Image Stabilization

[Deucalion]

I've been reading a lot... never thought there was a LOT to actually read about in digital photography... wow ...I've been wondering about the differences in the lenses of canon, from the EF-S and the "L" series specifically about image stabilization or in Nikon lenses the VR.

it seems that most EF-S lenses have IS (what most people seem to call "cheapies" are all EF-S lenses "cheap"? where I come from the EF-S lenses aren't cheap the "L" range is actually out of reach of most people except for those with deep pockets) while most "L" lenses excluding the high end telephoto lenses don't have IS...

but "L" lenses are supposed to be better than the EF-S counterpart... is the absence of IS in the "L" series means that it doesn't suffer from the occasional camera shake when you take a picture or does it mean that Canon intended the "L" series of lenses be mounted on a sturdy tripod/base when taking a photo?

is IS really necessary? and if so why does the low end of the "L" series not have IS?

[ktuli]

Reginald,

Wow - you've got a few different questions there and they all seem to get mashed up, so I'm going to try and pull them apart and answer them separately.

First, when dealing with Canon lenses, there are EF and EF-S. This refers to the mount used. Canon cameras with an APS-C sensor (1.6x crop) can all use both EF-S and EF lenses. However, full frame Canon cameras can only use EF lenses. The reason for this is because the EF-S lenses actually protrude further into the camera when mounted. EF vs EF-S does not necessarily distinguish whether they are "cheapie" or "pro" lenses. Sure, there are plenty of EF-S lenses that are "kit" lenses and are often considered "cheapie" and are lower quality. But as the EF-S lineup grows, I believe Canon is starting to take a more serious look at them as being used by serious (though maybe not pro) photographers and producing better quality lenses for the EF-S lineup.

As for the L series of lenses, though (I believe) right now they are limited to the EF lineup of lenses, they are simply Canon's higher grade of lenses. I honestly think that within the next three years, we'll see some EF-S lenses created for the L series.

Moving on to IS, that is completely separate from any of the previous letters discussed. IS (Image Stabilization) is Canon's distinction for it... Nikon uses VR (Vibrarion Reduction), Sigma uses OS (Optical Stabilization), Tamron uses VC (Vibration Compensation). They all do the same thing. But there are IS lenses in every series - including the L lenses. I own a Canon EF 100mm f/2.8L IS USM Macro lens. There's also a Canon EF-S 18-200 f/3.5-5.6 IS lens. They're worlds apart in terms of quality, but both have IS systems included. Seeing a lens without IS at this point in time, almost just means it is an older lens as manufacturers are putting the technology into almost everything these days. Sure, there are exceptions to that, but whether a lens has IS is not decided by which lineup it is in. For the "low end" of the L series not having IS, it is basically just because these lenses were designed/manufactured before the IS technology became as prolific as it is now.

Whether IS is really necessary is up to you and your style of shooting. If you tend towards hand-held, there are a lot of people who will argue that IS is absolutely necessary. Others will argue that IS can actually introduce problems to your image (especially when tripod mounted). But the bottom line is that the pros/cons of IS vary greatly based on the specific lens, its IS implementation, and that lens' usage.

Hope this helps.

- Bill

[Geoff F]

Just a couple of extra points to add to Bill's comprehensive reply. Stabilised lenses have an off switch so you might as well take advantage of the option and simply use that switch when required.

Most lens manufacturers seem to be gradually updating their lens line up and in most cases the newer models are stabilised as a standard fixture.

[ktuli]

Just a couple of extra points to add to Bill's comprehensive reply. Stabilised lenses have an off switch so you might as well take advantage of the option and simply use that switch when required.

Just another point to add to Geoff's additional point to my apparently less-than-comprehensive reply... Did you follow all of that?

Anyway, some IS lenses have multiple modes to their switches. The modes are usually, off / on (mode 1) / on (mode 2).

Mode 1 would be used for full IS mode, and attempts to correct camera movement in both the horizontal and vertical axises (someone correct me if I'm wrong, but I don't think IS usually does anything in the Z axis - ie: camera movement toward/away from the subject).

Mode 2 would be used for IS only in the vertical plane. This mode is used when you want the benefits of IS, but also will be moving the camera intentionally - say when panning with cars at a race track, or tracking birds in flight.

Off is used obviously when you do not want IS at all. There is a lot of discussion based on whether IS should be used when the camera is mounted to the tripod. And again, this is often times lens specific. But the argument is that while on a tripod, the IS sensors interpret certain micro-vibrations as motion, and attempt to correct for them, but in the process actually cause further problems. Some lenses are actually reported to over working the IS system in these scenarios and producing poorer image quality.

You would have to refer to the documentation for the specific lens you are purchasing/using.

- Bill

[koolkat]

Hi, I agree with Bill and Geoff and their reasons are solid and good reading. I tend to shoot a high percent of time with a monopod and as such turn the IS off of the Tamron 18x270 lens. I think I get faster autofocus with the IS off and with the monopod still good solid shots. A lot goes into a decision whether IS is necessary or warrented. I would think that most macro shooting would be done from a tripod and as such, IS might not be needed. I've shot the 300mm F4 Canon lens with IS turned off, using a monopod with good results. So, the question remains, what type of shooting will you do and is the extra cost worth what you want to do.

[LeoLeo]

Hi Reginald,

The "S" from EF-S comes from shortfocus, and is ment to be used on APSC camera. One year ago I've asked if I really need IS or not, for the 70-200L F4 USM (IS). The answer was a big YES from Colin and others, and it's very true.

As the others have asked you, it is most of the time related to the type of photography you made.
As Bill mentioned, IS/VR/VC/OS or whatever stabilization is used can help you with X axis (horizontal) and Y axis (vertical) but not on Z axis (toward/away to you).

Let me give you one example about (images are direct from the camera, imported and saved as jpeg)


Canon 40D, EF 70-300, 160mm, 1/320, F7.1 ISO 100


Canon 40D, EF 70-300, 70mm, 1/400, F7.1 ISO 100



and a crop from the picture above (100%)
Is the second picture motion blured ? I don't think so, looks more like out of focus, my son was moving towards me (actually, to a dog behind me ... poor dog..)

So, my point is : it's all about what sort of photography you make. For me ... maybe not even 70-200 L F2.8 IS II USM will help alot... some glue may help ..to keep my model steady

For me, the true advantages of L series and some of EF-S are FTM Autofocus and distance scale, except technical /optical capabilities.

I hope it helps you, not confusing you

Leo

[pipflash51]

Can anyone explain to me what photographers did before IS? How did they keep their pictures sharp then? Can the price of IS lenses be justified as being nearly double the price of non IS lenses?
Also would an f4 lens with IS be more useful than a f2.8 with no IS?
Lots of questions I know but please forgive me as it's my first post!!

[Donald]

Can anyone explain to me what photographers did before IS?

Resolved that they couldn't shoot at such slow shutter speeds when hand-holding. Used faster film or higher ISO speeds to get shutter speeds up.

How did they keep their pictures sharp then?

Shot a sufficiently high shutter speed that would ensure no motion blur

Can the price of IS lenses be justified as being nearly double the price of non IS lenses?

Putting IS in a lens is pretty complex piece of engineering. The fact that people buy them suggest that teh marker considers the price differential to be justified and fair

Also would an f4 lens with IS be more useful than a f2.8 with no IS?

That depends. Depends on what you want to do with it. For example, I have a Canon 70-200 f4 L IS. I don't need f.2.8 at that focal length. The S gives me an option of hand-holding at shutter speeds not possible with a non-IS lens.

Lots of questions I know but please forgive me as it's my first post!!

No problem. That's one of the best ways of learning.

[ktuli]

That depends. Depends on what you want to do with it. For example, I have a Canon 70-200 f4 L IS. I don't need f.2.8 at that focal length. The S gives me an option of hand-holding at shutter speeds not possible with a non-IS lens.

Or - if you turn it around, the IS gives you the option of hand-holding the same shutter speed, but several f/stops higher than a non-IS lens. So if you wanted to use f/8, the IS might make that possible for you to hand hold it, whereas without the IS, maybe the best f/stop you could manage would be f/5.6 without ending up with a shutterspeed too slow to handhold.

So in that respect, the widest aperture of the lens is not necessarily poignant to the comparison. You're not comparing apples to apples. Yes, there might be a case where an f/4 lens with IS (at maximum f/stop) would allow you to take an exposure without camera shake, and an f/2.8 lens without IS (at maximum f/stop) would also let you get an exposure without camera shake in the same lighting conditions - however, the two resultant photos would not be identical because of the different depths of field you would receive from the two.... thus making the comparison not really valid in my opinion.

- Bill

[thequacksoflife]

one thing to bear in mind, I've read that IS/VR doesn't work over 1/500 and indeed at these speeds it's better to turn it off.

is it a good thing? well if you are shooting in very poor light where you can't shoot with flash or use a tripod (some cathedrals, National Trust properties) then yes its handy. I rather have it than not have it, I can switch it off then!

[rpcrowe]

Bill, Mike, Donald and others have provide some vey comprehensive and valuable information. Here are a couple of additional remarks...

1. The older/oldest of the L lenses did not originally have IS capability. Among these. but not limited to, are the original 300mm f/4L, 70-200mm f/2.8L; 70-200mm f/4L; 24-70mm f/2.8L and the 400mm f/5.6L. Each of these lenses is of excellent quality but, was introduced before IS was a common factor in lenses. I used a 24-70mm f/2.8L and did not miss IS because of the relatively short focal length. However I felt constrained by the lack of IS capability in my 70-200mm f/4L lens because I like to use this lens for general hand held photography, often in low light situations. A tripod or a monopod "could" be used but, it would slow down my shooting and make my rig not quite as versatile. I really wish that my 400mm f/5.6L lens had IS capability. That would really help in hand holding and when using the lens on a monopod. However, I mostly use this lens on a tripod (occasionally on a monopod) and the lens has excellent image quality wide open so, shoting at f/5.6 using ISO 400 can provide some very high shutter speeds given I am in decently bright conditions. I just attended a seminar by a local bird photographer who hand holds an AF-S NIKKOR 200-400mm f/4G ED VR II lens on a D300 camera and uses a 1.5x teleconverter. His results were outstanding. I would definitely need at least a monopod to get equivalent results from my non-IS 400mm f/5.6L lens. I "could" use my Image Stablized 300mm f/4L IS lens with a 1.4x TC but the auto-focus speed of that combination (although reasonably fast) falls short of the bare 400mm f/5.6L AF speed.

2. There are several versions of Canon's IS. I don't think that they have identfied the versions with a numeric designator , that range from the original IS which is found on the 28-135mm f/3.5-5.6 IS lens to the most recent IS which can be found on the 70-200mm f/2.8L IS ii lens.
http://usa.canon.com/cusa/consumer/s...s_Advantage_IS
Each version has increased capabilities in the number of f/stops advantage, different modes for panning and still shots and the ability to identify when the camera/lens is tripod mounted. Each version is better but, even the earliest version helps.

3. It has often been claimed that shorter focal length lenses do not need Image Stabilization. I once thought that the IS capability of the 17-55mm f/2.8 IS lens was just icing on the cake. However, since using it, I have fallen in love with that capability. The combination of consant f/2.8 aperture and IS makes my 17-55mm f/2.8 IS lens a very viable low light glass.

4. Some lenses with IS (or other manufacturer's satbilization capabilities) did not provide as good image quality as their non-shake dampened counterparts. This is said to be true of the first model 70-200mm f/2.8L IS lens which was said to be inferior in image quality to the non-IS version. Tamron's 17-50mm f/2.8 VC is said to not produce as good an image quality as the non-VC version. However, when Canon introduced the 70-200mm f/4L IS lens, that tendency was changed. The IS version actually produces better IQ than the previous non-IS version. The IS version of the Canon 18-55mm kit lens also produces better IQ than the previous non-IS version. The 70-200mm f/4L IS lens costs almost double that of the non-IS version while the price of the 18-55mm IS kit lens increased only slightly over the previous non-IS version.

5. I personally think that there would be a market niche for a 30mm or 50mm f/1.4 IS lens for some serious low light imagery.

I LOVE IMAGE STABILIZATION! Sure, photographers shot with non stabilized lenses and got good imagery. Photographers once shot using wet plate technology - no one would advise going back to that way of shooting now! Many photo books from the 1950's and 1960's recommended 135mm as the longest focal length that should be hand held. Now, because of IS technology, we talk about hand holding a 400mm lens on a 1.6x camera which equates to a 640mm lens on a 35mm camera.

My bird shooting acquaintance hand holds his 200-400mm f/4 lens with a 1.5x TC. That is an equivalent 900mm capability on his D300 Nikon. Canon is introducing a 200-400mm f/4L IS 1.4x TC lens. This lens will have a 896mm equivalent focal length on a 1.6x crop camera. I would not be surprised if, due to the IS capability of that lens, photographers will be able to hand hold it with no more support than a Bush-Hawk mount. That type of capability was absolutely unheard of in the era prior to Image Stabilization...

[Graystar]

one thing to bear in mind, I've read that IS/VR doesn't work over 1/500 and indeed at these speeds it's better to turn it off.

That’s a myth. VR works just fine above 1/500s. I’ve yet to see any explanation as to why the VR system would be able to keep an image aligned to the pixel at 1/120s or 1/250s, but yet lose it with 1/1000s...especially when there’s practically no difference at all in the shutter actuation. I've never seen a problem with my lenses at those speeds.

The only difference occurs at shutters speeds above 1/450s, where the second curtain starts to close before the first curtain finishes opening. No one has yet to explain if or how that would affect VR. Otherwise, the operation of the first curtain is the same for all shutter speeds, and even at 1/8000s the camera requires around 1/425s to complete the shutter actuation due to the way focal plane shutters operate. The only difference is the timing of the second curtain.

[thequacksoflife]

That’s a myth. VR works just fine above 1/500s. I’ve yet to see any explanation as to why the VR system would be able to keep an image aligned to the pixel at 1/120s or 1/250s, but yet lose it with 1/1000s...especially when there’s practically no difference at all in the shutter actuation. I've never seen a problem with my lenses at those speeds.

The only difference occurs at shutters speeds above 1/450s, where the second curtain starts to close before the first curtain finishes opening. No one has yet to explain if or how that would affect VR. Otherwise, the operation of the first curtain is the same for all shutter speeds, and even at 1/8000s the camera requires around 1/425s to complete the shutter actuation due to the way focal plane shutters operate. The only difference is the timing of the second curtain.

well my primary source was http://bythom.com/nikon-vr.htm

i've read it elsewhere and seen some test on one of the various forum of course it may be wrong.....

[Graystar]

well my primary source was http://bythom.com/nikon-vr.htm

i've read it elsewhere and seen some test on one of the various forum of course it may be wrong.....

Yeah, I aleady had it out with Thom Hogan over that article. Supposedly the first version was written based on information from Nikon engineers. Then it had to be corrected based on information from Nikon engineers. Hmmm...

And his description of VR is still wrong. His Nyquist explanation is completely debunked with a just a few moments thought. As I said, a 1/8000s exposure still requires at least 1/425s to complete. At 1000Hz, the VR will perform at least 2 corrections during that exposure. At a shutter of 1/500s the VR will complete 4 corrections. So how is it that the VR system fails across 2 corrections when it has proven that it can keep an image in place on the sensor across 4 corrections?

Regardless of shutter speed, the VR is correcting for human motion, and that doesn't change with shutter speed. Also, there is no Nyquist sampling and no predicting as Thom Hogan described. This is a serial system...detect, correct, detect, correct, and so forth.

Nikon has written quite a bit about their VR system on their websites. Never have they mentioned that the system uses the data collected to make predictions. This is all Nikon says of the detection process...

"These sensors detect camera movement as angular velocity every 1/1000 second. Angular velocity data is sent to a microcomputer built right into lens, which then calculates the amount of compensation needed. After this data is transmitted to the VR unit, the adjustments are made instantaneously."

http://www.nikon.com/about/technolog...vr_e/index.htm

In another forum someone was wondering if there was something wrong with his new Nikon 70-200. He posted shots of kids playing baseball, and they were full of double images and looked pretty terrible. He suspected the VR system. His next post was full of clear, sharp images from his 70-200. What did he do? He removed the UV filter he had installed to protect the front element (and as I recall it was a good one.)

[thequacksoflife]

In another forum someone was wondering if there was something wrong with his new Nikon 70-200. He posted shots of kids playing baseball, and they were full of double images and looked pretty terrible. He suspected the VR system. His next post was full of clear, sharp images from his 70-200. What did he do? He removed the UV filter he had installed to protect the front element (and as I recall it was a good one.)

I'm puzzled by filters..... buy expensive glass and then put something in front of it.

thanks for that the VR is interesting. I did check out what Nikon said and there was nowt.

[Donald]

I'm puzzled by filters..... buy expensive glass and then put something in front of it.

Just ask yourself which one you'd rather get scratched.

[Graystar]

I'm puzzled by filters..... buy expensive glass and then put something in front of it.

Ditto. A hood offers good protection and is much better for IQ.

[Donald]

Ditto. A hood offers good protection and is much better for IQ.

Sorry, but I believe that to be a myth. You cite one example above. But where is the evidence that demonstrates that in real world shooting and viewing situations, that a good filter causes discernible image degradation?

[Steve S]

Sorry, but I believe that to be a myth. You cite one example above. But where is the evidence that demonstrates that in real world shooting and viewing situations, that a good filter causes discernible image degradation?

I really don't understand, why people would pay a premium for a high quality lens, and place a $50 piece of plastic over it. It's really comon sense.......................if you were shooting out of a window???? Would you place the lens right against it???? Or simply open the window before you take the shot??? Covering your lens is like shooting through a window, all the time. Personally i would open the window

[Tringa]

Some camera manufacturers put the IS in the camera body which means the lenses can be simpler and all become image stabilised.

Dave