplified vibration) and frequencies where the IS simply doesn't have time to react before the motion is in the opposite direction and the IS is sent there (smooth operation and might as well be off).

I suspect turning off IS while on a tripod is to prolong battery life and perhaps the amplitude and frequencies of vibrations are not within the range of the real effectiveness of IS.  It seems, without any proof or calculations, that IS would be most effective with relatively low frequency vibrations, such as found with hand holding the camera.  To reduce blur due to mirror slap, IS would have to be rather sensitive to higher frequency and lower amplitude vibrations.

I think the best thing would be to try your system (camera, lens, tripod, etc.) with and without IS on, under your expected shooting conditions and use of the images.  In the real world, vibrations are usually the sum of many frequencies at differing amplitudes and without a lot of knowledge of those frequencies and amplitudes and how the IS works (reacts to that input), determining how a certain sytem will react is difficult, if not impossibe.
[a href=\"index.php?act=findpost&pid=90731\"][{POST_SNAPBACK}][/a]

Having designed loudspeakers for a living (one of several jobs) I am familiar with vibration isolation techniques and materials. I have built vibration isolators for both photographic heads and telescope mounts they work well. Decoupling the camera/telescopes from the earth helps. Same with speakers, coupling a loudspeaker to a 300 sg ft suspended floor doesn't sound too good. Damping the resonances from the camera (shutter, button press, wind etc) helps too. Our house was 15 miles from the last earthquake on the Big Island I can attest to the amplitude and spectral content of those vibrations first hand.
Marc
Logged
Marc McCalmont

howiesmith

  • Guest
Image stabilization
« Reply #12 on: December 16, 2006, 02:23:23 pm »

Quote
Having designed loudspeakers for a living (one of several jobs) I am familiar with vibration isolation techniques and materials. I have built vibration isolators for both photographic heads and telescope mounts they work well. Decoupling the camera/telescopes from the earth helps. Same with speakers, coupling a loudspeaker to a 300 sg ft suspended floor doesn't sound too good. Damping the resonances from the camera (shutter, button press, wind etc) helps too. Our house was 15 miles from the last earthquake on the Big Island I can attest to the amplitude and spectral content of those vibrations first hand.
Marc
[a href=\"index.php?act=findpost&pid=90855\"][{POST_SNAPBACK}][/a]

I don't think IS is tries to isolate the camera from vibrations.  Rather IS attempts to compensate for vibration by reacting to motion and "reaiming" the camera.
Logged

marcmccalmont

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1780
Image stabilization
« Reply #13 on: December 16, 2006, 03:29:16 pm »

Quote
I don't think IS is tries to isolate the camera from vibrations.  Rather IS attempts to compensate for vibration by reacting to motion and "reaiming" the camera.
[a href=\"index.php?act=findpost&pid=90862\"][{POST_SNAPBACK}][/a]

I guess my point was not that Vibration Isolation and Image Stabilization are the same but they both serve the purpose to get a sharper image and as in the audio industry the imaging industry seems not pay it much attention. I'm surprised that no one has commercialized "Vibration Damping " heads.  By the way none of my audio equipment (all resting on EAR vibration isolators tuned to below 20hz) even moved during the earthquake while everything else ended up on the floor. When I first bought my 5D I purchased with it a 24-70 2.8 and a 70-200 2.8 IS. within a month I replaced the 24-70 with the 24-105 IS. The IS works that well, especially in the cockpit where there is both turbulence and engine vibrations to deal with.
Marc
Logged
Marc McCalmont

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
Image stabilization
« Reply #14 on: December 16, 2006, 03:59:00 pm »

Quote
I guess my point was not that Vibration Isolation and Image Stabilization are the same but they both serve the purpose to get a sharper image and as in the audio industry the imaging industry seems not pay it much attention. I'm surprised that no one has commercialized "Vibration Damping " heads. [a href=\"index.php?act=findpost&pid=90871\"][{POST_SNAPBACK}][/a]

Perhaps the reason is, with photography, image stabilisation solves only half the problem of getting a sharp image. It can't do anything for subject movement. In fact, it can be a bit of a trap causing one to use slower shutter speeds than one would otherwise use, sometimes resulting in blurred images due to a lack of awareness of subject movement when taking a shot or a miscalculation of the significance of subject movement in a particular shot.
Logged

marcmccalmont

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1780
Image stabilization
« Reply #15 on: December 16, 2006, 06:27:09 pm »

Quote
Perhaps the reason is, with photography, image stabilisation solves only half the problem of getting a sharp image. It can't do anything for subject movement. In fact, it can be a bit of a trap causing one to use slower shutter speeds than one would otherwise use, sometimes resulting in blurred images due to a lack of awareness of subject movement when taking a shot or a miscalculation of the significance of subject movement in a particular shot.
[a href=\"index.php?act=findpost&pid=90876\"][{POST_SNAPBACK}][/a]

Very good point. One of the hardest pictures for me and I have not been succesful yet is to get a sharp picture of the whales breaching during the winter months. I use a 5D adapted to a Nikkor 1000mm reflex, high shutter speed, high ISO and a single lever Slik pan head. With the wind and manual focus it is near impossible, I hope this year is the charm. I have to prefocus, pan and pray for a break in the wind.
Marc
Logged
Marc McCalmont

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
Image stabilization
« Reply #16 on: December 16, 2006, 09:57:10 pm »

Quote
Very good point. One of the hardest pictures for me and I have not been succesful yet is to get a sharp picture of the whales breaching during the winter months. I use a 5D adapted to a Nikkor 1000mm reflex, high shutter speed, high ISO and a single lever Slik pan head. With the wind and manual focus it is near impossible, I hope this year is the charm. I have to prefocus, pan and pray for a break in the wind.
Marc
[a href=\"index.php?act=findpost&pid=90893\"][{POST_SNAPBACK}][/a]

With the 5D, I would have no hesitation in using ISO 1600 in those circumstances. A sunny day on the open sea can permit some very fast shutter speeds.

Ever tried taking photos whilst riding an elephant?  
Logged

Ben Rubinstein

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1822
Image stabilization
« Reply #17 on: December 17, 2006, 08:11:22 am »

My 70-200 f4L IS is less sharp than the non IS version I had and so is the 70-200 f2.8L IS that I used to use, a drop less sharper and a drop less contrasty, I haven't tested it with the IS switched off though to see if that makes a difference. That said IS more than makes up for it!
Logged

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
Image stabilization
« Reply #18 on: December 17, 2006, 09:49:41 am »

Quote
My 70-200 f4L IS is less sharp than the non IS version I had and so is the 70-200 f2.8L IS that I used to use, a drop less sharper and a drop less contrasty, I haven't tested it with the IS switched off though to see if that makes a difference. That said IS more than makes up for it!
[a href=\"index.php?act=findpost&pid=90953\"][{POST_SNAPBACK}][/a]

Well, that's a disappointment, Pom. The non-IS version of the 70-200 f4 has been out for a good many years. I would have expected that the new lens would be at least as good, considering technological improvements.

Maybe we'll see a few detailed comparisons on the net in coming months.
Logged

gkramer

  • Newbie
  • *
  • Offline Offline
  • Posts: 27
Image stabilization
« Reply #19 on: December 17, 2006, 10:02:55 am »

Quote
One of the hardest pictures for me and I have not been succesful yet is to get a sharp picture of the whales breaching during the winter months. I use a 5D adapted to a Nikkor 1000mm reflex, high shutter speed, high ISO and a single lever Slik pan head. With the wind and manual focus it is near impossible, I hope this year is the charm. I have to prefocus, pan and pray for a break in the wind.

Sounds like your're doing just about everything wrong--no image stabilization, no autofocus, using a tripod instead of a shoulder stock, and using a full-frame camera--which for most telephoto wildlife work is a disadvantage, since one is usually lens-limited--can't get close enough to the subject to fill the frame, so must crop the image. A 12.8MP image from the Canon D5, cropped to APC size, will yield a final 4.6MP image--compared to a full 10MP from several current DSLR's using Sony's latest 10MP APC sensor, or 12.2MP for the Nikon D2X.

Do you also use an incident-light meter?
« Last Edit: December 17, 2006, 11:52:08 am by gkramer »
Logged
Pages: [1] 2 3 ... 5   Go Up