[ElJamoquio]

Nope! That's the problem. I use lots of my old Pentax glass on my 5DII and they already show vignetting and serious border and corner problems on a static FF sensor. What if it starts moving, even just a tiny little bit!?

(Admitting: I am a big fan of 100% corner to corner sharpness. What's the use of going FF, with an "FF compatible lens" if it requires your to crop off the bad borders aftwards? Then I can just as wel keep mounting them on my crop sensor camera and shoot the best bit straight away.)

Most lenses IME have better corner sharpness on FF than similar lenses on APS-C. Will it be worse if a FF starts moving? Completely hypothetical of course, but I'd confidently take my chances with the FF having better corner sharpness, even when moving, for all but the widest lenses.

[Kunzite]

Not long ago, Ned's comments about new lenses being necessary generated a lot of heated comments. But I'd say it's the quality, not some hypothetical SR vignetting, that will make them necessary.
So, Pentax should release new, high quality lenses, sharp corner to corner (which would, of course, work well with SR). They will have to do it anyway; then, what's the problem? That old lenses, designed ages ago, won't give optimal results?

[Fogel70]

Nope! That's the problem. I use lots of my old Pentax glass on my 5DII and they already show vignetting and serious border and corner problems on a static FF sensor. What if it starts moving, even just a tiny little bit!?

(Admitting: I am a big fan of 100% corner to corner sharpness. What's the use of going FF, with an "FF compatible lens" if it requires your to crop off the bad borders aftwards? Then I can just as wel keep mounting them on my crop sensor camera and shoot the best bit straight away.)

Are you saying that Pentax FF lenses are much worse on this than Canons own FF lenses? If it's large aperture lenses used with large apertures, then you will find the same problem on most lenses, and the cure is usually to stop down the lens more. You will find same type of problems on wide aperture APS-C lenses too, but Pentax don't have many of those.

What are your solution for Pentax going FF? Should they forget about it altogether, because some old FF lenses might not give you 100% IQ under some circumstances?

[Clavius]

Are you saying that Pentax FF lenses are much worse on this than Canons own FF lenses? If it's large aperture lenses used with large apertures, then you will find the same problem on most lenses, and the cure is usually to stop down the lens more. You will find same type of problems on wide aperture APS-C lenses too, but Pentax don't have many of those.

What are your solution for Pentax going FF? Should they forget about it altogether, because some old FF lenses might not give you 100% IQ under some circumstances?

No, just that they need to develope a whole new line of lenses. It's a much better explanation why Pentax hasn't joined the FF bandwagon all this time, despite the fact that it makes so much sense. A much better explanation then them being taken over multiple times, becaue that didn't stop them from making world's best APS-C DSLR, or the 645D.

Mind you, if I was them, I would make the new FF DSLR mount incompatible with old lenses. Just to avoid people whining about how poor the new Pentax FF DSLR performs with old lenses. They don't understand that it's the other way round. Luckily, I'm not them though.

[Fogel70]

No, just that they need to develope a whole new line of lenses. It's a much better explanation why Pentax hasn't joined the FF bandwagon all this time, despite the fact that it makes so much sense. A much better explanation then them being taken over multiple times, becaue that didn't stop them from making world's best APS-C DSLR, or the 645D.

Just like any other manufacturer of FF they need to update lenses to optimize them for digital and to keep up with IQ of the cameras. But it's the same for APS-C and 645D system too.

I believe the biggest reason for Pentax not entering FF bandwagon is that they will struggle to make any profit out of it. With the small user base Pentax have they either has to make higher profit on each sold camera (higher price than competition), or they have to wait until FF technology is cheap enough to sell them in large enough quantities. They also need a system that can grow Pentax user base in the long run. So Pentax need to be much more successful than Sony has been on FF.

Mind you, if I was them, I would make the new FF DSLR mount incompatible with old lenses. Just to avoid people whining about how poor the new Pentax FF DSLR performs with old lenses. They don't understand that it's the other way round. Luckily, I'm not them though.

I agree with you, I'm glad you're not Pentax.

[RobA_Oz]

So... The extra large image circle of the DA lenses is just for kicks?

0.3mm? Are you serious? Take the lens off your Pentax DSLR, switch it to LV and stare down the mount. You'll see the sensor float around A LOT more then just 0.3mm. (Seriously cool to see though!)

BTW, the 3mm sensor movement is not something I made up. It is confirmed as the SR sensor movement.

We've been down this particular rabbit hole a number of times. The full sensor movement available is not what's used for shake reduction – the 0.3mm may even be a slight exaggeration. The full amount of movement is available for level correction, Live View shifting and astro-tracing. Let's not get hung up on this yet again.

[isenegger]

Pentax had in the old days of film a very good reputation as think tank of the big companies and the lens between 1970 and 1990 were (my opinion) better than Canons FD and on par with Nikons F. A lot of professionals worked with the manual Pentax.

But with the race for the best AF they lost new customers to Nikon and Canon and even more as they forgot to invent a digital SLR.

I changed from my LX to Canon 20D and 5D, because Pentax had nothing comparable. And the last professional model with the Z1n was long ago. I waited with changing the system, because the lenses I had were superb.

For my work I use today Canon with EF-lenses and Olympus mFT, sometimes my SMC Takumars on 5D and for fun and creative work I use Nex and Canon with my old Zeiss, Nikkors and Takumars.

I think it is important for sustainability of the company that Pentax has a FF Cam and lenses for professionals. Equipment for $500 (zoom plus body) you will change every 2 years, no matter which company it is. A couple of lenses for $1000-$2000 you will keep a long time and put again and again the new body behind.

[jogiba]

No, just that they need to develope a whole new line of lenses. It's a much better explanation why Pentax hasn't joined the FF bandwagon all this time, despite the fact that it makes so much sense. A much better explanation then them being taken over multiple times, becaue that didn't stop them from making world's best APS-C DSLR, or the 645D.

Mind you, if I was them, I would make the new FF DSLR mount incompatible with old lenses. Just to avoid people whining about how poor the new Pentax FF DSLR performs with old lenses. They don't understand that it's the other way round. Luckily, I'm not them though.

That would be the dumbest decision Pentax made ever if they did that. There are people getting old Pentax lenses like the SMC Pentax-FA* 200mm F4 Macro ED [IF] converted to Canon and Nikon mounts to use on full frame DSLRs.

[bossa]

Lens selection would not stop Pentax from going full frame. Even if they dont launch a single new lens they still have a wider selection of lenses for the new full frame camera then Fuji X-system, Sony NEX, Panalympic mFT, Sigma and Samsung NX had at their launch dates. Still, im pretty shure Pentax would launch at least a few extra full frame lenses simoultanously. It would make a great lens selection from day one and only grow.

The Pentax FF lens selection was designed for static film cameras and so if they implemented IBIS via SR on a new FF camera I would expect the FA limiteds etc to require switching SR OFF. We would probably end up with random dark artifacts around the edges of images due to the Image Circle not being large enough to handle the moving sensor.

[Clavius]

The Pentax FF lens selection was designed for static film cameras and so if they implemented IBIS via SR on a new FF camera I would expect the FA limiteds etc to require switching SR OFF. We would probably end up with random dark artifacts around the edges of images due to the Image Circle not being large enough to handle the moving sensor.

Thank you, Bossa. I'm convinced of that too. But others here prefer to think the sensor only needs to shift 0.3mm for SR functionality, and that the extra 2.7mm was implemented by Pentax for fun or to show off.

Off course, Pentax will be showered with bad reviews and bad user experiences for those dark corners and borders, despite the fact SR can be switched off.

[Fogel70]

Thank you, Bossa. I'm convinced of that too. But others here prefer to think the sensor only needs to shift 0.3mm for SR functionality, and that the extra 2.7mm was implemented by Pentax for fun or to show off.

If you read or understood my earlier posts on the matter you would realize that more than 0.3mm movement is needed, But the total distance of 3mm is needed for acceleration of the sensor before the image is captured. But it does not matter if acceleration is done outside the image circle of the lens, as it will not be captured on the images.

My calculations is that with a 24MP FF sensor the sensor movement for 4 stops SR is 50-100 pixels. Which is 0.3 - 0.6mm, or rather 0.15 - 0.30mm outside the image circle on each side as the sensor should be centered during capture.

Off course, Pentax will be showered with bad reviews and bad user experiences for those dark corners and borders, despite the fact SR can be switched off.

I highly doubt that.

[Paul MaudDib]

The Pentax FF lens selection was designed for static film cameras and so if they implemented IBIS via SR on a new FF camera I would expect the FA limiteds etc to require switching SR OFF. We would probably end up with random dark artifacts around the edges of images due to the Image Circle not being large enough to handle the moving sensor.

Why would you end up with artifacts even assuming there was not an extra 0.3mm of wiggle room in the image circle? If there is no light there, there is no light to activate the sensor and nothing to generate artifacts. Worst case you would get a darker corner/side in the direction of travel. It could probably be corrected automatically in Photoshop/Lightroom if you made a plugin and knew the acceleration used and the coverage of the lens (or just look for an intensity gradient).

And many lenses have an image circle much bigger than strictly necessary. For example, the Fish Eye Takumar has an image circle more like 45mm (someone built a rollfilm camera to capture the full circle). The SMC Takumar 35/3.5 reportedly has a much larger image circle, one the order of 50-55mm IIRC, but with sharpness loss towards the edges. Someone else was doing stereo images with the fisheye and the 35/3.5, but I can't find the page at the moment. They sold their rig on APUG a while back.




http://photo.net/classic-cameras-forum/00YQrJ

Some large format lenses are mechanically vignetted to prevent light from bouncing around the body and causing flare, but I don't know if that is common on 35mm. My gut says not, because with retrofocus lenses smaller angles of coverage should be necessary. The few examples I've found of people doing full-image-circle capture suggest they are not mechanically vignetted.

edit: Dug up the other site:

Above is a photo of the camera with a real M-42 (Pentax) screwmount, replacing the original plywood friction fit mount. (Thanks to Tony Z). I have tested it using my 17mm f4 Fish-Eye-Takumar and my SMC Takumar 35mm f3.5 lenses with excellent results. The fisheye produces a sharp 45mm circular image. The 35mm lens produces a 54mm circular image with some colour fringing in the outer 2mm. The camera is basically completed except for gluing down the flocked paper that light seals the front and permanently attaching the fisheye viewfinder.

http://www.angelfire.com/ca/erker/fisheye66.html <-- site
https://www.pentaxforums.com/forums/sold-items/162920-sale-sold-custom-made-f...r-17mm-f4.html <-- sale thread

[Clavius]

If you read or understood my earlier posts on the matter you would realize that more than 0.3mm movement is needed, But the total distance of 3mm is needed for acceleration of the sensor before the image is captured. But it does not matter if acceleration is done outside the image circle of the lens, as it will not be captured on the images.

Yeah? You still really think 3mm is needed for acceleration of such tiny parts?

Last try: If you take a very fast lens, aperture wide open, switch on LV and stare down the lens. Shake your camera a bit, you can just make out the floating sensor. It never vanishes behind something. Not even the corners. With that reality in mind, who cares about calculations and theory?

[ElJamoquio]

Yeah? You still really think 3mm is needed for acceleration of such tiny parts?

Last try: If you take a very fast lens, aperture wide open, switch on LV and stare down the lens. Shake your camera a bit, you can just make out the floating sensor. It never vanishes behind something. Not even the corners. With that reality in mind, who cares about calculations and theory?

Huh? I'm not sold on the 0.3 mm yet (Pentax doesn't know when you're going to click the shutter... is SR only active after you hit the button?) ... but in any case your argument that the sensor doesn't disappear behind anything doesn't seem to be relevant... could you explain to me how it IS relevant? The biggest rear lens element I own is less than 29mm in diameter. The light spreads out from there.

[Fogel70]

Yeah? You still really think 3mm is needed for acceleration of such tiny parts?

Not always, but the longer the focal length is the faster the sensor need to move. As Pentax support SR up to 800mm there is need for a quite long acceleration. But I guess it also is a bit over sized so the sensor don't hit the end stops in normal use.

Last try: If you take a very fast lens, aperture wide open, switch on LV and stare down the lens. Shake your camera a bit, you can just make out the floating sensor. It never vanishes behind something. Not even the corners. With that reality in mind, who cares about calculations and theory?

Not sure what you point about this is?

SR in LV is very ineffective as it has to move much more than in normal use when the sensor only move when you press the shutter button. Try removing the lens and set FL to 800mm in LV, and you will feel what force it takes to move the sensor the whole 3mm distance in full speed. I don't really think that Pentax had LV and video in mind when they designed SR, so it do not work very well for them. Especially with long focal length.