What Is Image Stabilisation?

What Is Image Stabilisation?

Image stabilization - Wikipedia - Absolutely "nice to have" feature. Unfortunately it makes lenses more expensive and is another thing that might break some day.

Image stabilisation is the European version of image stabilization! :)

If you try to hand hold a camera and lens on a subject without IS (Image Stabilization) you may notice you can't hold the camera perfectly still while focusing so the image in the lens will appear to jump around. This is more noticeable on telephoto lenses zoomed out, the farther out you zoom the worse it will appear, the slightest movement can make the image jump around. If you are shooting at 600mm it is extremely noticeable. This becomes a problem at slow shutter speeds. IS will help lock the image a bit more and keep the image stable. (This only works to point but is nice to have because having it is better than not) This can allow you to use a slower shutter speed, useful in low light where you would like a slower shutter speed because there is less light. Camera shake is less noticeable at wide angles and you do not see it offered on wider angle lenses. It does add to the cost of a lens.

 

Canon calls this IS (Image stabilization), Sigma brand lenses call it OS (Optical stabilization), Nikon calls this VR (Vibration Reduction), Tamron brand lenses call this VC (Vibration Compensation), it may be called Sensor-Shift Stabilization for some brands of cameras that have the stabilization built into the body of the camera.

One of the nice things about having IS built into the camera body is that any lens attached will have IS. You won't have to buy the more expensive lenses with IS if IS is built into the camera. Many Canon and Nikon shooters wish Canon and Nikon would start incorporating this feature into the camera body. Hope this helps.

Mark, thank you for a stabile answer.

Here's a good article (pros and cons) on lens vs in-body stabilization: https://photographylife.com/lens-stabilization-vs-in-camera-stabilization

Here's a good article (pros and cons) on lens vs in-body stabilization: https://photographylife.com/lens-stabilization-vs-in-camera-stabilization

 

This article fails to mention that in-body and in-lens stabilization can be used at the same time, as I do with my Sony a9.

This article fails to mention that in-body and in-lens stabilization can be used at the same time, as I do with my Sony a9.

Curious on that. How is that working? Do you find an advantage of having two IS systems working together over just one? Do you get more stabilization using your big 600mm with IS and the Sony? Dual stabilization sounds impressive if it works. I have read on some cameras you can't use both, but you say you can with the Sony. Sure would be nice hand holding my 150-600mm Sigma C and being able to use a slower shutter speeds for some shots.

Edited November 8, 2017 by Mark Keefer

Does the Sony IBIS work with a Metbones adapter and a Canon IS lens?

Does the Sony IBIS work with a Metbones adapter and a Canon IS lens?

 

Mark, I'm sorry to say, but I've not really tested that. I have a couple of lenses with 4-stops of in-lens stabilization and the body claims 4-stops in 5-axis. Maybe next weekend, I can do some with and without testing.

 

I doubt that the effectiveness is additive, but,if it is, I should be able to hand hold 1000mm at 1/50-sec. I'll try that, with and without, shooting at a sign. That should reveal if there's a difference. I can say that it feels very stable, but it did with only the in-lens IS.

I assume that IBIS will add stabilization axis to IS lenses; i.e. compensate the camera rotation or such. - I wouldn't assume 5+3 stops but hope for an extra stop compared to lens IS only. - Pentax didn't even try stabilizing anything beyond 300mm in camera, so I guess dcstep's birding glass will benefit a lot from IS. - Let's wait for results! & TIA for sharing them.

It's not likely Nikon or Canon would add in-sensor stabilization to their DSLR cameras because they already have many lenses with VR/IS and the viewfinder stabilization is an important feature of in-lens VR/IS on DSLRs; a moving sensor would lead to mismatch between what is shown in the viewfinder and the actual image so it would be difficult to compose precisely and hold the AF point on the subject. Additionally there would be patent licensing fees.

 

However, in a mirrorless camera, in - sensor stabilization makes sense as the viewfinder would accurately reflect the stabilized image and then it is mainly a question of how much extra cost it would be to implement and license the technology and whether customers prefer a more expensive camera or lenses.

 

Personally I tend to find that a tripod gives the best image quality on static subjects and fast shutter speeds give the best quality on moving subjects - in both cases stabilization plays only a minor role at best.

From everything I've seen when an OSS (sony) lens is used with an a (IBIS) body w/ the 5 axis sensors on the current generation of Sony cameras, the lens is handed the 'Pitch' and 'yaw' axes, and the sensor retains control of the X, Y, and Roll axes. When used with a 3rd party lens (say an EF) or lens w/o any OSS, the sensor handles all 5 axes. One would assume that doing so is advantageous to the overall effectiveness, as controlling pitch and yaw from further down the optical path could hypothetically yield improved results (vs. those axes controlled solely at the end of the optical path). This of course would require the lens and body talking to each other... Interestingly, Sony doesn't seem to claim any improvement in effectiveness... everything is conveniently 5 stops...

 

 

...Using a long EF IS lens though, I'd imagine that you def want to turn the IS off...since it's controlled completely in the lens, and could easily cause diminished results since the sensor and the lens are separately detecting movement and likely either cancelling the stabilization, or making things worse. Of course maybe the metabones is smart enough not to turn it on when the IBIS is on? IDK - but curious...

It's not likely Nikon or Canon would add in-sensor stabilization to their DSLR cameras because they already have many lenses with VR/IS and the viewfinder stabilization is an important feature of in-lens VR/IS on DSLRs; a moving sensor would lead to mismatch between what is shown in the viewfinder and the actual image so it would be difficult to compose precisely and hold the AF point on the subject. Additionally there would be patent licensing fees.

.

 

Maybe it's because I routinely shoot super-telephoto focal lengths, but I easily see the impact of IS in the optical viewfinder. Shooting mirrorless, when I turned off IS, I saw it in the viewfinder also. They seemed roughly equivalent to me, with no problem holding AF point, IS on or off.

 

Don't know about royalties, but it seems like in-lens IS would be subject to royalties also.

they already have many lenses with VR/IS and the viewfinder stabilization is an important feature of in-lens VR/IS on DSLRs; a moving sensor would lead to mismatch between what is shown in the viewfinder and the actual image so it would be difficult to compose precisely and hold the AF point on the subject

Dunno; I see the entire thing differently. - First of all: Not all SLR VFs cover 100% FOV, so moving sensors might not matter in their case? 2nd issue: While there are some IS/VR lenses, we are still stuck with a shopping dilemma: 35mm f2 IS vs f1.4, similar about 28, 24mm and even the 24-70 zooms: f4 IS or f2.8? - I'm less familiar with the Nikon lineup but guessing Nikon don't make money by selling stabilized Tamron primes?

I'm no birder, so the AF point aiming issue is less relevant to me when I am not shooting anything beyond 200mm. And in nasty low light we tend to pick shorter lenses anyhow?

IBIS would add a lot of value to the bread & butter primes like 50 & 85mm f1.8 in the dark. I also believe that lens based VR/IS demand extra elements to be added to the construction which come with an optical price to be paid too? I have no clear idea what "micro contrast" bokeh and similar traits are based upon but in the old days lens makers tried to avoid elements for some reason, so IBIS might be worth framing a wee little bit generously and cropping in post.

When I look at Sony's A 6*00 series I would be tempted to spend a few 100 extra $$s on the 6500 for IBIS' sake. Canon & Nikon could probably attract a few extra customers by offering it too. - I stuck to Pentax for quite a while because I considered having SR worth it.

They seemed roughly equivalent to me, with no problem holding AF point, IS on or off.

 

On a mirrorless camera you see the effect of both optical and sensor-shift stabilization in the electronic viewfinder but on a DSLR you only see the effects of optical stabilization whereas the effect of sensor shift (if present) is not reflected in the viewfinder image.

it seems like in-lens IS would be subject to royalties also.

 

Of course but Nikon and Canon have been developing and patenting the optical stabilization technology since the 1990s and if they have to pay royalties it is likely to each other whereas if they were to use sensor-shift stabilization they'd also have to pay Sony (Konica-Minolta?) and Olympus for their part in the development of IBIS. This would likely increase the camera body price along with its complexity. By implementing only one technology for stabilization I believe they save costs compared to a dual implementation. Without doubt when Nikon and Canon implement large-sensor mirrorless they will also consider different stabilization technologies but my guess is they will stick with what they know, i.e. put it in the lens.Over time of course patents expire and it's then possible that the technology becomes more widely used.

 

Not all SLR VFs cover 100% FOV,

 

True, but 100% viewfinders are common in mid to upper tier DSLRs.

 

While there are some IS/VR lenses, we are still stuck with a shopping dilemma: 35mm f2 IS vs f1.4, similar about 28, 24mm and even the 24-70 zooms: f4 IS or f2.8?

 

While I use stabilization when it is helpful and available, its presence or absence in a lens of a short focal length doesn't pay a role in my decision making for which lens to buy or use. This is because a shutter speed that is fast enough to freeze subject movement in a typical situation where I would hand-hold the camera is also fast enough to avoid camera shake. I don't like the effect of a blurred subject in a sharp surround; I prefer the opposite (sharp subject in a blurred surround) although I realize controlled movement blur is utilized by some photographers effectively, particularly in dance and bird in flight photography, it's just not my thing. I do think stabilization is useful for long lenses because there it is likely that a shutter speed that is fast enough to stop most subject movement of a stationary or slowly moving subject may not be sufficient to avoid camera shake, and just pointing a hand held long lens is difficult without stabilization, hence there are tangible benefits. But: in long lenses I believe the in-lens stabilization is more effective than in-camera stabilization because of range of movement issues and even in the presence of in-camera stabilization you'd probably want to buy long lenses with optical stabilization, thus in the area where stabilization is most useful the implementation of in-camera stabilization would not result in cost savings to the user if they want the most effective implementation of the technology.

 

I'm less familiar with the Nikon lineup but guessing Nikon don't make money by selling stabilized Tamron primes?

 

I've read that Tamron and Nikon have some co-operation (shared patents implying some lens development together) and it may be that Tamron pay Nikon license fees for technology and the mount (whereas Sigma does not, which explains why Sigma were sued by Nikon for use of optical stabilization in the 70-200/2.8 and Nikon won the case, leading to some compensation). So to some extent Nikon do get money from third parties for use of technology Nikon developed in their lenses. Interestingly Nikon's 19 PC seems to be covered by a patent which was filed by Nikon and Konica-Minolta many years ago. So there is not just competition but also collaboration between companies in various areas of development.

 

IBIS would add a lot of value to the bread & butter primes like 50 & 85mm f1.8 in the dark.

 

Would it? I often photograph events such as post-phd-defence dinner parties and they are typically in dimly lit restaurants. With f/1.4 lenses I get 1/200s at ISO 6400 and the image quality is good. When photographing people giving speeches, even 1/100s would mean the facial expressions would likely show some blur and hand movement would also be exhibited as blurry trails. Thus I'm not convinced that IBIS would make much of a difference to the outcome. It might allow some shots of groups at stopped down apertures (say f/2.8) at 1/50s where individuals are smaller in the frame and blur might not be as evident as in telephoto close-ups but then I can use the 24-70/2.8 VR and Canon users could use a 24, 28, or 35mm IS prime for such shots. True, the Canon 24mm/28mm IS and Nikon 24-70 VR are only f/2.8 lenses and someone might ask for f/2 or f/1.4 but those extra stops in faster primes allow an increase in shutter speed that alleviates the need for stabilization in such a lens. I just don't see a pressing need in practical use. I appreciate that users of IBIS enabled cameras find uses for stabilization in shots made with short focal length lenses - you make what you can with the tools that are available to you. For me the optical viewfinder is a tool which I cannot do without and IBIS is not.

 

I think both silent shooting and IBIS are features that are well suited for implementation in mirrorless cameras and also I believe the effectiveness of stabilization should be improved if there is no mechanical shutter or mirror in play, as these cause additional vibration. If the camera only has to compensate for human hand shake the compensatory algorithm should be easier to make. Thus these features while they can be implemented on DSLRs they are not especially compatible with the central differentiating feature of the DSLR which is the optical viewfinder. If live view is used (either using the back LCD or with an accessory EVF which are in fact available for DSLRs), then IBIS and silent shutter could be used similarly to how they are used on mirrorless cameras but it would just render the DSLR into a very large camera with mirrorless-like functionality. I think Nikon and Canon see the advantage of the DSLR the same way I do: comprehensive lineup of lenses, optical viewfinder, and are not necessarily going to add a lot of mirrorless camera style functionality to benefit what is essentially a secondary mode of operation (live view) while not helping when the OVF is being used. There is some development of live view specific features in DSLRs but they don't get the same priority as they do in mirrorless camera development.

 

I also believe that lens based VR/IS demand extra elements to be added to the construction which come with an optical price to be paid too?

 

This may be true, depending on the lens design, but at least in longer lenses where VR and IS has been added, at the same time there has been in improvement in image quality so probably those extra elements can play a role in improving image quality and not just implement stabilization. However, I do believe in a short focal length lens jam-packed with elements it is difficult to implement stabilization. Nikon added VR to the 24-70/2.8 and the lens became somewhat longer and more expensive but they were able to improve corner sharpness, reduce color fringing, improve bokeh, improve color and contrast, reduce flare, improve AF speed in the upgrade as well, while there was a slight reduction in center sharpness in some conditions (i.e. at 70mm, at close distances). I think it's a very successful design in terms of the beauty of the images that result. I guess the main tradeoff here is not so much the image quality but the fact that VR necessitated a lens design that is noticeably longer in size and going through so many trial lens designs must have added to the final cost of the product.

Edited November 14, 2017 by ilkka_nissila

I would think there might be a problem with a sensor stabilizer and a lens stabilizer fighting each other.

How did we ever get by before IS and AF.

They seemed roughly equivalent to me, with no problem holding AF point, IS on or off.

 

On a mirrorless camera you see the effect of both optical and sensor-shift stabilization in the electronic viewfinder but on a DSLR you only see the effects of optical stabilization whereas the effect of sensor shift (if present) is not reflected in the viewfinder image.

....

 

I only know the Sony a9, but IBIS-only stabilization is visible in the EVF.

 

I was going to do some demonstration shots of no IS, IBIS-only and in-lens IS only, but didn't realize until I started testing that my FE 100-400mm would not allow IBIS-only operation. I will actually do a comparison, but I'll need to use a Canon lens on the Sony body. I'm travelling now, but I'll get to it later this week.

I would think there might be a problem with a sensor stabilizer and a lens stabilizer fighting each other.

 

Not a problem with Sony/Canon.

In my day, image stabilization involved the treatment of silver prints with a platinum solution. The platinum replaces the silver, and ia virtually immune to chemical attack, in addition to imparting a rich, warm tone to the print.

 

Now that we are beyond all that, image stabilization in cameras and lenses uses accelerometers to detect camera motion, and moves the sensor (IBIS) or elements in the lens to counteract that motion. There is no reason either or both cannot be used in a DSLR, since the sensor is usually not used to detect motion. There is an exception where digital stabilization is used, in cell phones and simple video cameras.

 

In-body image stabilization ((BIS) works best on shorter lenses, but can be effective on any lens compared to hand-holding alone. On the average, camera shake is about 2 deg/sec. The 1/F rule of thumb results in uncertainty in the film plane roughly equal to the full for depth of field - just below detection by the unaided eye in an 8x10" print held 10" from the eye. The largest part of camera shake is tilt and yaw. Lesser factors are rotation, vertical and horizontal translation (5 axes). The longer the lens, or the closer you go, the greater the effect of camera shake. The greater the effect at the film plane, the more displacement of the sensor is required to counteract it. It takes less physical displacement for optical image stabilization, which is shy longer lenses tend to incorporate IS in the lens. Some shorter lenses use it too, but it adds relatively more bulk and cost.

 

Sony and Olympus are the only cameras that use IBIS, to my knowledge. I have a Sony A7Rii and A9, in which IBIS and lens IS are highly effective. Sony advertises 5.5 stops of stabilization, compared to the 1/F rule (200 mm = 1/200 sec minimum). I have one test with a 200 mm lens which lends credence to this claim.

 

Sony A9 + Sony 70-200/4 at 1/5 second

[ATTACH=full]1219384[/ATTACH]

 

[ATTACH=full]1219383[/ATTACH]

Edited November 14, 2017 by Ed_Ingold

While I appreciate stabilization in longer lenses >80mm, I am rather in agreement with Ilkka on this. I have a stabilized 16-35mm, and to be honest I think it does not add much extra to the lens. One can easily handhold a 16mm lens at 1/60th and get sharp results. The IS allows me to hand hold at 1/15th, but then human and any moving elements are blurred, so I probably will not like the result, so it is no advantage. The same applies even with a 35mm lens: 1/60th will be nice and sharp with IS, but moving objects will not be. Stabilization is very useful with longer lenses when high shutter speeds such as 1/1000 sec are needed to ensure freedom from camera shake, but for normal and wides, I am not really convinced of the necessity.

The 1/F rule is a bare minimum. To get pictures sharp at the (nearly*) pixel level, you have to shot 3 or 4 times that fast.That's where image stabilization proves useful, especially for landscapes. A 16 mm lens at 1/60 would fall in the "safe" category. It is also useful in low light, even with short lenses. Even though people may move enough to blur in a longer exposure, it's not particularly difficult to catch them at relatively stationary moments. You just have to take more shots and hope for the best. Camera shake, on the other hand, is always there, and doesn't change much even if you concentrate.

 

* I say "nearly" because IS seems to impart slight disturbances at best, and if used on a tripod, very noticeable disturbances. With both IBIS and optical stabilization at my disposal, and the Sony does not produce noticeable artifacts described in the previous sentence, I rarely use my Sony cameras on a tripod, and then mainly for consistency or hand-free operation, rather than ultimate sharpness.

Most fairly recent IS Canon lenses can sense if the lens is tripod mounted and switches off the IS, or at least Canon do not feel it necessary to say anything about it, but the earlier Canon IS systems were suggested to be turned off. It is advisable to read the instructions for each lens in this regard.

 

I am not a slave to "ultimate sharpness" which I generally think is a never ending and rather pointless pursuit, but in general I do like an suitably sharp image when required.