Nikon explains why in-lens VR is better compared to in-camera VR

In a recent article, Nikon explains why in-lens VR system is superior compared to in-camera VR:

  1. Corrected finder image makes photo composition easy.
  2. Each lens is optimally tuned to achieve reliable correction.
  3. Image information captured by the AF and metering sensors is corrected with in-lens VR.
  4. Patterns of image blur are not the same with all lenses.

Here is Nikon VR in action (view from inside the lens):

Thanks Broxibear!

This entry was posted in Other Nikon stuff. Bookmark the permalink. Trackbacks are closed, but you can post a comment.
  • D40-owner

    That diagram is clearly a D800! It’s here! Yay!

    • The invisible man.


      • Mim


    • R R

      not really , a D800/900 would have an eye cup pretty much like the D700 removable and capable to accept accessories like the DK-19 rubber eye cup.. the one in the graphic resembles a D300 rubber eye cup , so it maybe a D400 in my opinion. But then again.. is just us wishing , and waiting.. wishing .. and waiting.

      #IAmNikon #IAmWaitingForEverNikon

      • I didn’t know this was Twitter.

        • The invisible man.

          what’s Twitter ?

        • R R

          Oh yeah, you are right, this is not Twitter, my bad.

          #IAmNikon #IAmWaitingForEverNikon

      • They’re willing to tell ya. They’re wanting to tell ya. They’re waiting to tell ya!

    • Can Wishingwell

      The Dream Machine –> zzzzzzzzzzzzzzzzz

    • NanDub

      It could simply be an N80… Who said it’s a digital SLR? 😛

      • WoutK89

        Practically saying it is an image sensor gives it away…

        • Lio

          Film is not an image sensor ?!?

  • Thnx for the explanation. I like f. 2,8 without vr better than 4,5 with vr though

    • sade

      I like 2.8 with VR better than both. Nikon give us a 24-70 VR.

      • Jeremy


        • The invisible man.

          I like to take pictures of the pyramids with my 20mm f/1.4 at ISO 3200 and my VRII on
          (well, sometimes I also use the tripod)

    • pavel

      I liked 1.4 with VR on my Pentax than 1.4 without on my Nikon. And then I like my wide angle 2.8 with VR. Oh, and my old 400mm Tokina with VR too. And also my manual old russian lens… with VR!

  • The invisible man.

    Yes it make sens !
    For some reasons my wife don’t like vibrations reduction, I wonder why.

    • twoomy

      Ha ha… *sigh*

    • Artur Kozłowski

      I guess she likes the buzzzzzzzz….

  • Spotpuff

    These points are true but there are obviously counterarguments:

    1. Mirrorless cameras mean viewfinder image CAN be stabilized with in-body stabilization (IBS)

    2. Cost; obviously, non-vr lenses are cheaper and there’s less cost to buy a body with IBS than every single lens having to have it.

    3. Lens size: VR adds to lens size. (of course a consequence is that IBS adds to camera body size, but you’re only carrying 1 camera and several lenses usually).

    I would tend to agree for now that in-lens is better though. EVFs are horrible and laggy in my experience, and it doesn’t get better in low-light.

    • Spotpuff

      Oh right not to mention IBS tends to heat up the sensor. I believe the Sony cameras have sensors that heat up really fast if you turn on IBS and video.

      • PAG

        IBS? Irritable Bowel Syndrome? Man, that in body stabilization must be frustrating.

      • The invisible man.

        I just tell people not to move when I take the picture.

        • PAG

          Do you do that right before you fire the flash powder?

          • The invisible man.


      • Chester

        Sony cameras do heat up during video if sensor shift stabilization is enabled. However, in my experience with the Sony a580, it takes a while (well over 10 minutes even on a warm day) to heat up to the point when it gives you a warning.

        I’ve rarely taken real videos longer than 10 minutes, but I did do some stress testing to see how long I could go before I got the overheat icon.

        • Whether or not it gives you a warning, heat = noise

    • Sky

      There’s more:

      4) Stabilized picture in viewfinder == you never know how much you actually shake and how far stabilization is stressed, so you actually never sure if VR helps or if it doesn’t.
      5) in-body VR cameras usually have added vibration-meter to viewfinder which is uber-helpful
      6) You have VR in tons, tons of lenses that don’t have it with systems where VR is placed in lenses (ever seen stabilized fisheye? or even: 50mm f/1.4 lens? ever? than you have lost)
      7) In-body VR doesn’t mean you won’t have access to in-lens VR. Sigma produces lenses with build-in stabilization for you to choose which one is preferred.

      But there’s only one real disadvantage of in-body VR, something Nikon forgot to claim, something that’s an actual disadvantage, not FUD spread by Nikon:

      You will never get it upgraded. Nikon or Canon users can enjoy better VR as they buy newer lenses. Oly, Pentax and Sony users will always have the same. Which in case of most recent high-end lenses is worse than in-lens stabilization.

  • sirin

    which makes the lens overall more expensive, but hey, this is why we’re with Nikon – to get stuff up to a quality, not down to a price.

    • Tony

      I was going to say that making VR in lenses benefits Nikon the most. I changed my mind because what you had said is true (though not for me but in general).

    • +1

    • Mock Kenwell

      +1. Ain’t it a bitch, though?

  • Parci

    What they don’t explain is that Nikon’s VR system operates/samples at around 1/500s frequency, in other words, it is actually counterproductive at faster shutter speeds. This is not necessarily the case with IBIS.

    • Paul

      What is your source for this? I suspected that was the case but haven’t seen a concrete number associated with it.

      • nobody

        Here is a detailed explanation by Thom Hogan:

        • Jesus_sti

          But over 1/250 you can cut the Vr fonction ….. 400 mm and more use a monopod it’s to heavy anyway !

          • Exactly—in 99% of all practical shooting situations, the spec is perfect. That 1% (or whatever) isn’t something to haggle over.

            • PHB

              Nikon know the parameters of their system. Surely they turn it off for short exposures in any case?

              They can’t know if the lens is on a tripod or monopod but they can check that one and I will bet they do.

          • The invisible man.

            The only lens I have with VR is the 105mm macro, and guess what ?

            – When I take portraits at f/2.8 I don’t need the VR.
            – When I do macro with the Flash I don’t need the VR !
            – When I do low light pics it’s with the 24-70 or the 14-24mm !

            Can I transfer that VR on my Af-s 300mm f/4 ?


        • Marc

          Why do these article “don’t use VR articles” never have pictures attached to them showing said artefact in a picture?? I have no idea what I am looking for? I’ve shot high speed with VR and never noticed any thing different??? I like to fling from one end of the exposure range to the other so I leave it on most of the time. Also I don’t understand his argument either. The reason we shoot fast is to stop things moving the. The VR element is a moving object ergo it just like every thing else is going to be just as frozen as every thing else. If your shooting over the sample frequency its not even going to have time to start moving before the shutter has closed. The centring may be done for this exact reason to remove any issue of an odd angle. I’ve try shoot my camera on high and taking photos with the VR on and I am noticing nothing different. I have a D7000 and 18-105mm vr I’ve tried both end of the lens and nothing. If any one has some examples of this artefact can they please post it?

      • Soap

        The interesting thing is not the source for why/how VR samples at 500hz, but that IBS does.

    • TaoTeJared

      It is basically the same as using a tripod with VR on (with most lenses). If you are shooting fast enough to counteract any vibration the lens doesn’t know that and still trys to compensate. Anything above 1/500 turn VR off, you shouldn’t need it anyway.

    • I don’t think I’ve ever shot anything at 1/500s and thought to myself “I could really use some VR right about now”.

    • Jabs

      @Parci and others here.
      The simple advantage Nikon has over others with In Body systems is that you need to program the camera for too many variables when it is built-in. Jack of all trades and master of none.
      In lens = better tuning for the actual lens itself and then the body is not a throwaway when newer VR algorithms come about.

      Nikon actually has lots of real Pro lenses and greater variety than anyone out there (including old lenses) and thus in body systems would limit compatibility. How are you going to program in old non-electronic lenses then and with Sony’s system, you are stuck in ‘firmware update land’ or ‘maybe’ Sony’s knows all the lenses that they are going to produce in 5 to 10 years?

      Putting the cart before the horse = Nikon has a history of products and Sony does not, so they can put the cart first as they are NEW to photography and it shows in their choices and results. Nikon puts the horse where it needs to be – after the fact or in the ever changing lenses = the cart.
      Body = horse, lens = cart!

      • Ben

        “Jack of all trades and master of none.” True for humans, but not for cameras. There’s absolutely no degradation to the quality of VR parameters for your 50mm f/1.4 when you add VR parameters for your 18-200mm. None. Whatsoever.

        “and then the body is not a throwaway when newer VR algorithms come about” But all your previous lenses are now throwaways. Since the body is guaranteed to be a throwaway every 2-3 years anyway (sensor/metering/video/processing/etc change much faster than VR algorithms) it’s stupid to worry about making just one feature of a digital body obsolescence-proof–Especially by offloading that one feature to a part that would otherwise have a 10- or 20-year lifespan. Unless you can upgrade the firmware of your old lenses…

        “Nikon actually has lots of real Pro lenses and greater variety than anyone out there (including old lenses) and thus in body systems would limit compatibility.” Huh? An in-body system that turns off when it sees a lens for which it hasn’t been programmed would still be available for many of your more common old lenses. An in-lens system is available with NONE of your old lenses. You’re saying you prefer the latter?

        Actually, you were being a sarcastic troll, weren’t you? Sorry, there are just so many fanbois on this site. It’s hard to tell who’s serious.

        • Sky

          “An in-body system that turns off when it sees a lens for which it hasn’t been programmed” – it doesn’t work this way. If IBS doesn’t have any profile specific to the lens than it reads it’s focal length – and that’s enough to stabilize lens with no noticable disadvantage. Alternatively you can manually calibrate IBS to focal length (in case on Pentax body + M42 lens) or use generic 50mm profile (which also works, but isn’t as efficient as dedicated methods)

    • Jabs

      1/500 sec is not Frequency but time.
      Plus sampling would not necessarily be tied to shutter speed as sampling is cycles per whatever and shutter speed is fixed in fractions per second or such.
      So – baloney!

      Sampling is often fixed while shutter speed varies!

      • Soap

        1/500 sec is not Frequency but time.

        I’m boggled!

        • Jabs

          @Soap – if it were a frequency measurement then it would probably be in MHZ or such and NOT in fractions of a second or 1/500 = bogus information or THEY misunderstood or made a wacko point – lol

        • Jabs

          @Soap and others here.
          Too much guesses and not enough data FROM Nikon = edumacated guesses by bloggers.
          1. Nikon did not say what their data specifically referred to and only said one measurement, and since there was NO information forthcoming from them as to WHAT the measurement referred to (due to them NOT being morons and give away their Trade secrets), then interpolating data or supposed ‘acts’ from that makes you an idiot. If I don’t tell you the specifics of something like HOW I got the frequency measurement and whether it was using a feedback or feed-forward circuit or using comparative analog or DIGITAL data from what part of the camera or lens plus its’ bit structure, then you are lost. Nikon is not stupid to tell what and where the measurement came from, so guessing means YOU are clueless.

          2. Taking out your calculator and broadcasting your findings all over the Internet, does not make you right. Do yourselves all a favor and go look at and STUDY at a web site like or the web site of National Instruments (plus go to an Engineering College) and get up to speed in the REAL world of Engineering and not ‘calculator gymnastics’.

          Moral = IF I have not told you what I have measured and how plus from what place in the circuit and now you are making assumptions of ‘facts’ from this, then you are showboating to get web site hits from gullible persons. PROVE your claims from Nikon’s patents and go work at Nikon in their Lens Division and then give away their secrets (jail time, hello!) to prove that or quit jiving us.

          OR simply ask Nikon HOW they did it and then come tell us here and we all will have a laugh at YOU, if you really believe that they will tell you exactly how! Yeah, Nikon is dumb and does not make semi-conductors nor is NOT one of the best and most respected semi-conductor manufacturers on Earth – GET REAL.

          People seem to not know the difference between ‘marketing speak’ and real information in which Companies tell YOU data but not how or what of an item to keep their Trade secrets INTACT!

          Frequency of WHAT – oscillator or what? – digital or analog signal? Light or copper wire signal? multiple instances of the signal, comparative signals or what? – answer that and see your folly – GEEZ, damn Internet showoffs!

          • Perhaps frequency isn’t the right word. Maybe latency. Anyway, I don’t think it’s too far fetched to think that it would take as 1/500s for the motion sensors to observe the motion and for the motors to move into place. What I don’t understand is how can that latency hurt you when your shooting at a high shutter speed? Any motion blur that the vr system would make would be also be demonstrated at longer exposures, only several times. Maybe because the motion blur steps make up such a small amount of the overall exposure?

            • Jabs

              Answering both your posts.
              The so-called ‘sampling frequency’ tells ME nothing whatsoever as I don’t know HOW it is employed, so guessing makes me as clueless as the blogger is.
              WHAT are they sampling comes to mind and from where, as in what stage of the circuit does this refer to!

              Sorry to be blunt but I am like that, as I work in Industrial places and thus, I am like Dragnet – ONLY the facts and nothing else.
              I also do not now whether it is digital converted to analog or whether it stays completely in the digital domain or what and thus I have nothing to go and guessing would only feed MY ego and not render facts.

              I seriously doubt that you would sample anything at that LOW a frequency as in MHZ or such and then have the data come and go WHEN the camera itself is shooting at a HIGHER RATE than 1 operation per second, so the mistake made by the blogger is that they used a calculator and NEVER factored in the exposure SEQUENCE of the camera in question (D3s = 9fps) and thus you MULTIPLY your findings by at least 9 times as he erroneously USED 1 second as their basis – hence BALONEY!

              1/500 sec = 1 second divided into 500 equal or distinct parts while the BASE reference there is ONE SECOND. Up that occurrence and you clearly see the mistake as the camera’s maximum framing rate did NOT stay constant or 1 second, but exceeded one second intervals, hence ‘edumacated guesstimations’ – since the camera takes multiple images.

              Latency refers to the DELAY that occurs when a signal is supposed to be somewhere and is not there like as in you sending two 8bit data structures in different pipelines and they are supposed to reach a certain point where they are now combined to make a ‘pseudo 16bit (8+8bits = 16 bits) homogeneous signal’ and something is late, hence latency or lateness as to where it was expected to be.

            • @Jabs,
              What field of engineering are you in? My degree is in computer engineering so here is my thinking: I don’t know how the motion sensor works but neither does the processor. All the processor needs to know is how to read the current position or change of position or whatever. How frequently the processor reads this is sampling frequency (how often you take a sample from your input device). Maybe a frequency of 1000Hz was decided because the sensor won’t pick up on faster movements or maybe they just thought that level of fidelity was high enough. Anyway my point is that the sampling frequency probably has nothing to do with the hardware.

              As for latency, although what you described is an example of latency, that’s not really the definition. I was using the word to describe the time it takes you to see the results of the operation you initiated (I know, not the best use of the word). But to correct your definition, latency, in terms of signals, is how long it takes a signal to get to its destination, not a signal being late to its destination. Every signal has a latency.

            • Jabs

              @ Lewis and hopefully my last post on this matter.
              1. Most people who do electrical wiring = number of colors, junctions, devices, wire gauge etc. – think that electrical and electronic wiring are the same even though they have similar references – therefore Internet buffoons. The VR system by Nikon clearly has both analog and digital signaling plus special ‘motors’ that really are not motors but the term was used for consumer consumption – like how Nikon moves it’s lenses with special silent wave ‘motors’ or even coreless units!

              2. Most people who read things and do not have a basis of knowledge in the item, cannot understand the fine details nor terms, as they take certain words that have different meanings in different fields as having the same meaning and thus it confuses them.

              3. Latency has too many definitions for me to further comment as there is latency in a digital circuit and then latency in an analog circuit from say resistance or such. There is latency within the device and latency in the external interconnects, signal pipeline, circuitry – example 64 processors can send data as two 32 bit streams to add up later on to 64bits or they can send 64bits through a wider pipeline (say a 128 bit pipeline) and then now not have to combine them later, hence less inherent latency. Computer latency is one thing and latency within the external links is another type. CPU internal latency is not the same as say interconnect latency or FSB latency.

              4. I get serious and then I don’t want to deal with people and their nonsense, so I leave them alone.

              5. Engineering background – Electrical, Automotive, Electronics, Computers, Designer of equipment, Material Science, FS controls, Computer to Equipment interfaces and a working person solving daily Engineering problems in equipment – don’t like to talk about self but come here more to relax and be an ‘idiot’ or have fun at a nice rumor site. I thus only show my skills when challenged and I am a specialist in my field. I do Technical photography in my field also but for ‘fun’ I shoot street photography, nature photography and such plus video work, some 3D animation, graphics, compositing and more – had enough!

              Sort of like the geek who hides their capabilities so they don’t always talk over people’s heads and make them uncomfortable – clear enough!
              I prefer to be left alone and enjoy life plus not be some damn guru to people – leave people at their own levels and not look down on them = my motto! Life is too short for crap and conflict, hence a ‘mellow’ person – lol.

          • Just read the blog post linked to above and the frequency that they mention is the sampling frequency (1000Hz). That doesn’t have anything to do with the electronics of the VR device. Sampling frequency is just how many times/s you read your sensor (1000 reads a second in this case). I think your over thinking this one.

          • Carl

            Dear Jabs
            being an engineer myself I can understand some of the frustration you have with people posting their guesswork or simply wrong things. However, didn’t you say that you have 30 plus years of experience in photography? So you are not a teenager anymore but a mature person and I would wish that this fact would bear down a little bit more in your wording.
            We may not be on the same page all the time but we (hopefully) be elaborating in the same book (photography) to remain in this image. While living in the US for some years I’ve learned a very helpful saying: “eat the meat and spit the bones”.
            There may be a lot of junk in a blog like plain guesswork, wish-lists and personal hopes. But there are also some bits and bites I personally have benefited from. Besides it is sometimes interesting WHY people care or care not for stuff that I haven’t recognized or do have quite a different approach. Just take the introduction of the 40mm Micro Nikkor: I personally wouldn’t see much application for this lens since I’d prefer more distance to the object. However, I would care to hear different opinions.
            Nikon does not only keep silence about trade secrets (which is understandable) but I haven’t seen Nikon publishing information relevant for the users like the initial opening of a Micro Nikkor goes smaller when approaching makro distances (from f/2.8 to f/4.8 with the 105mm VR), or that the 28-300 is max a 130mm at the closest focus distance. Bottom line, I already could benefit from the one or another comment, although some more of the serious and informed bloggers wouldn’t hurt.

            • Jabs

              @Carl – thanks for understanding but I don’t relate to them nor do they relate to me, so I prefer to NOT discuss this anymore as you can’t fight misinformation on the Internet and then try and explain to people who are not of your background. Simplicity is for the uninformed and thus back to work for me.

              Yes, I am a photographer, so the artist side of me understands their quests or mania, but I cannot get blood from a turnip nor am I willing to waste more time on this issue.

              What I have learned in life is that you cannot teach people things that they cannot grasp and perhaps that is why there is so much lack of Engineers in America and YES, I do live and work in America and being in an Industrial environment, I don’t relate to consumers and thus often I just prefer to keep quiet here.

              The oddest thing to me is that people think that 1+1 always equals 2 and then take this same mindset to electronics and fail to understand that the principles of electrical measurement is NOT the same as electronic circuitry – so they ‘google’ information and have no idea what it is, as they don’t know if the information is real or bogus – when you tell them that, they shoot the ‘messenger’ for waking them up – well not a target anymore and thus ignorance = bliss!

              It has nothing to do with age but people’s lack of wanting anyone to tell them that they are wrong – do you see the Administrator of this site getting into arguments with people here – smart person and thus time for me to keep quiet and perhaps after it sinks in, they might or might not get it.

            • @Jabs
              I’m sorry you feel this way. I know I don’t know everything and I probably have a different way of thinking than other people so I look at this forum as place for me to grow in my understanding.

              Let’s say I think I know exactly how VR works and I post a comment about it but I’m completely wrong. If someone corrects me and I have to do some digging to find out how it really works, I come out more enlightened (although I don’t always get it right away). If I think I know and I keep it to myself, I stay ignorant. Of course the other side of this is that if someone else is misinformed and I do know what I’m talking about, they may come out more enlightened.

              Yes some people are rude in their responses but you have to take the good with the bad.

              p.s. I assume by your logic your background is in electrical engineering? I work in software so I like to turn things into black boxes as much as possible.

        • Jabs

          Here is what Nikon says about their own VR system and not rampant guesses!

          • Jabs

            Explaining it to you simply perhaps!
            Nikon uses X, Y and Z coordinates to do their in lens VR via two voice coil based MOTORS and feedback from BOTH the lens and the camera’s electronics.

            X, Y and Z refer to side to side, top to bottom and front to back (Z), same as when you do 3D work in say Lightwave 3D or Maya or even computer based 3D animation.

            Body based sensor VR cannot or may not move in the Z direction or they might increase or decrease their distance from the sensor to lens element or the focal plane of the mounted lens and render autofocus unreliable, thus they often use mirrorless bodies and basically ‘shake their sensors’ to compensate. NOW, do they move them front to back or NOT = what you need to answer or shut up.

            Real Engineer here and not blogger!

            • I read the link you posted and it says that Nikon’s VR has two sensors and two motors. I don’t think it would be wise to move lens elements in the z-direction anyway as this would affect your focus (unless you were correcting focus but that wouldn’t be a VR element, just the AF motor which is conceivable since it’s all wired the lens’ Microprocessor).

            • Jabs

              The Z direction is FRONT to BACK in this case and refers to the lens movement perhaps.

              You cannot calculate things in essentially a 3D environment (3 axes, axes = plural of axis) with a basically 2D calculator = the blogger’s mistake.

            • Soap

              Besides your appeal to authority fallacy it if bloody F^&*(ng obvious they don’t use coordinates. Blur is an angular problem and one would never reduce a native angular problem to coordinates before applying an angular solution. It is a ridiculous step to take and shows a lack of understanding of the problem and solution.

            • @Soap
              When you say coordinates do you mean quadratic coordinates (ie X, Y, Z)? What about polar coordinates?

            • I think you meant to say:

              “What about polar bears?”

              And, indeed.

            • Yes, polar bears. Sorry for the typo.

            • Soap

              @Lewis. Either.
              Polar coordinates implies caring about distance, which one doesn’t. Angular velocity is what is cared about, not any sort of projection the word “coordinate” implies.

  • nikigo

    wow. thank you Nikon!!! I guess Nikonians are so dumb that they need drawings for kids to understand such complicated stuff like sensor vr vs lens vr.

    • Darkness

      Guess only Nikonians buy Nikon cameras then?

      • Sky

        Some suggestions that only Nikonians are smart people? lol

  • Money money money. I sold my only lens with VR. I never really used the VR function as I preferred fast shutter speeds, and (even with a D200), higher ISO. Too bulky for me in a 2,8 lens.

  • I wishing for a funny 24-70 2.8VR, 17-55 2.8VR and 50 1.4VR.

    I think it will sell a lot because people buy a camera & equipment from specification.

    • How about a 10.5mm fisheye VR III? 🙂

  • In lens stabilization may be technically better, but it is much more expensive for the end user, and a bit fragile as well. It cost me $165 to repair the VR on my 70-300VR.

    Besides, it is great for planned obsolescence.

    • Sky

      In-body VR is also fragile. For me it was 150$ to repair it, which isn’t bad as I see, but never the less – it’s no less a moving part than in-lens VR. OK – maybe the in-lens VR has one disadvantage in that area: If a bar full of lenses falls on the ground than people with in-body VR have nothing to worry about 😉

  • Eric

    What about a comparison between in-camera VR and no in-camera VR?
    1) How much would it cost to get e.g. a D7000 with VR?
    2) I would love it for my 50mm 1.8, 35mm 1.8, 80-200 2.8, 105 2.8, 11-16/2.8,…
    VR alternatives don’t exist for every lens, and upgrading more than 3 lenses would probably cost more than buying a body with in-camera VR.
    3) How much in-camera VR is achievable? Even 1 or 2 stops would be very welcome.

    • Chester

      With the Sony a580 and a 200mm lens, I’ve gotten sharp images at 1/60s, provided that I’m in a steady position. I usually try to keep it above that for moving subjects, though.

      • With the Nikon D7000 and an 80-200mm lens, I’ve gotten sharp images at 1/60s, provided that I’m in a steady position.

        Only the D7000 doesn’t have IBS. 🙂

        I have found that most times that VR comes in handy, it is usually only reliable when I hold really still. Which is what I had to do when I DIDN’T have VR. So…..

  • John

    This doesn’t seem to make sense. Since when is the image written to your eyeball. If you lift the mirror and take a picture doesn’t the diagrams look pretty close to the same? They are just using the prism assembly to make their system seem superior.

    • Artur Kozłowski

      – they’re trying to show the physical difference when composing with IS on (in the lens) and off (in body).

      • – they’re trying to show the physical difference when composing with VR on (in the lens) and off (in body).

        That’s better.

        • John

          Their attempt to make it for the average person to understand is underselling the technology and borderline false.

          • Care to elaborate? I can’t really see anything false or even “borderline false” about the 4 points they’ve made. Ok, so maybe #2 is underselling in body VR, but what about the rest?

            And are you saying that it’s underselling their own technology, or that of competitors who have decided to offer IBS?

        • Artur Kozłowski

          8^) I shall not write such heresy in the future, master.

    • iamlucky13

      Read point 1 in the post summary – A stabilized image delivered to the eyeball makes composition easier.

      This is mainly true with telephotos rather than wide and normal lenses. Personally, I don’t find it valuable so much for framing, but it is helpful for keeping the autofocus sensor over the intended target.

      It’s also nice simply for watching a subject. 300mm on DX makes a decent substitute for binoculars, and it’s easier to tell what a deer is doing, for example, with VR on smoothing out your jitters than with VR off.

      Lastly, points 2 and 4 are not demonstrated in the graphic, but are also valid.

      There’s pros and cons each way, although none of the above reasons are ultimately why Nikon chose in-lens VR. They did so because of the large numbers of current Nikon owners they didn’t want to leave in the dust.

      • Sky

        “but it is helpful for keeping the autofocus sensor over the intended target.” – would be if there existed any proof for that other than over-thought theory. I never had any problems with keeping AF in point while using tele-lenses and suffering shaky-hands. 😉

        So it’s not really a good point. No better than one that IBS is better at stabilizing shake caused by mirror slap than the in-lens VR.

  • dan

    I can see why Nikon are keeping it in their pro lines as when you’re working at the top of the field you need the very best, and when it comes down to it, that is in-lens VR.
    However I imagine the difference is slight at best, and to be honest, the huge size and cost increase for a VR vs non-VR lens, plus the fact it’s only featured in certain lenses…I’d much rather in-body stabilisation with the ability to turn it off for better lenses.
    I want VR for my AIS f/1.4 primes dammit.

  • Hellen

    the only trouble is that adding a variable element can lead to poorer optical performance and, as it was the case with one of my VR lenses, they tend to fail after a few years of operation.
    they should have kept from adding VR or even internal focusing motor, as these things degrade quality and life cycle.
    the old manual lenses are still around and working. these plasticky bits with motors and variable somethings won’t last the decade…

    • Chester

      Sensor shift stabilization isn’t perfect either. If you move the sensor too far to one side, you get closer to the edge of the lens’s imaging circle, and optical problems with lenses tend to get worse as you get close to the edge of the imaging circle.

      • To that I’d add that the relative mass and size of the 35mm sensor means that more power and acceleration/deceleration would be needed. I’m not saying FF would need a car battery, but more juice certainly, than in-body stabilization in comparison to DX.

        I’d expect that we’ll eventually see in body stabilization for FF eventually from some of the smaller manufacturers who don’t want the expense of rejigging their lens manufacturing buy FF will have to be well lodged in the consumer market for that to happen.

        • Heidfirst

          Sony launched the A900 FF with IBIS in Sept. 2008.
          Canon said that it couldn’t be done … 😉

          • sandy

            And just how many FF DSLRs would you guess Sony has sold since then compared to Canon? They probably surpassed Pentax FF though.

            • Sky

              And what does it have to deal with the stabilization? You know that it’s one of least important factors in buying a camera, don’t you?

    • Artur Kozłowski


    • M35G35

      I have the first generation 18-200VR and it still works great. 5 1/2 years old and still operational.

    • Darkness

      Oh, “they tend to fail” do they?

  • The SR system in my Pentax K-5 works pretty well. And the advantages such as automatic horizon leveling correction, a small rear shift system when in live view on a tripod, motion tracking with the movement of the earth when shooting astrophotography, and the cost reduction on lenses is what I see as MAJOR advantages over uber expensive VR systems in each lens.

    • Harry

      But Pentax hasn’t reworked their ancient lens formulas in 15 years. And, by this point, may have lost the IP and skill set to do so. Even the 645D lenses were sitting in a warehouse for half a dozen years before Hoya bought them so adding in body is probably not only the cheapest route for them to add features, but the only route.

      Here’s hoping things improve for them under Richoh.

      • I’ve found my DA* 16-50 2.8 is nearly as good as the 17-55 2.8, weather sealed, and MUCH less expensive. And the DA and FA primes are ta huge reason why I switched from a Canon 5D Mark II to the K-5. Considering the build, AF speed, and lens quality of the K-5, I’m happy to sing it’s praises.

      • Guy

        You talk as if all Nikon lenses are new optical designs.
        Especially for primes many lenses are very old designs.
        Only the Aperture shape and the Coatings have been updated.

        With using such an old mount Nikon would actually benefit from VR.
        Although it would probably need an AF lens.
        I know that the Minolta/Sony system needs the focal length, so an adapted MF lens can only get VR if you have a correctly chipped adapter.
        But well it works with all Minolta lenses since 1985 (well I think not with the 3x macro)…

  • Slon

    Of course in lens VR is better ! This way you sell it with every lens instead of selling it once inside the body…Way to go Nikon, you make more money from your fans this way!

    • The invisible man.

      Nobody force you to buy Nikon, there is plenty other brands !
      I will never buy a Ford or Crysler car….

      • iamlucky13

        I find the inclusion of your second sentence a little odd…my Ford Taurus takes great pictures.

        • The transmission is somewhere along the 101, but the camera, no doubt, rocks.

        • The invisible man.

          you must work for the US gvt, USPS ?

    • Analyst


      Nikon and Canon produced in-lens stabilization because they knew they could get away with it not because it is technically better. The other manufacturers with less market power who are trying to gain share are the ones who put it in the bodies.

      Every reasonably scientific test out there shows the same thing every time – they’re about equally effective.

      • Slon

        good point!

      • pavel

        Unfortunately, you’re right. I bought Nikon because the picture quality and decided to live with their lack of in-body VR. But I still hate it.
        My (former) Pentax was better not only because it stabilized everything you put in front of it but it also did it better than Nikon!
        Yes, I said it. It was 1.way more consistent, 2. way more efficient, 3. it worked instantly and 4. recognized tripod and I never needed to switched it off.
        If Pentax had system range similar to Nikon, I’d have never switched.

      • El Aura

        Nikon and Canon produced in-lens stabilization years before they produced digital cameras. And I don’t think anybody could say with a straight face that film stabilisation would have been the better option back then.

      • Mock Kenwell

        Totally disagree. I think anyone can tell that on paper, ILS is better for the myriad reasons stated in the article and posts above. But it will of course cost more. Nikon always errs on this side of thinking. This is precisely why they haven’t produced an f/1.2 lens in years—too many quality compromises required. If you hate this mentality, move on. Canon takes more risks, and some of them actually pay off—but many don’t. Pentax et al make IBS decisions based on available cash and resources, not quality concerns. If you revel in quality over quantity/cost-effectiveness—however minimal or incremental you believe the improvements to be—then you must pay. And wait. This is the Nikon way.

        • Thank you, MK. Nikon is a premier brand, and can command a premier price. Let’s say VR is not revolutionarily better than IBS, but enough of an improvement to make a noticeable difference. Isn’t it GOOD that Nikon commands the kind of market value to be able to charge more for the better solution?

          I’ve long said that Nikon will take their sweet time getting it right (it being whatever they bring to market). It’s not because they’re lazy or stodgy as some have implied, but rather because they care too much about quality to make compromises that only amount to short-term wins.

          If this is the meaning of old-fashioned, then to all the other companies I say BE A LITTLE MORE OLD-FASHIONED!

          • Ryan

            I have no worry on this approach, at some point, most consumer will be sold by the quality you deliver.
            Just look at how apple do with their iphone, they take their sweet time to polish the existing phone features, and slowly launch it one after the other. But they are still conquering the market share out there, eventhough they are lacking behind the competitors in term of features.

        • Sky

          “I think anyone can tell that on paper, ILS is better for the myriad reasons stated in the article and posts above.” – Epic fail in pretending to be Nikon fanboy.

          Oh way… there was no pretending! :O

  • broxibear

    Here’s a youtube video that shows Nikon’s VR system at work, sounds like a Kraftwerk soundtrack too (I think?)

    • broxibear

      P.S. I’ve just seen it says “Afrika Bambaataa & the Soulsonic Force” but that’s definitely a Kraftwerk sample they’re using.

      • Jabs

        It was called ‘Planet Rock’ and yes based on that group – interesting.

    • The invisible man.

      I just watched the video, does not worth it, thank you anyways.

      • BaoKeWen

        ….then you didn’t watch it 😀

    • Thanks again for all the links broxibear!

  • I fear to say that VR affects image quality on every setting of exposure pair. Of course, it helps for handheld shooting but I did short tripod test on this question and test revealed such problem. Image quality is acceptable when VR is ON but it is slightly better when stabilizer disabled. Results may vary depending on what lens you’re using. With DX kit lens 18-55 VR difference is non-existent, since resolving power of this lens is limited, but with 105 VR difference is more visible. Also, bear in mind that 18-55 is under old VRI technology and 105 micro is under VRII technology. I didn’t tried other lenses with stabilization. Hoping for your responses.

  • Darkness

    Looks like no-one here has tried to lock a focus pin-point on something small lately? Oops, Sony cameras dont have focus points like that do they?

    • Heidfirst

      seems to work OK on mine …

  • Pat

    What ever happen to just learning how to hold a camera and lens? I routinely shoot weddings with an 80-200 (2nd gen from 1992 that I paid $400 10 years ago) on a d300. Settings rarely get much better than 1/15 – 1/30 @ f2.8 with an ISO of 1250 shooting at 200mm (300mm eq.). Sure I get a fair amount motion blur, that’s why I shoot in bursts of 3, at least one is always in focus.

    I own the 18-200 VRI and I have to say that the VR feature seems to blur more photos than a solid hand hold. In contrast, it helps my wife out a whole lot (who still can’t figure out how to get sharp photos at 1/30 while hand holding at 50mm)

    I understand the purpose of VR on the longer zooms, but why does anybody need it on a 24mm or an 18mm? In my opinion, you’re better off just learning how to hold still, buying old glass, and saving enough money to buy another body or lens.

    I am interested in trying out the 70-200 VRII, that sounds like it makes more sense for what I do.

    • Yes, yes! That’s it. In my case 105 mm isn’t long enough to exercise any problem with shooting handheld. For example, shooting at 1/60 wasn’t problem for me with 200 mm lens (personal record was 1/20!). More than that, shooting at 1/8 wasn’t problem with 50 mm lens, I can make crisp photos consistently at such fairly long shutter speeds in this focal range. BUT. But there is a confusion with AF-S Micro-Nikkor 105 mm f2.8 G IF-ED VR: I can’t provide steady images without VR at 1/500. Why? What’s the problem? The only answer is that weight balance of this lens is so awful that camera dives downwards every time regardless how steadily you hold a camera. Sturdy hands is not enough to hold combo of any camera and 105 VR and since VR affects IQ slightly (this is not an issue but little inconvenience), this lens became a tripod-dedicated tool. This confuses me alot and leads to disappointment. I had experience with fairly more heavy cameras and lenses and didn’t feel any bad about using them. Sadly, there is no word of encouragement for me about usage of 105 VR. 🙁

      • Pat

        Finally! I thought I was alone on the VR front.

        • GregD

          I use the AF-ON button, that unfortunately means that I pay for VR but never get the benefit but I get to pay the price.

          But I agree everyone calls for VR on the 24-70 but it seems like those people, unless they have a medical issue, would be better served by some instruction on how to hold a camera.

          I have a tremor but I can hold 70mm down to 1/30 with good results.

          I do wish VR did kick in with the AF-ON on the longer focal lengths though.

      • GregD

        How close are you to 1:1, as your magnification increases the usefulness of VR decreases.

        To you note that it is easier to get clean shots at infinity vs close focus?

        Also magnification (which is what is letting you get to 1:1) has an effect similar to focal length where it becomes more difficult to get sharp pictures at low shutter speeds.

        • Pat

          I rarely do much macro work off a tripod. I have a nasty habit of using an old Nikon 200mm Medical with a SB900 during some wedding shoots (namely when women are getting ready and putting on make-up). It’s an unwieldy lens with no focus mechanism other than physically moving the camera. Even then, I’ve managed some 2:1 and 3:1 shots at about a 1/30 of a second. Working distance is about 6 inches for 3:1 and maybe 9 inches for 2:1 (I’d have to check the manual for the real distances). Usually I just stick with the 1:4 or 1:6 ratios if I’m using that lens.

          If I’m doing serious work, I through it on a bellows set with an extension tube. I do this more for the focusing rail than for anything else.

          • GregD

            I use a focus rail to move the entire camera, I do have the super expensive 200 f/4 micro, it is great for other work though if you can deal with the very slow focus.

            For 4:1 work I just tape a 50 1.4D backward on the front of the 200 and focus them both at infinity.

            I find that moving the entire camera on the focus rail actually helps reduce the perspective change.

            • Pat

              I use the focusing rail on the PB-6 to get my primary focus and then the bellows to frame the shot how I want it.

              The medical will do 3:1 on it’s own through a series of close-up lens attachments. With the PB-6 and tubes, the field of view is pretty close to 3 mm or 4mm wide. Really, the only benefit that I can think of of the medical to the new 200 micro is the aperture, an impressive f/45 and it’s analog (in the sense that you have infinite control of aperture from 5.6 through 45 rather than click stops) and it has a built in ring flash (which from what I read from the manual, is pretty much worthless for what I do). I think I would rather have the new 200 and a good set of close-up lenses over the medical. It has been a great cheap ($175) insect lens that I just use whenever I can, mostly to prove to my wife that it wasn’t a worthless purchase.

            • GregD

              I guess this is too deep to reply to you directly, but here is a 4 times life size image taken by just taping a 50mm lens to the front of a 200 🙂

              Example IMage

              if you have a focus rail it is a great way to get a huge magnification.

              My 200 will do f/35 but I like to stay with f/11 and just go slower on the stacking to get more detail.

        • GregD, that explains alot and might work in most situations. The problem, actually, comes in when I shoot portraits. I haven’t enough money to afford both macro and portrait-dedicated lenses so decision was to buy the macro one in reason to have more universality and freedom in shooting. God, what this was a mistake! I’m not relying on VR if there’s need to capture something in 1:1 magnification, because tripod serves better, obviously. Especially in close-distance shooting slight loss of crispness caused by VR action is more evident, so stablilizer stays off most of the time. On the portrait side I can’t find a solid answer on the question why this particular lens behaves so unpredictably. Conclusion is that the heart of problem could be find on a weight balance skyline, but how true it will be? So, to answer your question:

          > To you note that it is easier to get clean shots at infinity vs close focus?

          that is easier to get clean shots at close focus. There’s just more techniques to be handy: from the tripod to the most improbable handheld yoga-style steady-poses. 😀

  • FM2Fan

    can anyone please explain to me, why you want VR for anything having focal length below 105 or 135mm? (at full frame)

    the key question is: if there is motion, than it may be interesting to adjust i.e. either follow the movements, but distinguish between frequencies of moves – that is exactly, that VR in a lens can do very precisely.

    I’d say: the technologies to move sensors have other benefits rather than only VR –
    think of the multi-shot modes of Hasselblad, think of other advustments, being useful, because the lens transmits information about absolute focal plane etc. …

    • broxibear

      Hi FM2Fan,
      I think you’ve misunderstood what VR is.
      VR isn’t to stop any vibration/movement/motion in the subject…it’s to stop the vibration of the photographer, camera blur caused by the photographer moving at low shutter speeds.
      The point of it below 135mm would be the same as any focal length, VR in say a 24mm would mean I could hand hold at shutter speeds below 1/30th. Depending on which camera body you have vibration levels from the shutter and mirror vary…the D3 series cameras make a hell of a slap.

  • Mark H

    Now substitute “focus motor” for “VR”

    Yeah, the advantages and disadvantages are different, but how would people reconcile the their preferences if they differ between focus motors vs. VR?

  • Camaman

    Nice reasoning, Nikon.
    But give us both… 🙂 😛

  • SSS

    But the reality is that having any lens work with adequate in-body stabilization is still better than select few lenses with really good VR…unless you smoke crack and drink a lot of coffee – then nothing works.

  • who knew

  • I think premium unleaded gas is better than regular but every company still offer me both options. We get it Nikon, VR in lens is better. The new Canon 100 macro has superb IS that sensor shaking won’t do… Good for them. Good on you, and me, Nikon for the sweet VRII on my 70-200.

    Still, why not put an in-body VR as well so we may use it with all lenses. We could simply choose which to use. The body would shut off in-body stabilization when it connects to a VR lens.

    Obviously the reason for not putting in-body stabilization is the same as for Canon. The other big guy doesn’t do it. If we stick to our guns, we won’t lose out on this cash-cow that is IS/VR.

  • Mk19

    The way I see it…..

    Generally, Nikon is a company that sells lenses and camera bodies at various (inflation adjusted) price points. It keeps its revenue stream by getting customers to rebuy those items at those price points (or by giving reasons to move up price points). Nikon does this persuading customers that each iteration represent improvements over the last generation, or building feature sets so moneyed enthusiastic folks have reasons to move up price points. [familiar behavior among us Nikon owners, yes? 😉 ].

    Honestly, without VR upgrades, most Nikon’s older lenses generally look VERY competitive to their current iterations …….. having VR in the new model lens, or “X% better” VR in the new lens, makes it much easier for the marketing folks to argue that the customers should buy the new one.

    The only reason any company would put out a flyer like that is for marketing. IMO it’s clear that Nikon is trying to keep customers from going to (presumably recently launched) cameras with in-camera stabilization.

    Who are these competitors? I don’t think it’s Pentax, I’d say its the m4/3. And the target audience wouldn’t be customers of its US$7000 camera, any up to its US $1200 camera. The customers they fear losing the $800 camera customers, as well as new entrants who are wondering why they should get a D3100 over an m4/3.

    So, I vote that this article is FUD, no matter how valid the technical arguments might be. And it seems to me “VR in lenses” was and is a marketing decision.

    D90 owner and possible D400 upgrader. Who’s keeping his 18-200 VRI. 😉

  • Jabs

    Here is what Nikon USA tells us about their OWN VR system versus what guessers say – look, read and then decide yourself.

  • Eugene

    +100 for new nikkor lenses all with VR please!!!!

    Adds a new dimension to creativity!! If Nikon came out with more VR lenses, Nikon would have all my money 😉

  • Jojo

    While I agree with Nikon’s article, and accept lens VR is a better solution, it’s only fair to consider some disadvantages (especially as some manufacturers have different solutions):
    1. An in camera system applies to all lenses
    2. Theoretically, VR lenses will probably have a shorter working life
    3. As VR lenses have, in effect, a wobbly group of elements, it seems that group can (very rarely) be in the wrong position at the exact moment of exposure, leading to less sharp results. Has anyone ever tested out this aspect?

    • El Aura

      OIS does center the VR elements just before the exposure so that while they naturally have to move during the exposure, they do it being as close as possible to their optimal optical position.

  • Chris

    The best would be to have in-body stabilization plus lens stabilization. When using a VR lens, in-body stabilization would be disabled. But often, you’ll use lenses which don’t have VR – even new ones – like the 50mm 1.8. There, in-body stabilization would still be way better than nothing.

    • f/2.8

      Agree. Only the longer focal length lenses may need an optimized in-lens-VR. Most wide to medium telephoto can be handled by IBS.

      I think it is just posturing for now since Nikon and Canon are not ready. Sooner or later they will implement IBS. It will actually give them a competitive edge when the cost goes down on the lenses by not having to include VR. Whether the cost savings will be passed on to the consumers is a different story.

  • Florin

    It seems that most of you here are real specialists in techniques, mats & stuff with all sort of computing & algoritms…but I have one question to you:

    How much time do you spend shooting & creating instead of fighting on forums about all this bullshit?!?


  • cirtap

    I don’t believe Nikon. Here is why!


    They make more money selling VR Lens’s than they would having VR in the Camera.

    • @cirtap:

      Nikon is in business to make money and prosper—just like you, buddy. Get over it. If you can’t deal with the idea of deriving profit from effort, then I suggest you sell all your belongings, move to the woods, and burrow a hole in the ground while you live out the few remaining moments you’ll have left of your untainted, unprofitable life.

      I love to see a company that is profiting from their products. That means people want to buy them, which in some cases is even an indicator of their products’ awesomeness. I love awesomeness, and therefore hope Nikon can start showing some awesome profit results on their financial reporting.

      • cirtap

        IT is always lovely when Trailer Park people come to visit. Thanks Ron..come back often, so that we all can enjoy your vile.

        • Yes, I’m an American, so naturally I live in a house on wheels. I also shave my head and wear a beer stained wife-beater featuring a picture of bikini-clad women on the front.

          Do you really believe Nikon is a corporation full of deceptive leaders with purely selfish interests as you seem to be implying?

          • Greg Robert

            No, but their marketing department, as all marketing departments, is.

  • Q

    The classic American cars are all back wheel drive. The modern Japanese cars are front wheel drive. Both have their advantage. But the supercars from europe have 4 wheel drive. No doubt that 4wd is gives the best performance. Sure it is very good with VR in the lens. But Nikon cant make a optical 1.4 or 1.8 lens with VR. So in the house stabilizer works best with high end optical lenses. Nikon has a 2.0 with VR and that is awesome. But there is still a big gap. So I see this as a pure sales bullshit.

    • Yes, Q, and it’s about time they engage in some marketing volleys. I can’t help but feel like this is Nikon taking a page out of Canon’s playbook—meaning they are responding to market threats and confusion by adding their own perspective to the conversation.

      Nikon isn’t just about making money—obviously. Look at their earnings reports. They are as keen on making great imaging products as they are on making money. So if this is a show of their efforts to defend their assumed superior position so they can continue to sell their superior products to us, then I commend them. I hope it’s a first step of many.

  • @Admin: Is it just me, or does this VR response seem to contradict Nikon’s characteristically formal (and silent) style?

  • Sebastian

    They might be right, the problem is: In-Camera VR using a f/1.4 lens is better than having no in camera VR, and no f/1.4 VR lens to buy. To get the full advantages mentioned, Nikon would have to offer all lenses with VR, this includes the f/1.4 primes.

  • mshi

    how many times you tend to use on super-teles, such as 500/4 and 600/4, shooting sports? VR is purely a sales gimmicks to get you pay more for what you don’t need.

    • sade

      How many times you had too use ISO 1600 or more and still your pictures got blurry? How many times you could not use a tripod in museums?
      I always think why people have such a bad tech-phobia that ignore any sort of new technologies simply by saying that it is useless and is invented just for extra money they get.

  • joe schmoe

    them commie pinko liberals trying to take away our in-lens stabil-i-zation. over my dead body, hippie booklearners. {*sound of shotgun cocking*}

  • samp

    as a film shooter i am certainly glad nikon and canon both do in-lens stabilization! especially with iso 50 velvia…

  • tavi

    Stabilization is not a gimmick. But reasoning that is better just cause it will shift the image in the vf is just marketing. And how is an extra element (that could be missing in the first place) gonna improve the image quality? Oh, I remember, it’s great, ’cause it’s a Nikon.
    Admit it: it’s a workaround that N & C developed during film years, and had to continue with it during digital with all the r&d invested.

  • Alan

    Just come across this topic days late, but one point I didn’t notice in scanning previous posts is that Canon and then Nikon gave us stabilization in SLR lenses in the film era, effectively committing themselves to their current solution. In-camera sensor stabilization has only been an option in digital cameras.

    Significantly, I have never seen any evidence to show that one system generally provides better pictures than the other.

  • Carsten

    Hi, anybody here can help me in my work?

    I need to make my Nikon able to correctly give me feedback that image actually is sharp, when green is reporting ok in camera to be able to shoot.

    As of now, the camera report I can shoot but the image is blur eventhough tripod. This only happens when I go closest possible to the edge of where I cannot go any closer. The camera reports back wrongly, that I can shoot, when in fact I am too close, something the camera never tells me.

    I need to be as close as possible, to get the image as large as possible for my work.

    Any help will be very much appreciated, thanks.

  • Back to top