You made me look. My XH-1 is set for stabilisation to be on only when the shutter release is depressed. However the manual also says that if an OIS lens is fitted the lens switch overrides the setting - this then is ambiguous as the lens only has ON/OFF. It is clear that OFF must turn it off, but whether the lens set to ON sets the camera to continuous (defined as stabilisation ON) or Shooting only (ON when shutter release pressed) is moot.

 

I would be intrigued to know how an in-body stabilisation system, based on sensor shift could stabilise the image in an optical viewfinder.

 

I strongly suspect that the degree of sensor movement required to dislodge dust is a small fraction of that required to stabilise the image from a telephoto lens. Just because there is room for dust removal doesn't necessarily mean there is room for a sensor shift stabilisation system. Additionally the control required for a sensor shift stabilisation system is considerably more complex than required to simply shake the sensor. Without considerably more information I can't say with any certainty whether there is space in a Nikon body or not. The suggestion that there may not be is pretty much speculation.

The suggestion that IBIS has economic advantages for the customer implies that there is an economic disadvantage to the manufacturer who, after all is in the business of selling lenses. Which is an argument in favour of the manufacturer not incorporating IBIS.

The compelling argument is probably economic, there really is no competitor making SLRs with IBIS that will take sales from Canon and Nikon if there were I imagine that they would respond quickly and decisively to make it available.What ever the obstacles.

 

That review must have involved a mirrorless system; I thought it was relevant to the discussion so far because it showed that at least one manufacturer assumes there are potential customers who don’t share your wish to have their viewfinder image continually stabilized.

Of course I agree with all of the above. But I think it likely that Nikon will have gone through a range of different body designs to modify from film to digital sensors, then to add space to fit a vibrating mechanism to the sensor, to take revised shutter designs, and also to differentiate between high-end and low-end models and manufacturing from different materials. I suspect designing a new model to have more space behind the mirror box for IBIS would be a minor issue for the company.

I don’t think that’s correct. If two manufacturers were to offer equally good cameras and lens systems, but it cost one manufacturer using lens-based stabilization more in total to make his range of products than the other using IBIS, the manufacturer using lens-based stabilization would have to make smaller profits to remain competitive. In practice it’s more complicated if a manufacturer has legacy products for lens-based stabilization and is considering introducing IBIS and omitting lens-based stabilization from all but telephotos. But on the other hand, being able to omit the stabilizing elements from a lens will permit better optical quality as well as lower costs for the lens.

Yes, I’m sure the compelling argument has been economic, but manufacturers don’t always make the best design decisions from the economic point of view. I’m sure both Nikon and Canon are kicking themselves for not introducing full frame mirrorless cameras sooner!

They do have a competitor, although not very powerful in recent years, in the shape of Pentax. But I’m also sure both Canon and Nikon are looking for opportunities to steal a march on each other, rather than always following the same route. I think Canon’s early switch to the all-electronic EF mount is a good example of this, irrespective of which company made the best decision.


Chris

 

I know Nikon had the One system and I suspect learned a lot about mirrorless from it but, yes I think both Nikon and Canon could have produced their full frame mirrorless cameras earlier, even if only by a year. Though the mess Nikon made of the grip for the X6/Z7 suggests that there was a degree of rush involved.

As I said on page 1, in this context Pentax are irrelevant because of their very small market share and presence.

Nikon and Canon seem to share a lot where the top of the range cameras are concerned, batteries can use the same charger with adaptor plates, the sensor resolutions are very similar etc. I don't see them competing quite as hard at the top of the market as was once the case, but I may be reading it wrong.

 

Whenever somebody asks a question like this, the prices and numbers of lenses available for the newer types of camera often are carefully ignored. On another thread a few weeks ago, somebody was wanting advice on buying a first 'real' interchangeable lens camera, but I think I was the only person to ask if the person might also want to buy more lenses later. If the answer is year, and the budget is not large (whatever that means today), I suggested a second hand Canon of Nikon DSLR because these will allow the best choice of affordable second hand lenses when required. I have no personal preference, being a Pentax DSLR owner, but was encouraging the person to consider the purchase of a camera body (and probably a kit lens too) as just the first part of a long term investment in which the costs of lenses should be investigated before deciding on a particular brand or type of interchangeable lens camera body.

An alternative view is that if many of the wealthier owners of DSLR bodies and lenses trade them in to buy something different, the availability of decent second hand stuff will continue.

 

My dslrs are still going strong although the iso capabilities of the new generation is tempting!

 

Only if you need them. I'm still amazed how good shots from my old-model DSLR can look at 1200 or 1600 ASA.

 

We discussed In-Body and in-lens stabilisation systems earlier but one thing I forgot to mention in regard to this is that some movements tend to be around the centre of mass (gravity) of the lens/camera assembly. With in-body stabilisation, any movement about the C of G will be seen as a translation rather than as a rotation. The rate gyros in a camera body will be great for short lenses but it will be the accelerometers that are doing the detection for longer lenses. Stick the gyros in the lens and they can be nearer* the C of G and thus give a more accurate measure of the rate of rotation. In practice the difference will be in the efficacy of the vibration reduction, e.g. fewer stops, However, with larger, faster, shorter zoom lenses the reduction in efficacy may be greater than with physically smaller lenses, the reason being that a physically longer lens, with the weight in the front element, will shift the C of G further away from the sensor.

I suspect that is as clear as mud to anyone who hasn't any experience in the field.

*The rate sensors cannot always be at the exact C of G because the mass of, say, a Nikon D3300 is considerably less than that of a D4 so with the same lens the D4 C of G will be nearer the camera than it will be for the D3300.

 

I’ve come across quite thick mud which is clearer than that.

 

I have a camera body with 'in body' image stabilisation, and I've found it helps for slow shutter speeds with lenses of the wide angle to 'standard' type (up to 70 mm full frame equivalent), but is much less effective at longer focal lengths. After 35 years of using a 35 mm film camera and all-manual lenses (before owning a DSLR with the 'in body' type image stabilisation), and coped without it, I don't rely on it now. Given the complexity of your explanation of the process (if I understand correctly), this might be a good idea.

 

Having started more than 50 years ago with a secondhand Pentacon FM and a 50mm Meritar lens (manual aperture and no meter) I reckon modern cameras are just plain marvellous. Auto focus, auto exposure, image stabilisers, electronic finders... You name it, I'll use it. Ignoring convenience is something I'll leave to the masochists, while I get on with just making pictures!

 

I thought it was a simple explanation!
There are three axes Vertical, Lateral and Longitudinal, all three pass through the C of G
There are six possible movements, along the axes and around the axes.
Movement along an axis is detected by an accelerometer, the accelerometer does not have to be near the C of G.
Movement around an axis is detected by a rate gyroscope, ideally the gyro is mounted near the C of G, preferably on the axis about which it is to detect movement.

A gyro mounted any distance from the axis it relates to isn't detecting rotation about that axis, if you move a rate gyro sideways or vertically it will not detect the movement, it only detects rotation about a single axis. Thus if a camera/lens combination has a C of G 100mm from the C of G of the body any rotation about the vertical axis will amount to moving a gyro in the body sideways so little or no output. An accelerometer in the same location will give an output. A gyro mounted on the vertical axis will give large output but an accelerometer at that location will have a very small output.

The ideal is to have accelerometers in the body and gyros in the lens.

Confused?

 

It does appear to be a very complex solution to the problem of carrying a tripod or learning how to hold a camera still (and knowing when you cannot).

 

2020 was the first year for ages that I haven't used my main DSLRs.
I use mirrorless bodies most of the time but usually the DSLR comes out for the airshow & some motorsports. Sadly Covid has ensured these shooting opportunities haven't arisen in 2020. I desperately hope they'll be back this year. It's really depressing to be at a circuit while racing is going on & not be allowed to come out of the pit garage & watch the action - let alone photograph things.
FWIW I did still use an old SD14 a couple of times this year for IR, so I've still not managed the year without using a DSLR.

 

Indeed, though many times more portable than a tripod. The computing power required would make the Apollo astronauts envious and the size of the various components is a huge achievement. The gyros are simply vibrating rings, from what little I have seen, and they are tiny. It is hugely impressive technology.

 

I am not much interested how lens and body stabilization works.
But I am very pleased that it does, as it helps a great deal with my age related tremor.
Lenses with stabilization work very well indeed.
However in body helps with any lens, even those that are already stabilized, or for heritage lenses that are totally manual.

There is no doubt at all that both routes will continue to be developed, and with the results even better in the future.

Any company that neglects this will suffer commercially.

Not long ago Fuji were saying that their lens mount would not allow for in body stabilization. They are now on to their third generation solutions.

 

Stabilization is not a substitute for a tripod.

For maximum precision of image placement, and repeatability nothing beats a tripod or studio stand.

 

I honestly don't know what to think.
Got a email from local camera store advertising new canon R5, I know nothing of it except that I assume its mirrorless, what made me baulk was the price..£4199.00.... Whatttt !!
Body only too !!
Had no idea they were that much, thank god mirrorless does not tempt me.

 

That's progress. The lenses are, like for like, more expensive than their EF equivalents too. By all accounts it is a very good camera but the price is a sign of the times. The price will reduce in time as they all do. So much of the overall market is won over by phones that there will be a migration to the higher end of the market and that means greater cost. For perspective, the 5Div (which the R5 is the mirrorless successor to) was something over £3,500 body only when it was launched. and the 1Dx is something around £6k body only. I was pleased that I'd gone for the 5Ds, which was more like £2,500, a year before it was launched.