Image Stabilization vs. Image Quality . . Your Thoughts?

[xpatUSA]

On another site, someone's review of a compact camera is being blasted as we speak.

The reviewer commented on the absence of image stabilization. That comment was challenged by a poster who wrote:

This shows his [lack of] understanding of what "this camera" is. It has no image stabilization for a reason. Better image quality is possible from a lens with no image stabilization. That's why Nikon has never put VR in their 85mm f1.4 and Canon never put IS in their 135mm f2 L.

(my emphasis). Of course, he doesn't say what sort of quality is meant.

So, in theory, if have two identical lenses apart from IS, do I get rid the stabilized one?

I've just pretty much done the opposite, by the way, something to do with my shaky hands. But I digress.

Is better image quality really possible from a lens with no image stabilization?

If so, what is that quality and why is it "better"?

[tclune]

First, of course, you can always turn stabilization off. So it isn't the lens that is better or worse, but the stabilization. Second, there is a well-established artifact that is attributed to stabilization, especially at long focal lengths: you get a weird nervous bokeh sometimes. At its worst, you will see doublets of branches or other sharply defined thin objects that are out of the focal plane, but near enough to still be reasonably well-defined in the image. The reason for this is that the stabilization gimbals around the focal point, jumping to a new position every couple thousandths of a second. If you shake the wrong way at the wrong time in just the right image, you end up with the weirdness described above in the foreground and/or background. I have only observed it on occasion at long focal lengths (300 mm and above), but I don't know any reason why it would have to be limited to those images. FWIW

ETA: The attached link shows the effect about as completely as any photo I have on hand. But I've seen (and shot) even worse examples. Just look at the photo full-size and the background will give you vertigo: http://www.ipernity.com/doc/1295788/38341598/sizes/o

[Manfred M]

Ted - in theory an unstabilized lens should be sharper than a stabilized one. But in theory that means you would be shooting with a heavy duty and weighted tripod. Stabilized lenses have a floating optical element that rotate independently around the x and y axes, assuming the z-axis runs through the centre of the lens, perpendicular to sensor plane.

Accelerometers measure camera motion and use magnets (almost like a linear motor) to deflect the floating element to compensate for lens motion. There will almost certainly be some level of error (I suspect that both timing issues (lag), over / under damping and hysteresis would likely be the main sources of positioning error) in this corrective action, but it is certainly more than good enough to correct for camera deflection.

Of course, the stabilization mechanism can be disabled and the floating element left in it's "neutral" position. The question is whether or not the floating optical can ever be positioned as accurately as one that is mounted permanently. Intuitively, one would have to say no, but I have no way of confirming that.

Both Canon and Nikon use in-lens stabilization on their pro f/2.8 70-200mm lenses. For them to do this suggests to me that they feel that the benefits outweigh the liabilities of the in-lens stabilization technology. I have the Nikon version and have tested it with a weighted heavy-duty tripod against my f/2 105mm lens - the prime lens seems to be a touch sharper, but frankly that could be related to it not having the design tradeoff required for a variable focal length lens. Stabilization is going to have a create benefit on longer focal lengths, so fast short to medium telephoto lenses in the hands of an experienced shooter are unlikely to gain that much from stabilization technology (i.e. no need for it on an 85mm ir 135mm lens).

My opinion is that you could theoretically get a sharper image with an un-stabilized lens, but you had better stick to shooting off a very heavy duty tripod and head in order to achieve this quality. If you are going to hand-hold, stick to the stabilized lenses.

[pnodrog]

On another site, someone's review of a compact camera is being blasted as we speak.

The reviewer commented on a lack of image stabilization. That comment was challenged by a poster who wrote:



(my emphasis). Of course, he doesn't say what sort of quality is meant.

So, in theory, if have two identical lenses apart from IS, do I get rid the stabilized one?

I've just pretty much done the opposite, by the way, something to do with my shaky hands. But I digress.

Is better image quality really possible from a lens with no image stabilization?

If so, what is that quality and why is it "better"?

No you do not digress - A lens with image stabilization active in shaky hands is going to produce far superior results to a none stabilized lens. However on a 10 ton tripod and a windless day the lens without image stabilization will be vastly superior if you can see the difference.......

We worry far too much about the equipment and far too little about the skill.

[Manfred M]

Second, there is a well-established artifact that is attributed to stabilization, especially at long focal lengths: you get a weird nervous bokeh sometimes. At its worst, you will see doublets of branches or other sharply defined thin objects that are out of the focal plane, but near enough to still be reasonably well-defined in the image. The reason for this is that the stabilization gimbals around the focal point, jumping to a new position every couple thousandths of a second.

The phenomenon you are seeing is called "jitter" and is pretty typical in any control system that is constantly trying to make a correction and overshooting a bit, correcting for that, etc. If the system were perfectly damped, you'd likely not see this at all.

If you shooting on a tripod with IS / VR turned on, I would expect to see more of this as the lens tries to correct for shake that isn't there. I'd expect that this would be more prevalent on older lenses as IS / VR seems to have become more sophisticated with each new lens release. The earliest stabilized lenses might only gain you a stop or two, but the latest ones are claiming gains of four stops or more.

[chauncey]

FWIW, I've never have seen one of my images that suffered from using/not using IR when on a tripod.
Hand held...never not used IR even if a cranked-up SS. Additionally, in my personal experience,
I have never seen a zoom lens that, IQ wise, could equal a prime lens.

[xpatUSA]

So, if I understand folks correctly so far, OIS/OS/IS/VR/IR lenses can have unstable position control loops and the particular image quality in question is blur in the xy direction of the camera shake?

William, pardon my ignorance, what is a "cranked up SS", please? Interesting too, your comment on zoom v. prime, BTW, made just as I was thinking about getting the well-regarded Sigma 18-35mm Art model . . .

[Manfred M]

So, if I understand folks correctly so far, OIS/OS/IS/VR/IR lenses can have unstable position control loops and the particular image quality in question is blur in the xy direction of the camera shake?

Potentially yes, but the bulk of the camera shake will have been eliminated as a consequence of using OIS/OS/IS/VR/IR. The tradeoff is a good image but there appears to be a risk of artifacts based on Tom's experience. I have not experienced this myself.

William, pardon my ignorance, what is a "cranked up SS", please?

SS = Shutter speed.

Interesting too, your comment on zoom v. prime, BTW, made just as I was thinking about getting the well-regarded Sigma 18-35mm Art model . . .

In theory a fixed focal length lens means the lens designer has a single focal length that he / she has to optimize for, so in theory at least that could be true, but reality is a bit different. (Isn't it always)

Most fast zoom lenses tend to have a a smaller maximum aperture than the fast primes; often no faster than f/2.8 (the Sigma you are looking at is an exception at f/1.8). Faster lenses require more correction (again, I'm thinking of the f/1.4 and f/1.2 (or better)). So these fast primes tend to have performance characteristics that are around the same as the zooms, and the image quality wide open can be worse performers than the fast zooms. After all, why would I want a fast lens other than to shoot it wide open (or nearly wide open); because I want to shoot shallow DoF, lower ISO required or faster shutter speeds. Otherwise, I'd go for a slower, less expensive lens.

So William would likely be right if you were comparing two lenses that had roughly the same maximum aperture, but as you are not, his assumption is not quite correct.

I have heard that the Sigma Art primes are excellent; I haven't paid any attention to the zooms.

[John 2]

.............................We worry far too much about the equipment and far too little about the skill.

Amen to that.

[Manfred M]

We worry far too much about the equipment and far too little about the skill.

Yes, that can be true. I also am someone who likes having a thorough understanding as to how my equipment works so that I can get the optimal results out of it.

[pnodrog]

Yes, that can be true. I also am someone who likes having a thorough understanding as to how my equipment works so that I can get the optimal results out of it.

Yes I would certainly not go as far as ignorance is bliss. However often I see long discussions on the finer points of technology when in practice they have no observable influence on the result except in extreme circumstances.

[Manfred M]

Yes I would certainly not go as far as ignorance is bliss. However often I see long discussions on the finer points of technology when in practice they have no observable influence on the result except in extreme circumstances.

A hazard of the job. A number of the members (I have to include myself here) have technical backgrounds and our interests in some of the minutia as to how things are put together or work will influence the equipment we buy. As an example, I stayed away from both Sony and Olympus cameras because of how they implemented image stabilization.

[xpatUSA]

Generally my shots are so boring and poorly executed that the only real pleasure I get is discussing and knowing exactly how everything works - down to the nearest photon, nm, nV, mcd, lux, um^2, etc., ya get the idea . . .

[Manfred M]

Generally my shots are so boring and poorly executed that the only real pleasure I get is discussing and knowing exactly how everything works - down to the nearest photon, nm, nV, mcd, lux, um^2, etc., ya get the idea . . .

You'd get along well with one of my friends (a PhD physicist). Another photographer and I were discussing light output falloff from an electronic flash and Duncan jumped right in trying to figure out how the photons would be emitted as the electrons moved to another orbital to another in the xenon atom over time...

[chauncey]

Personally, I get a great deal of satisfaction knowing that it's not gear related
when an image is screwed-up.

[Donald]

Personally, I get a great deal of satisfaction knowing that it's not gear related
when an image is screwed-up.

Unfortunately that's true. When it is screwed up it's invariably us that have caused it, not the gear.

[Manfred M]

Unfortunately that's true. When it is screwed up it's invariably us that have caused it, not the gear.

Donald - that is not necessarily true. I had some significant issues doing something photographically around 18 months ago and just couldn't get something to work, so I turned to the experts here at CiC for help.

As things turned out, the issue I was having was 100% a result of the technology not working the way I thought it did. By having had this explained to me and by me understanding it, I was able to find a workaround.

A modern digital camera is an amazing tool, as long as one works within its design parameters. Work at the edges and things can go wrong in the most unexpected ways.

[rpcrowe]

I have never experienced the anomalies described above and I shoot hand held from 17mm to 55mm (on my 17-55mm f/2.8 IS lens) to 300mm (on my 300mm f/4L IS lens) with IS turned on. I also shoot from 70-200mm on my 70-200mm f/4L IS lens) also with the IS turned on.

IMO, Image Stabilization is more and more important as I grow older resulting in less steady holds.

If you can hand hold a long lens without image stabilization and do a better job than with image stabilzation, more power to you! I just know that Image Stabilization is an important tool in my photographic toolbox.

[xpatUSA]

You'd get along well with one of my friends (a PhD physicist). Another photographer and I were discussing light output falloff from an electronic flash and Duncan jumped right in trying to figure out how the photons would be emitted as the electrons moved to another orbital to another in the xenon atom over time...

Funny you should mention that - at one time I was the world's foremost expert on luminous watches:

http://kronometric.org/article/lume/#3.2

I remember struggling with that very concept of photons being emitted as electrons lost energy and went back to a lower band . .

[pnodrog]

Funny you should mention that - at one time I was the world's foremost expert on luminous watches:

http://kronometric.org/article/lume/#3.2

I remember struggling with that very concept of photons being emitted as electrons lost energy and went back to a lower band . .

It actually helps to be as thick as Planck.....