[maxfield_photo]

It's pointless to worry about corner sharpness if your image is blurry from camera shake. I'll take IBIS over nothing any day. (Still, it would be nice to have lens based IS for video). Furthermore, it's not the mount diameter that is important, it's the size of the image circle. On a mirrorless camera those two things are much more closely related, but on a traditional DSLR, the image circle has room to expand once it leaves the rear element. This article sounds like it was written by a college sophomore photography student, in other words, a wise fool.

[MadMathMind]

This dude is full of it...His size comparisons are not like for like (premium zeiss 55 1.8 vs budget canon 50 1.8..pppffffffft..they are only similar in largest aperture). Clearly he is just trying to justify his purchase by talking junk about other brands...and obviously he has no idea what he is talking about.

I think he had some good points but I agree that it definitely read like someone trying to defend every engineering decision his product of choice has made as the absolute best. I remember reading a lot of arguments like this one during the HD-DVD/Blu-ray format war:

"The people really want picture-in-picture!" [HD-DVD had it, BD didn't]
"Internet connectivity is for chumps!" [BD didn't have it, HD-DVD did]
"Lossless audio? No one cares about that!" [HD-DVD titles mostly didn't have it, BD did]
"50GB > 30GB" [BD capacity vs. HD-DVD capacity]

Truth is that none of these things decided who won the format war.

[clackers]

I think he had some good points.

He produced *no* evidence, MMM. ☺

Where are his sample pics?

Look, Fuji are trying to attack a technology they don't have, just like they did Full Frame.

From the company that gave us the ISO Rating Scandal:

Youtu.be

Youtu.be

[MadMathMind]

He produced *no* evidence, MMM.

Hmm, it seems the actual forum post itself was not posted. He begins by talking about the size of lenses. He illustrates the problem: fast FF lenses are huge and the mirrorless body is not. That was a pretty good point.

I read it elsewhere and though there was a link in the OP to the whole thing.

[Fogel70]

Hmm, it seems the actual forum post itself was not posted. He begins by talking about the size of lenses. He illustrates the problem: fast FF lenses are huge and the mirrorless body is not. That was a pretty good point.

I read it elsewhere and though there was a link in the OP to the whole thing.

But if the lenses are "equal" the size do not have to differ much.
(FI f/1.4 lens on APS-C vs F/2 on FF)

The fuji guy is a product manager/planner, so deeply involved in the priorities fuji made.
And it is his job to convince the market that those priorities are the best.

It do not really matter if a fast lens is designed for APS-C or FF, if you want the same performance on both (same max aperture size and same FOV) the size usually do not differ much.

Fuji X-pro 2 with 35/1.4 vs Sony A7 II with 50/1.8. These cameras has same price, but the Sony FF lens is half the price of the Fuji APS-C lens.

[clackers]

Hmm, it seems the actual forum post itself was not posted. He begins by talking about the size of lenses. He illustrates the problem: fast FF lenses are huge and the mirrorless body is not. That was a pretty good point.

I read it elsewhere and though there was a link in the OP to the whole thing.

Well, I read it, MMM.

He provides quite good pics for the size issue, if leaving out height.

He does no such thing for the IBIS point, which is his weakest.

I think we'd have to be pretty gullible to be taken in by that.

Pentax's engineers when interviewed said the IBIS challenge was to keep the body size similar to the K-3. They had to make the magnets stronger.

I have not heard of Sony A7II or A7IIR owners asking for a refund.

Fuji marketing throwing mud on other CIPA members seems ingenuous.

[Mountain Vision]

We should not forget, also, that IBIS gives us pixel shift, anti-alias filter simulation and astrotracing. Put that in your pipe and smoke it, Fuji!

I was thinking this. While the IQ or stability may or may not be superior in one or the other. I think if IBIS is close enough to lens based, factoring in the other advantages, including sensor cleaning (which might be in other cameras now, or might be a gimmick, but is still another feature) and what you posted above, it seems like there are a lot of pros in the pros and cons list for IBIS.

It also seems mirrorless has size advantage from the short register distance, but this might also create issues for IBIS on larger sensors on the same mount. Perhaps it's another benefit of SLR VS MILC that isn't often discussed.

[beholder3]

If the whole discussion here is about this here:
Why Sony's Full Frame Pro Mirrorless Was a Fatal Mistake
then you seem to be referencing the Fuji sales guuy stating

"It means the amount of light at the corners is reduced when the sensor is shifted. We could correct it digitally, but we don’t want to do it: we don’t want to compromise our image quality."

I do think this is a joke, when you look at a old FA50/1.4 which vignettes 0,7 EV in the corners uncorrected (Pentax SMC-FA 50mm f/1.4 - Review / Lab Test Report - Analysis & Verdict) and then also have a look at the Fuji 56/1.2 which vignettes 1,7 EV in the corners. And "surprisingly" Fuji does all it can to reduce this whopping 1 EV stringer vignetting by digitally preprocessing it in camera for their loved JPGs:
Fujinon XF 56mm f/1.2 R (Fujifilm) - Review / Test Report - Analysis

In RAW mode the vignetting is heavy with a peak of ~1.8EV (f-stops) at f/1.2. ... In auto-corrected JPEG mode, the camera digitally corrects the vignetting down to 0.9EV at max. aperture and 0.6EV at f/1.6. Interestingly it is fairly obvious that the camera "presses the brake" at these settings

So, yes Fuji does compromise image quality with regards to vignetting.
And yes, the hide it in camera by digital software processing.

It is simply impossible to create worse results using SR.
And it is extremely likely that SR especially on fast glass in dark ambient light offers you 3-5 stops less noise = image quality. Remember: The Fuji APSC without SR will have the same noise as the Pentax APSC with SR at 4 stops lower ISO. That is go compare Fuji ISO 3200 with Pentax APSC ISO 400 or even K-1 ISO 400.
Good luck, Fuji.

Not building in a SR means you are actually sacrificing a huge amount of image quality for many users all the time.

[Mountain Vision]

I do think there are sour grapes on Fuji's part. They have an APS-C only camera system. How well that sells in a few years when full frame prices are well below a thousand dollars is hard to say. Certainly APS-C will be here for awhile, but Fuji wants to sell professional cameras with high prices and it might be tough to do down the road.

As someone who was seriously considering a switch to fuji for the smaller bodies, I have no immediate plans to move to full frame. I do really like the design and tech in the k-1 and really hope a K-2 is released with it. But I don't like the size of the body (which isn't big, just bigger than a K-3, which I consider the largest I want to go).

I think APS-C is good enough and I think if I needed more resolution and DR, the 645D is affordable enough.

Too me, FF was always an incremental upgrade. All that said, I'm not down on it. I think the he K-1 looks like a winner, I think it's good for Pentax, and I think it's good for Pentax photographers. I also think Fuji, like Olympus, painted itself into a corner. Both make great products but they didn't leave themselves a lot of room to grow, if need be.

[Rondec]

As someone who was seriously considering a switch to fuji for the smaller bodies, I have no immediate plans to move to full frame. I do really like the design and tech in the k-1 and really hope a K-2 is released with it. But I don't like the size of the body (which isn't big, just bigger than a K-3, which I consider the largest I want to go).

I think APS-C is good enough and I think if I needed more resolution and DR, the 645D is affordable enough.

Too me, FF was always an incremental upgrade. All that said, I'm not down on it. I think the he K-1 looks like a winner, I think it's good for Pentax, and I think it's good for Pentax photographers. I also think Fuji, like Olympus, painted itself into a corner. Both make great products but they didn't leave themselves a lot of room to grow, if need be.

I don't really think Fuji feels threatened by Pentax (although maybe they should be ). I think the big threat to them is Sony, who has managed to make full frame mirrorless cameras that are roughly the same size as Fuji mirrorless APS-C cameras, focus faster and have bodies that offer stabilization on the sensor.

The end result is that their paid guns have gone on tirades about how there isn't that much difference between crop and full frame sensors (there's a little better than a stop difference in dynamic range and high iso performance) and how in body image stabilization harms image quality (it doesn't seem to so far).

I'd like to see Fuji stick around for the long term. They have interesting bodies and really nice lenses, but I'm afraid that they have chosen a specific niche that is only going to shrink over time and selling their X Pro series for 600 or 700 dollars like the K3 series is doing now probably doesn't give them a lot of return on investment to continue the process in the long term.

[Simen1]

And exactly how far does the sensor "move around"? Obviously it is only a very small amount.

On both APS-C and FF its 1,5mm in all directions from center (up, down, left, right). That means the active sensor area operates within a frame of 38,9 x 27 mm (FF) or 26,7 x 18,7 mm (APS-C). In other words image circles with diameters of 47,35mm and 32,60 mm. Note that most lenses have slightly larger actual image circle then the sensor size its designed for. Its the worst case settings (focal length, aperture and focus distance) that defines the image circle it is designed for. At other settings then worst case, the image circle will be larger.

If it extended beyond the image circle I am sure the shake reduction software would compensate. But all the talk of mount diameter is a bit of a red herring. The image circle is formed on the sensor (the focal plane). On any FF lens I have seen, the diameter of rear element was far less than the image circle would be because it is doing the final focussing of the image on the sensor. As long as the mount can accommodate the rear element, it will do the job. If it were not so, we could not adapt M39 lenses to K-mount. I would say the interview is either a beat up or an April Fool piece. In-body SR works.

I just measured the rear lens diameter of an old Sigma 100-300mm for FF: 19mm, and a DA50mm f/1,8: 23mm. Both far less then the full frame sensor size they support.

[Na Horuk]

On both APS-C and FF its 1,5mm in all directions from center (up, down, left, right). That means the active sensor area operates within a frame of 38,9 x 27 mm (FF) or 26,7 x 18,7 mm (APS-C). In other words image circles with diameters of 47,35mm and 32,60 mm.

Um, how did you get 47,35mm? Sensor is around 35mm, add 2*1.5mm, you get 38,9mm. And I'm not even sure the SR needs a full extra 1.5mm for all lenses - I suspect for normal and wide angle lenses this number should be smaller; unless you are really shaking the camera. But yes, 1.5m on all sides shouldn't be a big problem. I'm sure most FA lenses will fit this comfortably, and even some DA lenses (and all DFA lenses). And these are all still fairy compact designs, compared to Sigma, Sony, Samyang..

I don't really think Fuji feels threatened by Pentax (although maybe they should be ). I think the big threat to them is Sony, who has managed to make full frame mirrorless cameras that are roughly the same size as Fuji mirrorless APS-C cameras, focus faster and have bodies that offer stabilization on the sensor.

Yeah. Sony was the main target of that article, Pentax was just mentioned to keep things "fair". I have no love for Sony, but they definitely have so much money and are producing so many different technologies (they had DSLR, SLT, now mirrorless, with PD AF; from regular sized cameras to super compact FF; super fast burst rate, etc. etc.) that some are bound to be interesting, and thus threatening to other brands.

[RobA_Oz]

Um, how did you get 47,35mm?

Pythagoras: the square root of (38.9 x 38.9 + 27 x 27) is 47.35.

---------- Post added 5th Apr 2016 at 09:53 PM ----------

On both APS-C and FF its 1,5mm in all directions from center (up, down, left, right).

I suspect that 3mm envelope is probably much more than is required for Shake Reduction. It represents 8.3% of the longer side and 12.5% of the shorter side of a 35mm sensor, which wouldn't be so much blurring of an image as smearing it. For an APS-C sensor, it's even more pronounced. The full envelope is probably more for Astro-tracking, perspective (image shift) correction, or horizon correction. By the way, do you have a source for the 1.5mm shift figure? Speculation about the actual SR requirement caused a few heated exchanges prior to the announcement of the K-1.

[PJ1]

On both APS-C and FF its 1,5mm in all directions from center (up, down, left, right).

Thanks for that. I searched and couldn't find it.

[Fogel70]

I suspect that 3mm envelope is probably much more than is required for Shake Reduction. It represents 8.3% of the longer side and 12.5% of the shorter side of a 35mm sensor, which wouldn't be so much blurring of an image as smearing it. For an APS-C sensor, it's even more pronounced. The full envelope is probably more for Astro-tracking, perspective (image shift) correction, or horizon correction. By the way, do you have a source for the 1.5mm shift figure? Speculation about the actual SR requirement caused a few heated exchanges prior to the announcement of the K-1.

If you accept 20 pixel blur on K-1 as a sharp image then the sensor need to move +- 1.5 mm (~600 pixels) to compensate for 5 stops of SR.
If you accept 5 pixel blur on K-1 as a sharp image then the sensor need to move +- 0.4 mm (~150 pixels) to compensate for 5 stops of SR.

But I doubt we will see a 5 stops improvement in every shot captured, it's probably more like 2-4 stops in real use.
Like on most other image stabilization systems the CIPA rating is most likely over rated compared to real world usage.