5

In an interview done with a Fujifilm senior product planner, it was said that XF-mount lenses are incompatible with IBIS, and that Fujifilm does not want to sacrifice image quality. But in the recent launch of the Fujifilm X-H1, IBIS was introduced. I'm not sure whether the image quality is compromised with the IBIS, or has Fuji finally found a solution that can do this without any degradation of IQ?

  • I'm guessing from context that IBIS is in-body image stabilisation, but it might be useful to actually clarify that. – Tetsujin Mar 11 '18 at 10:32
  • Yes, IBIS = In-Body Image Stabilisation – rcs Mar 11 '18 at 12:21
  • 2
    The point of any image stabilization solution, whether in body or in lens, is to improve image quality in the presence of unintentional motion (hand shakes, vibration, etc.). The better question might be more along the lines of "under what conditions will this particular stabilization solution fail to improve the image quality by eliminating/reducing such factors", or "under what (hopefully less common) conditions might the stabilizer actually introduce errors that wouldn't have been there otherwise"? – twalberg Mar 11 '18 at 13:06
  • @twalberg No, the point is that with IS we willingly trade one problem for another, We allow intentional misalignment of the lens/camera system in exchange for reducing motion blur. So the question is, does the benefit of reducing motion blur outweigh the IQ penalty? The answer to that depends on several variables. – Michael C Mar 13 '18 at 0:49
3

In the interview, Takashi Ueno of Fujifilm explains why he thinks the XF lens mount is inadequate for in-body IS - shifting the sensor means some parts of the sensor will be shaded by the mount. In other words, the ideal lens mount should be larger diameter.

Now, is the image quality compromised in the recent Fuji X-H1, featuring IBIS? According to the above statement - yes, it is compromised. But I doubt the difference is easily noticeable. It is just a little bit more vignetting. After all, and we get some vignetting all the time, regardless of the image stabilization. Fujifilm's small APS-C sensor enjoys plenty of space behind their XF mount, compared to the Sony full frame cameras using only 2 mm wider Sony E-mount.

Alternatively, we can guess that the Fuji guy simply wanted to turn the camera's deficiency into merit. Look, our competitors offer IS, but we don't and this is good, because it means better quality. Now, 2 years later, let's forget about it, IS is now good because we have it as well.

  • So which is it? The first two paragraphs or the last paragraph? – Michael C Mar 11 '18 at 15:03
  • 3
    there is no contradiction really. vignetting is true, the image quality will be influenced. whether the influence is a real concern depends on who is talking. manufacturer of the non-IBIS camera wanting to point out disadvantages or the manufacturer of IBIS camera neglecting disadvantages. can be even the same person 2 years later :-] – szulat Mar 11 '18 at 15:10
  • The contradiction is that two years ago a high ranking executive of Fuji said "We can't use IBIS with the XF mount without a reduction in image quality and we won't do that." Now, two years later, with the introduction of the X-H1 they are saying, "Never mind." – Michael C Mar 13 '18 at 0:56
  • @MichaelClark Seemingly, the reduction in image quality is now acceptable to them, either through adjustment of standards or improvement of IBIS or both. That is not a contradiction. – Belle-Sophie Mar 19 '18 at 8:53
  • If the image quality reduction was not acceptable to Fuji two years ago and is acceptable to Fuji now that is a contradiction. What they say now is contradictory to what they said two years ago. Even if they adjustested their standards or even if they had a technological breakthrough the two statements made two years apart are contradictory to one another. Contradiction: 1) a combination of statements, ideas, or features of a situation that are opposed to one another. 2) the statement of a position opposite to one already made. – Michael C Mar 20 '18 at 9:17
2
  • We consider a lens with unwanted tilt or shift to be less than ideal because either affect image quality in one way or another.
  • We also consider motion blur caused by camera movement to be less than ideal.

In the case of image stabilization, whether based in the camera or in the lens, we accept a very slight amount of misalignment of the lens/sensor system in exchange for reduced blur due to camera movement. This is true of all image stabilization systems, either lens based or camera body based.

  • Lens based systems use tilt movements.

When the lens' IS element is tilted in any way away from exactly perpendicular to the center of the lens' optical axis there will be a slight reduction in image quality in terms of overall sharpness. The slight tilt induced by lens based IS will also affect things such as astigmatism and other aberrations as well.

  • IBIS (In-Body Image Stabilization) systems use shift movements.

Movements that keep the imaging sensor perpendicular to the lens' optical axis would, in theory, not affect image quality in the same way that tilt movements do. But shift movements introduce issues of their own. By moving the center of the imaging sensor away from the center of the lens' optical axis one side of the frame will be moved into the edges of the image circle that would otherwise not be used. For most lenses, image quality is greatest at the center of the image circle and poorest at the edges. There are exceptions with regard to some aberrations for some lenses, but for the most part this is generally true. Vignetting, for instance, would be affected by shift movements. The direction towards which the sensor moves would become darker on that edge. The opposite edge that the sensor moves away from would be brighter.

From the promotional materials linked in the question, we are told the IBIS system will only be active when used with Fuji lenses that do not have in-lens stabilization.

The correction is based on five axes (up and down / right and left pitch, yaw angle and optical axis rotation) to achieve more than five-stop (up to the equivalent of 5.5 stops* ) image stabilization when the camera is fitted with any FUJINON lens that does not feature the optical image stabilization functionality.

So it appears that either the XF lenses with stabilization will not be compatible with the new IBIS system or, at the very least, Fuji has decided to disable IBIS when lens-based IS is available. One would assume that this is because Fuji considers the lens based IS to be better than their IBIS in some way.

This is an indication that either:

  • The lens based IS without IBIS is more effective than IBIS without lens based IS. At least for these lenses.

OR

  • The reduction in image quality due to tilt of lens-based IS is less significant than the reduction in image quality due to the shift of the new Fuji IBIS.

OR

  • There is a fundamental way in which the lens based IS systems in Fuji's XF-mount lenses with IS are implemented that makes them incompatible with the new Fuji IBIS system introduced with the Fuji XH-1

In either case, now that Fuji has introduced a camera with IBIS, they are backstepping a bit away from their earlier statement that the XF-mount can not accommodate IBIS without some loss in image quality.

It seems to mostly be marketing doublespeak.

When Fuji didn't offer a camera with IBIS (In-Body Image Stabilization), the negative effect on image quality, however minute it might be, was emphasized as being significant enough to affect the user experience.

Now that Fuji offers a model with IBIS, the same effect on IQ is, according to Fuji, no longer significant.

  • When the image sensor is tilted in any way away from exactly perpendicular to the center of the lens' optical axis there will be a slight reduction in image quality in terms of overall sharpness. Technically, wouldn't it be less a reduction in sharpness, as much as it a change in the orientation and angle of the focus plane according to Scheimpflug (i.e., the slight tilt in IBIS is the same as film plane tilt in a view camera)? When shooting against a test target on a flat wall, this will show up as loss of sharpness gradually across the image (assuming focus point is the center). – scottbb Mar 11 '18 at 12:43
  • Why the downvote? – scottbb Mar 11 '18 at 12:45
  • the answer is not just incorrect guessing (the true reason is explained in the interview) it does not even make sense logically. assuming the sensor really can rotate in pitch/yaw axes (as far as i know, no IBIS systems does that), the mount type does not influence the result, XF mount will not make the lens less perpendicular than m43 or sony. – szulat Mar 11 '18 at 13:18
  • 1
    According to the graphic in the linked promotional materials, pitch and yaw are two of the movements the camera uses. They also claim that the "flatness of the sensor is not compromised" but they say nothing about its alignment with the lens. Sounds like marketing doublespeak to me. Just like when they did not have IBIS they said it was impossible without sacrificing IQ. Now that they have it, somehow it can be done without sacrificing IQ. Hmm... – Michael C Mar 11 '18 at 14:54
  • @scottbb Yes, every time one moves a view camera or t/s lens one gives up some IQ. Whether it is perceptible at normal viewing sizes is debatable. It's even more the case with typical 35mm and smaller cameras where the rear lens elements are much closer to the imaging plane than with typical view cameras that have the lenses fairly far forward proportionally attached to the lens board. – Michael C Mar 11 '18 at 14:57
0

The interview was from sales&marketing, engineers may well of been slapping their foreheads and groaning.

My guess is that given a potentially perfect image, if IBIS is switched on then it will be degraded, but given a shaking camera and so potentially imperfect image, introducing IBIS will make it less bad.

So the sales guy was kind of right with his facts, he just came to the wrong conclusion. It must be difficult when everything you say is recorded and played back when you contradict yourself. So they managed to do a thing they said they could not do, to me that is a good thing.

Rob.

  • your guess might be right, but the question is not if image stabilization can be harmful in some conditions (if that's true, it can be turned off, so it's not really a disadvantage). the question was, if and why the IBIS is supposed to sacrifice image quality when implemented in Fujifilm XF mount. – szulat Mar 17 '18 at 22:05
  • if the amount the tight fit damages image quality is less than the amount that IBIS improves the image by – Rob Mar 17 '18 at 23:49

Your Answer

By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

Not the answer you're looking for? Browse other questions tagged or ask your own question.