[flippedgazelle]

Interesting. Now that you mention it, the last time I seriously agonized about switching to Nikon (middle of last year, I think), I did look at the prices of some lenses and was surprised that some of the VR lenses cost less than I had expected them to. Perhaps, by selling lots of them, they've found a way to bring the costs down, which would make sense.
l

That's how I felt a couple years ago when lenses for the Nikon D40 - which required in-lens focus motors - dropped in price to match the lenses which didn't contain focus motors.

One thing I'm *really* curious about is which IS technology will prove more durable. I think most of us would be more likely to hang onto good glass longer than a camera body, and if a bunch of these IS-lenses start crapping out in 5 or 10 years...

[WMBP]

One thing I'm *really* curious about is which IS technology will prove more durable. I think most of us would be more likely to hang onto good glass longer than a camera body, and if a bunch of these IS-lenses start crapping out in 5 or 10 years...

That's the most interesting observation I've heard on the subject. Hadn't ever thought about the possibility that the IS mechanisms in the lenses would, what? wear out? break? Guess I don't know anything about how in-the-lens IS works. I know only a wee bit about how in-body works. But I see your point. If something breaks in the body, well, you think of the good times you had with it, put it on the shelf, and use the tragedy as an excuse to buy something new. But I'm hoping to hang on to my lenses for a long time. And given the relatively simplicity of the lenses I have, I think that expectation is reasonable.

Will

[sterretje]

I will happily drop SR if it reduces the price of a camera, no questions asked. I used to be able to shoot 200mm at 1/30sec in the film days with a bit of support (wall, streetlight,...) and with a bit of practice I might get that back again.

I don't know what the prices were of the K100D and the K110D when the models were released; 10% difference possibly?

[geezer52]

One thing I'm *really* curious about is which IS technology will prove more durable. I think most of us would be more likely to hang onto good glass longer than a camera body, and if a bunch of these IS-lenses start crapping out in 5 or 10 years...

I've heard of one of the potential issues w/ in-lens IS. To maintain image stability in-lens requires micro shifting of at least one of the lens elements. This means there's an inherent lack of a fixed centering. Small motors moving large glass over time sounds like a lot of wear and tear. And once the lens starts to fail or slip I don't think you could reliably use that lens at all without repair.

Bit of a philisophical argument. If the in-lens IS on one lens fails, it doesn't impact the other lenses but you have to buy lenses with the extra gizmos every time to have IS. In-body IS is a one time purchase (arguably less precise) but if it goes you have no more IS (but the camera is still probably usable.)

[jct us101]

I've heard of one of the potential issues w/ in-lens IS. To maintain image stability in-lens requires micro shifting of at least one of the lens elements. This means there's an inherent lack of a fixed centering. Small motors moving large glass over time sounds like a lot of wear and tear. And once the lens starts to fail or slip I don't think you could reliably use that lens at all without repair.

Bit of a philisophical argument. If the in-lens IS on one lens fails, it doesn't impact the other lenses but you have to buy lenses with the extra gizmos every time to have IS. In-body IS is a one time purchase (arguably less precise) but if it goes you have no more IS (but the camera is still probably usable.)

Couldn't the lens based IS be compared to the power zooms that were mildly popular in the late 80s and 90s? Those lenses (if they haven't already) are probably going to stop losing their power zooming ability fairly soon and become totally useless until you fix them.

[geezer52]

Couldn't the lens based IS be compared to the power zooms that were mildly popular in the late 80s and 90s? Those lenses (if they haven't already) are probably going to stop losing their power zooming ability fairly soon and become totally useless until you fix them.

I'm thinking if in-lens IS lenses keep coming down in cost, the savings are coming from somewhere. They will, at least initially, want to appear sharp, so I assume the glass will still be reasonably good. Maybe they will come with an life span of xx years and that will just be expected and accepted.

[Rondec]

The reality is that it isn't particularly expensive for companies to make in lens IS, they just choose to charge significantly more for it as compared to lenses without it. I do know a couple of people who have had Canon (cheap) IS lenses lose the IS functionality and had to send them in to get them fixed. Not certain that it happens very often, but it definitely made their images blurry and unusable.

I don't see taking SR out of the body of the camera changing its size very much at all.

[falconeye]

I compare the new Bigma (with IS, left) vs. the old one (cf. images).
<->
The new Bigma is not only heavier (1970g vs. 1842g), has a larger filter diameter (95mm vs. 86mm), is slower at the wide end (f/4.5 vs. f/4.0) and is more expensive (~20%+). It has more lenses too (20 vs. 18 if I count correctly) and has not necessarily what I would call a superior MTF.

So, optical IS seems to come at a price.

My gut feeling is that sensor shift IS could be a lot better if more value would go into it. However, it is probably difficult to market as nobody ever compares prices based on, say a stabilized 70-200/2.8 equivalent combo. So, sensor shift IS is cheaper than it must be and really complex IS system can only be sold embedded into expensive lenses.

I really hope Pentax makes a body with sensor shift IS which sends the optical IS vendors back to their drawing boards. Feasible. But is it likely?

[Lowell Goudge]

with respect to sensor shift and cost/size impact on camera bodies I have one observation to make.

I suspect all camera makers already have a form of sensor shift already inplemented, specifically related to dust removal.

Once you have done this, although the movement and frequency may be different, I doubt there is much change in overall volume of the mechanical assembly to add sensor shift for image stability. It is just something that other makers have chosen NOT to do.

For me, and I think I have demonstrated its use in my first post with this thread, sensor shift has its uses. There is no way, even if I can be perhaps proud of my technique of hand holding, that a shot at 1/40 with a 500mm lens is possible without SR. and the shot I posted was not the only one, just the best angle and composition of many clear shots I took in a sequence at 1/40th .

Long live sensor based SR

[Lowell Goudge]

That's the most interesting observation I've heard on the subject. Hadn't ever thought about the possibility that the IS mechanisms in the lenses would, what? wear out? break? Guess I don't know anything about how in-the-lens IS works. I know only a wee bit about how in-body works. But I see your point. If something breaks in the body, well, you think of the good times you had with it, put it on the shelf, and use the tragedy as an excuse to buy something new. But I'm hoping to hang on to my lenses for a long time. And given the relatively simplicity of the lenses I have, I think that expectation is reasonable.

Will

Will

I agree it is an interesting point.

I have a collection of bodies I no longer use, as I have replaced film with digital, but except for my vivitar 400F5.6 which I have not used other than to check exposure with my K10D, and my Xr rikenon 50mm F2 (I have an SMC50mmF1.4 to replace it) there is not a single lens in my collection (30 lenses and yes I suffer form LBA) that I have not used in the last 3 months. Some of these lenses I bought almost 30 years ago. As a result, Good Glass far outlives camera bodies. Therefore the less that can go wrong with a lens the better, since you will always buy new cameras.

[falconeye]

Once you have done this, although the movement and frequency may be different, I doubt there is much change in overall volume of the mechanical assembly to add sensor shift for image stability.

I doubt this to be the case.
Dust removal is best done with ultrasonic shift of the AA filter, not low frequency shift of the entire sensor.

The point I wanted to make (and I obviously failed): lens-based IS is already at the limit of feasibility, both in budget and performance. Sensor shift IS could outperform optical IS by a large margin if a similiar effort would be put into it.

Because the body has only one instantiation of the IS mechanism, you can pay for the best gyro sensors to start with. And the shift mechanism can be made almost arbitrarily precise and fast. After all, moved masses are smaller than with a lens element and again, only one instantiation can mean better precision and responsiveness.

[Lowell Goudge]

I doubt this to be the case.
Dust removal is best done with ultrasonic shift of the AA filter, not low frequency shift of the entire sensor.

The point I wanted to make (and I obviously failed): lens-based IS is already at the limit of feasibility, both in budget and performance. Sensor shift IS could outperform optical IS by a large margin if a similiar effort would be put into it.

Because the body has only one instantiation of the IS mechanism, you can pay for the best gyro sensors to start with. And the shift mechanism can be made almost arbitrarily precise and fast. After all, moved masses are smaller than with a lens element and again, only one instantiation can mean better precision and responsiveness.

I did consider this, and while there may be some differences in frequency etc, I still think that the size in camera is not a limiting issue. that was my point.

As for putting the best into the body I think there are a couple of issues here, and it is not gyro sensors but accelerometers, however that is splitting hairs. Generally I agree we still have room to move (no pun intended)

The first issue is that like focus adjust, it would be great to have programmable stability adjust. Let's face it, not everyone has the same stability, so the present setup is a one size fit's all splution.

Second not all lenses have the same shake around the same optical center. There is no way that the compensation for an SMC/Super Tak 500F4.5 is the same as for my 300F4 plus 1.7x AF TC, is the same as a compact 500mm Mirror.

I would be willing to bet that the results using the 3 options I give above would be completely different. As a result, you would really need 2 sets of sensors, so that rotation or movement of the entire optical system could be calculated, along with the optical center of the lens, which is different in each case above, although you only have a single focal length to inout at present.

Perhaps there could be additional data encoded into an FA mount lens that would be of use, as at least the camera would know what lens and SR could be tuned precisely to what pentax knows about those lenses. That would at least give pentax lens users a one up over aftermarket lenses.

[WMBP]

The point I wanted to make (and I obviously failed): lens-based IS is already at the limit of feasibility, both in budget and performance. Sensor shift IS could outperform optical IS by a large margin if a similiar effort would be put into it.

This is exactly what I was wondering. If it's true, then I think the makers who are using in-body image stabilization (including Pentax and Sony) should put their brains together and improve in-body stabilization so that it's demonstrably as good or better than in-the-lens. If necessary, charge more for bodies and put the improved mechanism into the higher end bodies instead of all of them. I'd be willing to pay $50 more (at least) for a body that had even better SR than I've gotten already.

Right now, there's a sort of stand off between Canon/Nikon and everybody else. The Canon/Nikon users believe that their IS/VR technology is superior and apparently there is some technical evidence to support that; and this superiority, in their view, justifies the extra expense it adds to some of their lenses. On the other hand, everybody else believes that the technical superiority of in-lens IS/VR is modest at best, and so users of Pentax, Sony, etc. bodies are happy to have what they've got. Well, if you make in-body image stabilization demonstrably, provably the equal of in-lens, then the advantage that Canon and Nikon users are paying for evaporates, at least in this one area. If you make in-body image stabilization demonstrably BETTER—even if only slightly—then you're really got a game changer.

The writers of reviews are interested in this. The sources I read regularly both online and in print mention in-body image stabilization as an advantage of every camera that has it.

And ironically, I think in-body IS could help sell cameras to consumers as well as pros. I know, the low-end Nikon cameras now come with a kit lens that has VR. (Do the Canon's?) Still, that's just one lens. You get VR in an 18-55 lens; big deal. Where you really need it is in that 50-200 lens you buy next, and right now, I think a lot of consumers aren't paying for VR in their second and third lenses. Pros and serious enthusiasts, yes, but not ordinary consumers.

Something I hope Pentax is working on.

*

By the way, does anybody know how different the technologies are for image stabilization that are used by Sony, Pentax, et al.? I gather from the reviews that I've read that they ARE different. I just don't understand the differences.

I guess it's unlikely that these competitors would get together and work cooperatively on a single Canon-Nikon-killing image stabilization technology. But I wish they would. They should realize that it's to the advantage of all of them, for ANYBODY to give Canon and Nikon a challenge. Right now the challenge seems to be coming mainly from Sony but only at the high end (with its full-frame bodies).

Will

[falconeye]

I just came across a study which was first discussed in a thread here:
https://www.pentaxforums.com/forums/pentax-dslr-discussion/91524-shake-reduction-effective.html
The direct link is this:
Study of the Effectiveness of Shake Reduction in the Pentax K7
I think this is a wonderfully done work. It has my admiration. What it lacks is a study of different focal lengths though.

I still think that the size in camera is not a limiting issue. that was my point.

Point taken

As for putting the best into the body I think there are a couple of issues here, and it is not gyro sensors but accelerometers, however that is splitting hairs.

In the case of the Pentax SR system (the one I studied) it is gyro sensors, not accelerometers. The gyro sensors (there are two in the K20D) work by measuring the corilios force and are pretty ingenious pieces of electonics
In some of my earlier posts, I gave the exact manufacturer and part number.

There is no way that the compensation for an SMC/Super Tak 500F4.5 is the same as for my 300F4 plus 1.7x AF TC, is the same as a compact 500mm Mirror.

Wrong. This is the nice part about sensor-based stabilization. You only need to know the angular frequency around the optical axis and the two axis perpendicular to it and the focal length for an exact compensation of an image at infinite distance. Lens details play no rôle here.

At finite distances, you need extra accelerometers and Canon did actually file a patent explaining the obvious.

By the way, does anybody know how different the technologies are for image stabilization that are used by Sony, Pentax, et al.? I gather from the reviews that I've read that they ARE different.

Will, I don't know.

[SpecialK]

So is shake reduction/image-stabilization that necessary a feature anymore?

It is a marketing/sales feature and therefore will never go away.