FF antishake?

What's the word on the feasability of applying KM's antishake

tecnology to a full-frame sensor?

 

Obviously, KM (as well as Nikon, Leica, etc) is working on it's own FF

dslr. Any mention of a time frame? The Canon 5D's shadow looms large!

Two points:

 

-A FF sensor is 2.25x larger (by surface) than their current sensor, i.e. 2.25x the weight to move around.

 

-It's unclear how much anti-shake would have to move the sensor toward the edge of the image circle, and how much that would affect image quality (though I believe that it's negligible)

<i><blockquote> Obviously, KM (as well as Nikon, Leica, etc) is working on it's own FF

dslr.

</blockquote> </i><p>

 

Maybe not. KM isn't investing much into its poorly-performing camera division

compared to the rest of its corporate business. It announced a partnership recently with

Sony (which provides most sensors to KM, Pentax and Nikon), which would allow for

cost-sharing, but whether Sony is able to produce affordable FF sensors is another

question.

Jean-Baptiste- Just because a FF sensor has 2.5X the surface area of the APS-C sensor, it doesn't mean that the weight would be 2.5X the weight. The thickness of a FF sensor is not going to be 2.5X the thickness of an APS-C one. And if the weight were a factor, KM's engineers could compensate for it some other way.

 

Also, KM has already said that they will concentrate a 'value-added' cameras, and I would certainly think Anti-Shake fills that bill.

it wouldn't need to be 2.5x the thickness to be 2.5x the weight; if it's the same thickness but 2.5x as large (surface area), then it's 2.5x the weight (presuming same density).

 

If it were also 2.5x as thick, that would make it (2.5 x 2.5) the weight.

Mark- My Bad!!! You're right, but I'm sure KM is well aware of the added weight, but I would hardly think that it would prevent a FF sensor from having Anti-Shake. It might take using motors with more torque or a lighter material on which to mount the sensor. Anyway, I'm willing to put money on a FF sensor camera having Anti-Shake.

AS on a FF sensor would require that the glass project a larger-than-FF image circle, which I highly doubt Minolta glass does.

Edward- Some one could mount a lens on a piece of cardboard and attach it to a darkened box with a piece of paper with the dimensions of a 35mm film marked on it, placed at where the focal plane is on a film camera and measure the image circle. The focal plane would be easy to determine, as it is engraved on the top of camera. You would then shine a light thru the lens and focus and measure the size of the image circle.

I doubt it. I don't really think Konica-Minolta saw far ahead to the FF DSLR + AS when they

designed the Maxxum AF mount. I would be surprised if we ever see a full frame DSLR from

Konica-Minolta or Pentax or Nikon for that matter.

<a href="http://forums.dpreview.com/forums/read.asp?forum=1035&message=14484146">On DPReview</a>

Interesting posts on that forum. Not only about a FF AS dSLR, but the possibility that the 9D will have a upgradeable design. This would carry on Minolta's idea that the Maxxum 9 would be upgradable. It would be easier on a digital camera as each new development leads to smaller components, so the body can accomodate the changes, just unplug the old part and plug in the new. The part of the posts I didn't care for, was 'KM SEA's' implication that Sony will be doing more that jointly-designing and selling dSLRs with KM. It sounds like Sony will buy the photo side of the business from KM. But, I guess if that happened, I will still be able to use my lenses.

If true, this makes it a war between Sony and Canon, as Sony is producing most of the

chips for the DSLRs sold by Nikon, Pentax and KM, in addition to its own top-selling

digicams.

 

Now only if Kodak could make a low-cost FF sensor, we'd have a real price war.

If they are to do it properly they will have to pay not only a lot of attention to AS issues (image circle and increased mass) but probably more importantly telecentricity. If they do not want their user base to be forced into buying hugely expensive new wide angle lenses they will have to make a breakthrough on layout of the micro lenses to solve the angle of incidence issue for sensors (unlike film) which have a finite 3rd dimension. i hope if they do go FF they include a D2X like option to, say, a 1.3X crop that allows us to use lenses with non astronomical price tags.

 

Canon has thrown down the gauntlet though, and Kodak has already had several goes at some very good sensors. There is also Foveon, Fuji and Panasonic who are pushing the envelope. So i should think Sony, perhaps with a little assistance from all her partners (Nikon, Pentax, Konica, Minolta...) should be able to make a very good sensor soon.

 

More i think about it the more i like the idea of a modular camera. Joe the Fork on DPReview likes to pooh pooh the idea noting that 90% of the cost is in the electronics, which pretty much needs to go as a unit (processing power needs to be commensurate with MP and processing requirements). But it sure would be nice to get a quality body which lasted as long as my XD11 has! Why discard and pay every couple years for professional weather sealing and controls? Electronics progress pretty much is forcing that. With delicate electronic cameras the weather sealing is all that much more important.

There is a question that has been nagging away at the back of my mind for some time, and this thread has brought it to the fore once again.

 

While it is desirable to have a full frame sensor so all those 35mm lenses don't go to waste, is the race to develop a FF sensor actually advancing digital photography or just taking it off on a side line?

 

There are clearly technical issues with producing a FF sensor (otherwise all DSLR's would already have them). I can't help but wonder if in the future a new 'digital format' will evolve and 35mm will just be a dim and distant memory for us film-loving luddites.

 

If the technical experts and R&D gurus were given a totally free reign, what size would they choose?

 

Any views?

Matt- I think the fact that the 35mm frame size has been around for so long and that there are many millions of lens around sized for this format makes it the de facto choice for many people. The proportions of the 35mm frame is very pleasant and with a little cropping is expandable to other print sizes. No other format seems to have had the impact that the 35mm frame dimensions have had.

Looking at a few basic test shots at dpreview, it would appear that KM's anti-shake mightbe able to compensate for approximately 8 pixels worth of shake, i.e. less than 1/10th of a mm, i.e. not something significant (the sensor of the Canon 5D is actually 35.8mm wide, i.e. you could move it by 1/10th of a mm and it would still fit in a 36mm film frame).

 

It would be interesting to quantitatively attempt to assess the effectiveness of anti-shake.

 

About the format that would be designed "from scratch": does it really matter? The E-1 and E-300 "designed-from-scratch-ness" does not seem to translate into a fast sweeping market advantage. The question seems fairly futile. Finding an absolute answer has little practical significance. There's no unique answer (there's a ratio of about 10:1 between the characteristic size of the smallest and largest sensors on the market). There are numerous examples where the existing legacy prevails on any attempt to bring logical change. A big historical precedent is the failure of the French revolutionaries to establish metric time: there were too many clocks out there that worked with 24 hours a day and that couldn't practically be changed to 10 hours a day.

JB- From the test I've seem in the photo magazines, the efectiveness of AS on the 7D is about the same as that found in the Canon and Nikon cameras, about 2-3 stops.

"about 2-3 stops" isn't a precise measurement, though. 3 stops is double the effectiveness of 2 stops.

 

I'd hope to see measurements done to 1/3 of a stop, with a reasonably well defined notion of "hand-holdable speed" and a significant number of samples. E.g. the tests at dpreview would reasonably suggest that without AS, shooting at 1/4s yields better results than 1/8s.

JB- The magazine was Popular Photography & Imaging, one of the leading photo magazine in the US. So, if the 2-3 stops is not good enough for you, take it up with them. They are not a fly-by-night operation, and has been in business for decades.

Clinton: I didn't want you to take it personally, I'm sure that you properly quoted a reputable source. In this case it seems that the reputable source is still too unprecise for my taste, I personally believe that it's reasonably easy to get better measurements.

 

To take another example, imagine what would happen if the result of a car test said "those cars accelerate about the same, about 5 to 10 seconds for a 0-60".

JB- I don't know what country you live in, but Popular Photography & Imaging is 'The World's Imaging Magazine' and therefore very reputable. In their February 2005 issue where they do a full test of the Maxxum 7D, they say that the AS function's "...test show that the system actually produces a 2 to 4 stop improvement over the minimum recommended shutter speed (based on 1/focal length)depending on how steady you are or how much caffeine you've had to drink (a big blur factor.)"

 

"That's the equivalent of boosting your ISO from 200 to 1600 without adding noise. Or changing your f-stop from f/2.8 to f/8 to extend depth of field without being forced to set-up a tripod."

 

If that isn't good enough for you, then I suggest you do your own tests.

I've read their review, and their comment about anti-shake just doesn't live to the standard of precision that they usually use (most of their numerical test results have 2 or 3 significant digits, their comment on AS, which you quoted, doesn't even have a single significant digit). Their methodology for the anti-shake test doesn't seem to be as advanced as what they use for their other tests, and since they don't publish the actual methodology I can't reproduce their tests. At least the test at DPreview has some information about the test environment (Phil mentions what lens and what focusing distance he used). Neither seems to define an objective notion of hand-holdable.

 

Unfortunately I don't own IS/VR/OS/AS equipment, which means that I can't actually perform such tests myself.

And, yes, before someone suggests, I could actually rent an image-stabilized lens and run tests with it.

JB- You must be the only one who is so obsessed with this matter. If you go to Canon's, Nikon's, or Sigma's website, none of them gives a one, two, or three decimal spec of their anti-vibration lenses. As the Anti-Shake function is not recommended for use of a tripod, you would have to come up with a machine that can be computer controlled to test the equipment that would simulate someone handholding the equipment. And even if they did give you a multi-decimal answer, what good would it do in real life? You're going to hand hold it anyway.

 

My suggestion to you is to let it go and get on with your life.

JB- To expand on my earlier post. What difference is it if the AS gives 3.85 stops or 4.01 stops in real life? The amount your hands shakes varies from moment to moment. Let's say, you are able to take a clear shot at 1/20 sec., then you drink a double shot of expresso. Five minutes later your hands shake more so that under the same light conditions, you have to shot at 1/100 sec. to get a clear pic. Or, you're shooting during a hurricane, and a gust of wind hits you and all of sudden you need a higher speed. Or, you're taking pictures of a beautiful girl in low light, and she starts to take off her clothes, so your hands start trembling.

 

As I said, the actual amount of correction is going to vary, as the amount your hands shakes varies. So, if one time you can hand-hold at 1/20 sec and the next time you need 1/30, are you going to blame the equipment? Your desire to put a definative number on the amount of correction, has little bearing in real life.