VR for wildlife: is it worth the risk?

[ecopix]

I'm interested in the opinions and findings of other wildlife photographers on the use of image stabilization on long lenses in low light (shutter speeds around 1/200th).

I find that my own experience, after much experimentation, is still a bit confusing. It's complicated, but I'll try to describe my uncertainty. These comments are more about the Nikon VR on long lenses than others. The Canon and Olympus (in body) versions seem much milder and less intrusive, and so don't really generate this uncertainty.

There are two characteristics of VR that create the question. The first is that the effect is statistical. VR doesn't mean you will get a sharp picture, it means that out of 20 pictures, you have greater probability of getting a sharp one if you use VR. This strikes at the heart of the question: is it worth the risk?

The second consideration is that no one tells you that it is a two-edged sword with long lenses. It can snag you a surprisingly sharp image in low light, and it can also completely destroy the sharpness of all the others, that might otherwise have been just a bit soft. Deciding to switch VR off or on seems like deciding to bet on a horse or a risky investment. The horse can only lose; the risky investment can take you down with it!

Now the heart of the matter: wildlife photography isn't about making 20 pictures and picking out the sharp one. That's what sports and fashion photographers can do. If you miss out, you can ask the model to do it again, and there's always the next game!

Wildlife photography is about waiting poised after a week's observation, finger twitching, ignoring the sharp rocks and the mozzies, for that one magic moment of behaviour, and making that one rare, and often unexpected, opportunity count.

Sure, we make hundreds of nice portrait shots of animals, some sharp, some not, but we know that it is only the one special moment that will make the shot that we want, that the editor wants, and that the viewing public will go ooh and aah over. It's a kind of harnessed serendipity.

For that one special shot, do I switch the VR on and hope my number comes up, at the risk of losing all, or do I switch it off and just hope my shot is "sharp enough"? How far do I push Serendipity?

Thanks for your thoughts.

[arthurking83]

Hoping is also a two edged sword. I've shot at faster than the 1/focal length recommendation, and still got plenty of unsharp shots, as well as having shot at shutter speeds seemingly ridiculous considering the focal length and they've come out beyond respectable in terms of sharpness(eg. 1/60 at 200mm).

I'm not sure exactly why you can't spray and pray just as the sports protographers do. I'm not a dedicated wildlife photog, I have followed a few around now(I think three maybe four occasions now, and using the continuous high mode of the higher end (Nikon) cameras seems to produce the best results(of at least one image worth displaying).

I'm inclined to think that even just past the threshold of 1/focal length shutter speeds... say 1/2-1 stop faster?? Leaving VR(or IS or whatever) on may be of benefit.

There is a line of thought that seems to think that VR may be detrimental to image sharpness, once you reach speeds past 1/1000s. This is because the frequency of the movement of the VR lens(or lens group) in the lens is believed to operate at about that speed. That is, the lens floats at speeds up to 1000Hz, so a shutter speed of about 1/1000s may counter the VR lens elements attempt at trying to produce a clear image. :confused:

I have two lenses with VR(18-105VR and 105VR) so no 'proper' telephoto lenses to check out some of the information I've read so far.

I'm a big fan of VR myself.. even at short focal lengths, for when the need arises for it.

FWIW, Canon have in lens stabilisation. Sony, Pentax and Olympus all have in camera stabilisation systems.

Also as 105mm is my (current) longest focal length, and even with VR on, I can still get a random number of blur affected images in a continuously shot series even at 1/focal length shutter speeds. Of course I will usually get at least one or two very sharp shots as well.

Also note that in the Nikon line of long lenses, some have several modes of VR.. something like panning or tripod/monopod mounted modes .. and stuff like that.

On a side note too tho.. one of the real benefits of VR(IS) in lens is the smooth viewfinder it gives when in operation. One thing i really hate a lot about my manual 500mm lens is that when I look through it, and apart from the darkness, the small vibrations and shakes I see through the vf can be bothersome after a while of use.

as for the question you posed(of how far to push...) really! only you will know the answer to that. The best way to figure out the answer(for yourself) is to test a lot. That is, set yourself the task of shooting at various shutter speeds and at various focused distances(common to wildlife shooting) eg. small bird at 5 meters, large wild grazing animal(such as water buffalo, at 20meters) and see how each variable affects the quality of the images you've captured. Of course this requires a static subject and a lot of patience and dedication. Once you've figured out a few basic sets of data (ie. that at 1/200s and 200mm and focused distance of 5 meters) that you get consistently sharp images over a series of 6 continuous shots in one second not using VR.. you have to commit to memory that VR is something you don't need .. etc, etc.

[farmer_rob]

Thom Hogan give (IMO) a pretty good wrap-up of when and when not to use VR here (http://www.bythom.com/nikon-vr.htm), with some pretty fair reasoning as to why.

In my experience, VR does not really give me anything except a warm and fuzzy feeling (and some of that is in the photo:p). I tried some direct experiments (single shot only) and found that nothing beats "tripod with mirror lockup" for sharpness. VR was better at slow shutter speeds (under 1/60) than no VR, but for me still not sharp enough if any substantial cropping was going to occur.

In my limited wildlife shots, I have found that VR does not give me enough enhanced sharpness at 300mm on a DX body if I am pushing the limits. (Nothing beats getting close enough not to crop and enough light for faster than 1/250 shutter speeds:o.) However, I am not a dedicated wildlife photographer, and don't have the big fancy lens to compare.

However, I am not bagging VR - I just think it is not a complete fix.

[ricktas]

Simple, VR/IS/OS lenses have a switch to turn the feature on and off. Use it to ensure you get the results you want. As the above posts say, shooting wildlife often involved firing off a lot of shots in quick succession, to get 'the shot'. With VR off or on, your results will come down to more skill, settings etc, than whether VR is on or off. Remember VR is only new, and we have some damn fine examples of wildlife photographer, in fact some truly stunning ones, from prior to the invention of VR, so it isn't the be all of wildlife photography.

[macmich]

spend a weekend of shooting in the scrub for two hours with it on one day and the next day do it with it off and compare your results
i no you wont have the same roo or bird in the the spot or light or whatever
but it will soon tell
cheers macca
bugger it i will do it tuesday and wednesday not that i am any wheere as good as you fellas

[kiwi]

I'm still not seeing the risk, or downside to vr.

Either on or off you still fire bursts...it's just better to do that on most things

[Allann]

I can only speak from a canon point of view and my own findings. I have shot many genre now but the one constant is my bird photography. I don't share many anymore as it has become a more personal crusade of relaxation than anything else. I mainly use my 600/4 non IS lens and think it's all down to proper technique. This isn't something you can be told to do, it comes with practice and knowing your camera/lens combination. In fact I don't own any IS lenses anymore, preferring to rely on my skills and knowledge to get the shot.
I am not saying VR (IS) is a bad thing and I have seen many photographers use it to good effect, mainly when hand holding at very slow shutter speeds. When I photograph birds I try to keep my shutter speeds as high as possible.
I'm sure you hit a raw nerve with some saying the moment isn't impotent in sports/fashion, I can tell you as a fact they are, and in some case, even more so. Wildlife behaviour is seldom a once of thing and if studied long enough you will see the same behavior again, in sports photography it is all about the moment, take for example the mini rolling end over end at the v8s at QR, or the fashion model loosing her dress, these shot would make their photographers a nice sum of money, the front page (instead of the second last). I'm not saying that wildlife doesn't have those moments though.

[kiwi]

The point of vr for wildlife is that for still subjects, say a lion sitting in the sun that for the same conditions you can half your shutter speed therefore halviing your iso or increasing your apperture, both of which might help clarity, if the objects moving a lt as Allan says your going to want a higher shutter speed and vr doesn't help as much, but it will still help.....as will dozens of other things like technique, tripod, etc etc

Just another crutch

[Steve Axford]

Interesting read. I think it may come down to personality and opportunity, rather than anything purely technical. I love IS and I wouldn't be without it, but I think of photography as a probability game. Anything that improves my chances of getting that great photo is good and I find that IS greatly improves those chances for wildlife. I use a 300/f2.8 usually with a 2x extender, which I can hand hold. If I had a heavier lens, the IS would be of less value as I would need a tripod anyway.

[Shelley]

I am not really against VR/IS, but I do just fine without it. I did use it once and maybe my lack of skill at the time was the downside. I am saving for the 70-200 IS - so maybe then I can give a better opinion. But, that lens is not for my wildlife shots specifically.

But, currently I use my monopod for wildlife shots and love the sharp images that I can achieve - that is what one goes for in birding. Even with high shutterspeed, the tripod/monopod achieves that little bit more in my image compared to handheld, not really sure if IS would do this, as it not necessarily done in low light. Its not a hassle for me carrying tripod/monopod around when birding. For me its only a small part of my photography technique as there are a lot of other factors to consider when getting wildlife shots, such as getting close to the subject, positioning and patience. I do not feel the need to rush out and get a lens with IS to improve my bird shots. I also agree with what Allann posted.

[arthurking83]

..... I think it may come down to personality and opportunity, rather than anything purely technical. ....

Definitely agree here.

I definitely want it in all my long lenses now.. maybe not so much at say 105mm, but at 200mm and beyond.. for sure!!

Although to a small degree, I think there is some technical aspect too.

When I was younger(say 20-25years ago) I used to have a rock steady hand.. so much so that I found it very easy to solder and desolder the most microscopic parts off an electronics board. Now, I seem to have a bad case of the shakes. Some time back, I had to repair my home DVD player, and replace a standard capacitor, with very easy to reach solder joints, but I noted as I did this, my hand shaking in small sharp fits and I didn't have the smoothness I once remember that I used too. Stil got the job done, and saved myself the pain of paying for another new electronic device that I end up not happy with.
The point being that my hand is no longer steady, or as steady as I'd like it to be. VR helps with that.
If you are susceptible to the shakes, even on a monopod optical stabilisation is handy(some lenses have a monopod VR mode remember).

I'm with Shelley. I currently have the Tamron 70-200/2.8 which doesn't have VR(or VC in Tamron terms) and will one day upgrade it to the Nikon 70-200VR lens. Tammy is great in terms of outright IQ, but in many situations, the VR would give me a lot more keepers when handheld, so I tend to use it on a tripod the majority of the time.

Also note what Farmer Rob said too.. re tripod and mirror lockup. I don't think I've ever seen any images with my VR lenses (18-105VR at the long end) that could be comparable in sharpness to that method(but of course in wildlife photography this is not an option).
The point is that VR is not a substitute for tripod and MLU.. it's an advanced feature(in lenses in Nikon and Canon circles) that simply allows acceptable sharpness in images at lower than usual shutter speeds.
It should really be seen in that maner, and not as a tripod replacement feature(as per KR's thoughts!! :rolleyes: ... who's KR? ... don't ask :p)

[ecopix]

Wow! What an array of generous expertise. I'm amazed at the insightful comments my obscure little question generated and am very thankful for them. Thank you all.

What seems to emerge is the same probability thing, rather than a cut and dried answer. Maybe VR’s quantum? No one seems to have found that it makes all their pictures sharper, up to that theoretical upper limit that Arthur speaks of.

But then, few bemoan the damaging effects of VR, either (apart from Rob’s warm and fuzzy one-liner).

Often I see a negative effect of VR, in which shots that should just be a little soft from micro-movement, are, with VR, decidedly blurred and even double-imaged, along with the miraculously sharp ones. That’s the risk I was asking about. It seems exaggerated with long lenses. It's as if it’s sharpening some that would have been blurred and blurring some that would have been sharp! Maybe Rick nails it – the most important feature is the off switch.

Regarding "spray and pray" (it's good that photogs can laugh at themselves like that), my D300's only do 6fps, and if you set your shutter to one second and note the duration, it’s obvious that one second is an eternity in the quick movements of animals. An awful lot can happen in a second, and most "magic moments" will be over before the second or third frame gets a chance. Only one frame captures the peak. I know I'm not telling anyone anything here!

I stand corrected implying that the moment isn't important in sports and fashion. My apology. Of course Allenn is right and a skilled, perceptive and ready photographer can make the most of the decisive moment everywhere.

My point was just that with wildlife, it's what the editors seem to expect all the time, and what we aim for. A sports game or fashion parade is news in itself and every dress is different, but a species portrayal needs a special moment to be different. If I come back from the bush with nice but ordinary pictures of species (difficult enough to make) and a few unexpected ooh and aah shots, the ordinary ones are passed over and ignored.

So I can't afford to lose those carefully fostered no-second-chance opportunities to a device that was supposed to help but hindered. I think the answer emerging is that for wildlife, it's better to stay with the techniques we've developed and just try VR when it seems appropriate, such as in Kiwi's example, and when there’s time to experiment with it on and off. Do I read the thread correctly there?

With many thanks,
Wayne Lawler

[Allann]

Would love to see some of these wildlife shots your talking about!

[farmer_rob]

I can see where you're coming from Wayne - and I think the best thing for you to try is experimentation. Not trying for the "wow" shot, or the hard to shoot animal - but instead an easy to find bird or pet dog. Try a few shots with and without VR, using your best technique and see what comes back. All your shots have a probability of error, as I am sure you know - does VR increase or decrease that probability? Only experimentation will really tell.

Also, IIRC, VR at higher shutter speeds - say 1/250 or more - starts to interact negatively with the shutter (simple physics/maths related to the operational speed of the VR mechanism). If you believe the marketing blurb, VR is meant to improve images at "below hand-holding" shutter speeds - although they don't quite state it that way. However, I don't think this takes into account telephoto 1/focal length speeds, which are really suffering from slightly different vibration/motion issues. I reckon that if you need 1/500 to get your image without VR, VR is unlikely to help; if you are shooting at 1/60 or slower, VR will help. You can test and prove (or disprove:o) this yourself quite easily.

[ecopix]

Thanks Rob. Sorry for the delayed response, Allann – Christmas ‘an all… You’d be disappointed in my pics if you’re expecting those amazing super-spectacular wildlife shots one sees (occasionally) in publications. I was the one who asked the question, but I mislead a bit in my lavish description of the one-off shot. I wanted to know what I should do when the opportunity presents itself!

Seriously, those great shots are normally made at fast speeds or with high speed flash, so they’re not VR territory anyway. What I had in mind are the more “ordinary” wildlife shots that are still fleeting one-or-two-frame-only shots, and even if they’re repeated often enough in the animal’s life, they won’t happen again any time soon while I can capture them.

The sort of shots where the editors say, “Oh, yeah”, not, “Oh, wow!”, but at least run them sometimes. Even these “ordinary” wildlife shots take a lot of time, effort and luck, as we all know! We try for them because we know that editors, customers, photo libraries, competition judges or whoever uses our pictures, consider them normal shots.

For examples (the flash file linked below), they don’t want a shot of a peregrine standing still - they want one stooping at the camera, because that’s what peregrines do best. Likewise, they don’t want a shot of a desert dragon just lying low, because everyone knows they stand tall to beat the heat; or a rufous whistler not whistling, or a rock wallaby not leaping, or an ant not dragging something ten times its weight, a shining starling not showing it’s crazy bulging red eye in display, or a dominant male wallaby not standing tall and flexing its muscles at a pretender.

Capturing this behaviour is the name of the game in wildlife photography (editorial wildlife photography, anyway), so we need to know if our tools and techniques will capture that one frame when it happens in front of our lens. They can be simple and everyday - a shot of a joey biting its mother’s tail, or a galah doing contortions trying to work out the viewer, bring wildlife into people’s hearts.

I made one of these more humble moments yesterday, photographing a pardelote investigating a hole in a piece of old plywood. Interesting behaviour, I thought, but then another pardelote came and they started fighting over the hole (pardelotes are turned on by holes). Not a world-beater, but the pick of the shots in that series, and the one most likely to be used. It was only a one-frame chance.

Many of these more humble moments can be made at middling shutter speeds at which VR can be useful with a long lens. That’s why I asked the question about using VR at moderate speeds with teles, that everyone answered so generously with their experience. Rob gives a good final summation.

Personally, I’ve yet to make that spectacular world-beating one-off shot we all have in our heads, but when the time comes, I want to know if I should use VR or not. After making thousands of pictures with and without it, I still don’t know if it’s a blessing, or just a mixed blessing.

Maybe there’s no answer to that question. Have a great Christmas-New Year!

http://www.mediafire.com/?ohdvxezyczv40x1

[Allann]

Would love to see the parody shot. One of my fab little birds.

[Kym]

Basic bird photography... (as discussed with Richard Hall at last group bird meet)

1. Av mode, f/8 or a bit less
2. Shutter ideally 1/500 or better
3. ISO 400 or 800 (whatever is needed to get that shutter speed)
4. 400 or 500 mm on an APS-C sensor camera, SR/VR/IS on if hand held
5. Monopod or tripod recommended
6. Burst mode shutter, birds move very quickly
7. Centre weighted metering
8. Usually centre spot focus
9. Single focus (not Continuous/AI Servo), esp. when lots of branches/leaves
10. Stalk quietly and at an angle to your subject
11. Wear drab clothing
12. Delete 98% of 500-1000 shots per shoot

:christmasparty:

[kiwi]

If you are waiting to capture the moment, I'd turn vr off, it will cause a lag in initial focus acquisition

[farmer_rob]

Kym, from what I've read, points 2 and 4 are in conflict. Thom Hogan writes (http://www.bythom.com/nikon-vr.htm):
...Care to guess what the sampling frequency might be? 1000Hz according to Nikon.... Nyquist tells us that we can only really resolve data accurately below half the sampling frequency, thus it can accurately only take out movements as small as 500Hz (1/500 second). While this sampling frequency is of the camera motion, it is not completely uncorrelated with shutter speed. For example, the shutter curtains only travel across the sensor at speeds above 1/250, exposing only a portion of the image at a time....

This makes sense to me, and fits with what I learned about frequency domain behaviour in engineering.

So back to the OP's original question which could be rephrased as "what gives the highest *probability* of a good shot - VR/IS On or Off?". I think by turning VR/IS on at 1/500, you have decreased the probability.

Unfortunately, the way the human mind works, 1 good result overrides probabilistic likelihood all the time. (Consider lotto, poker machines, gambling.) I think the occasional good photo at 1/500 with VR is not indication that there is a higher probability of getting more keepers with VR always on.

[Tannin]

Potentially an interesting thread, but the basic premise is flawed, almost to the point of being nonsensical. Of course you use IS/VR - it very rarely ruins a shot, and it very often improves them.

Now let's consider the few times when it "ruins" a shot - what is going on here? Usually (always?) it is a shot which just happens to be taken at the moment when the stabilisation system is re-setting - i.e., it has compensated for a larger movement and has "run out of room" because the camera movement is greater than its ability to compensate, so it has to pick a new centre point and start again. Perfectly normal, given greater than small movements. There are several points to note here:


We know that there is camera movement going on, and in fact we know that the movements are moderate to large (otherwise there would be no need for a reset) - in other words, we know that a shot taken without stabilisation or a very high shutter speed would be blurry anyway. So the shot the IS/VR has "wrecked" wasn't a very good one in the first place. Short answer: don't lose any sleep over that one, it was never going to be a keeper.
The higher your shutter speed, the less chance of it wrecking a shot.
The better your technique, the less often this will happen.


In summary, I'm tempted to write that the use of IS/VR "isn't even a question", because outside of certain very specialised circumstances, the answer is always "if you have got it, use it". But although the answer is very clear, is is nevertheless a question that should be asked and answered, and a very interesting one, so thanks for asking it!

[Wayne]

I have been playing with my 400/2.8 on a mono the past few days, shooting small birds at some distance, they barely fill the focus point in the VF. I find it is very hard to stabilise such a lens on a mono for such small subjects even with VR, and without it, near impossible. I need a gimbal for these little things....

[mbundock]

I've been trying to shoot birds with an EF 500, hand held, with IS on. I know a gimbal on a tripod or a monopod are useful, but not always feasible due to extra weight. I've been building up my 'Holding' time slowly, I'll probably end up with one arm like a Gorillas'! I for one am all for IS, any help is useful!!

[Patagonia]

Good discussion here, to put some more spice I can comment that in my last trip I found that VR on my new 70-300 Nikon Lens idoubles as anexcelent set of "binoculars" for spoting and enyoing wildlife, you focus and after a few moments the feel is as if you canched from being looking from a motorboat to looking from a big cargo ship, all the shake desappears....

With the VR on the Nikon 16-85 I notice absoluteley no difference up to the point now I doubt if it has ever worked!! :confused013

[arthurking83]

You won't notice it on shorter focal lengths as much as you will on say longer focal lengths.

it still works, but the obvious difference/advantages are at lower shutter speeds(as the focal length decreases), and usually in the images themselves and not so much through the viewfinder.

If you want to confirm that the VR feature is working, you should hear a small click noise(in the lens) as you half press the shutter button.

Also, to confirm the effectiveness of VR at say shorter focal lengths you woudl take two or more snaps of a static scene(wall, door, anything that is not going to move) and shooting for example at 16mm you woudl set the shutter speed at approximately 1/5s and compare the VR shots with the non VR shots.. all handheld of course.

[Max]

wow, valuable information here. But i think one point is being missed. VR or IS or in camera stabilisation is just another tool in your arsenal. Just like checking wind direction as many birds will take off into the wind. Line up sun set and west wind and you make for great light and hopefully a great shot.

How far do you screw up iso to maintain fast shutter speed? At some point some compromise is involved. As you often increase exposure shooting up into the light VR IS etc just helps sometimes and doesn`t other times, that`s all.

[fabian628]

The only thing i can say about the image stabiliser being a problem might be first aquisition of the target. However, since every target is different in its own way, it is difficult to get a controlled test.

I did feel i had problems when I saw a target, swung the lens into motion and saw the object in the viewfinder, and half pressed, the lens was still in motion, and the IS swung into ctivity causing me to lose sight of the target in the viewfinder. So unless you are violently swinging the lens and turning on IS at the same time it should be too much of a problem.
At fast-ish shutter speeds (1/400s) and faster using a 300mm lens i dont turn the IS on becuase of the small time it takes to start up.

[ecopix]

Coming back to this thread after many months of photography and now going through images on the computer, I can confirm that my VR, VC, IS definitely do degrade the image if used inappropriately. So I've made a rule for myself: only switch it on if shooting handheld at or below one stop slower than the reciprocal of the focal length. At higher speeds, the sharp non-stabilised images are sharper than the sharp stabilised images. Often the difference is slight, perceivable in the 'compare images' function of ACDSee or similar, but worth it. The danger zone seems to be around the reciprocal of the focal length, but can even be visible at 1/1000th with long focal lengths. Controlled tests have confirmed this.
My lenses are 100-400 Canon, 70-300 SP Tamron and 80-400 Nikkor, plus an 18-105 Nikkor. I can't comment on others.
What a great thread this has turned into!
Wayne

[Kym]

This is an interesting take on SR http://forums.dpreview.com/forums/read.asp?forum=1036&message=19647176