How effective is R6's IBIS with EF Lenses?

No need to get heated! juanmaasecas is correct: The R5/R6 has a bug with non-IS lenses when IBIS is activated. And he has brought it to the attention of Canon and many R6/R5 owners. The bug involves a twisting of the IBIS which shows most clearly in the edges of photographs. This phenomenon only takes place during the FIRST shot after the camera body has powered on or been reactivated after having timed out (or "woken up from sleep").

One easy workaround is to take a burst of shots, examining the first for the phenomenon, and then using a subsequent shot if needed.

Hopefully Canon is working on this! And do note that it is only very noticeable at the slower shutter speeds--but it is EASILY reproducible by anybody with an R5/R6 and a non-IS lens.

To answer the OP, the R6's IBIS is very effective when used with EF lenses--just keep in mind the bug that has been acknowledged by Canon.

Right, the bug appears in a specific scenario. You can easily reproduce this, and see it without pixel peeping, by simply handholding at a slow shutter speed, say 1/10th to 1/20th sec., take several shots. Compare the first shot to subsequent shots in the series.  It isn't only the edges, but that's where the blur shows most because of the twisting motion of the IBIS.

As for using IBIS set to "Only for Shot," in my opinion, that setting sacrifices one of the best aspects of IBIS--keeping the EVF image stable while focusing and composing.

The motion blur caused by the problem does show up at higher shutter speeds, just not as much, eventually getting into pixel-peeping territory.

I mentioned my workaround--just take a short burst of shots. And I've been doing that for years with the 5D series of bodies. Another workaround is to give up great non-IS primes, such as the very Rf lenses which prompted me to switch to Canon mirrorless.

Deal breaker? No. Identifiable bug that Canon should fix? Yes!

How could anybody ever test IBIS without handholding? And if the phenomenon is consistently reproduced among a large number of shooters, that seems fairly strong evidence.

The problem with testing on a tripod is that the R5/R6 IBIS "senses" when it is on a tripod within a few seconds, and adapts accordingly. (Though I always turn IBIS off when using a tripod, maybe just because I always have.)

You don't have to turn the camera off and on. You can wait a few moments when ECO MODE is on for the camera to time out. This also means that the problem is not limited to just when the camera is first powered on, but happens frequently during normal usage.

We agree: It is a bug that should be fixed. Does it detract from the overall amazing capabilities of the R5/R6? Each buyer/owner must decide.

Would Canon create an advertising campaign built around the promise: "World Class IBIS after the first shot!"? Of course not! They'd rather just fix the bug!

And, once more, thank you, juanmaasecas, for finding the pattern of the mysterious, occasional blurring of images that were expected to be sharp! I remember scratching my head at some soft images taken with my ef 35mm f/1.4L II, used in low-light indoors, then testing on a tripod, finding no problem. Until you demonstrated how the issue affects the first shot in a series, I assumed that I was jerking or jabbing or there was just some random IBIS problem.

No matter how sure we are that our technique is ok, a few visits to forums can easily convince us that we MUST be at fault, never the hardware. (Understandable, though, as there are more than a few influencers or cranks who just make stuff up!) So applause is due when a photographer keeps a clear head, a civil tongue, and then patiently explains how to demonstrate a bug. The best and sometimes ONLY way to get something fixed.

John Sheehy wrote:

1Dx4me wrote:

bodeswell wrote:

ArtHeals wrote:

I use a couple of Canon EF portrait lenses (135/2L & 85/1.8 USM) with EOS R and I don't have any major problems as i keep the shutter speed around 1/200 most of the time. But I was wondering how effective is the IBIS on R6 with these EF lenses. Is it still 5-axis with EF lenses or downgrades to 3-axis?

Some used R6's are popping up here and there and I am contemplating if its going to benefit my STILL images in any way - (No interest in video)?

Thanks in advance

I have an R5 and a 135/2L. It works great with the IBIS on. In fact, I haven't noticed any problems with IBIS using any EF lens.

yeah, that is what i am expecting. i have no plane to acquire RF lenses while i have a plethora of EF "L" lenses.

For lenses that don't have any stabilization, IBIS, even if not perfect, can allow much slower shutter speeds than without IBIS, but that does not necessarily contradict the idea that IBIS added to an IS lens can actually degrade stability slightly. So, for lenses without IS, IBIS is very helpful for slow shutter speeds, but may be even better if you could turn roll and/or X/Y off in some situations.

well, John, speaking for myself, half of my lenses are without IS, such as canon 24-70mm f2.8 II that i use quite a bit, and i couldn't careless if it doesn't have IS! long ago, i decided to work on my lens handholding techniques instead of being in the mercy of IS and i am very happy with the success i have attained. IBIS is a fine concept and i am sure it helps many but as i said, i couldn't careless. i have a couple zeiss lenses that i use in manual focus mode for static subjects with no IS but end up with great, sharp images.

the best.

I have noticed a similar effect with a couple of Powershot cameras. With IS enabled, it seems that the first shot of a series is more prone to one side or corner being inexplicably soft. The cameras in question are the S95 and G1X original. These cameras only have lens-based IS, but I believe that the softness could be caused by the IS element being at the extremity of its travel at the moment of exposure. It seems more prevalent at longer focal lengths, where presumably, the element has to move over greater distances.

I haven't tested exhaustively, but the effect seems more likely to happen when using 'continuous' rather than 'shoot only' IS, and due to the way that compact cameras tend to be used, the first shot of a series has usually been the first shot after turning the camera on.

The series are usually either focus or exposure brackets. The three shots are triggered with a single press of the shutter button.

While 8-year old compacts with lens-based IS may seem to be far removed from a top-end mirrorless camera with IBIS, the commonality is that their IS systems are both controlled by firmware written by Canon, so I can't help thinking the problems may be related.

To try to avoid the problem, I have started doing the following:-

1) use shoot-only IS rather than continuous

2) use a 1s self timer to avoid motion from the shutter button press on the first shot

3) turn IS off when not needed.

These measures seem to be helping, but I don't have enough to go off as yet to be sure.

If there is something special about the first shot after turning the camera on, then another measure would be to simply fire a single shot before doing the bracket sequence.

Anyway, this offers no solutions for the problem discussed, but I was struck by the similarities. Instances where stabilisation systems may actually degrade image quality are seldom discussed.

DJaye wrote:

ArtHeals wrote:

I use a couple of Canon EF portrait lenses (135/2L & 85/1.8 USM) with EOS R and I don't have any major problems as i keep the shutter speed around 1/200 most of the time. But I was wondering how effective is the IBIS on R6 with these EF lenses. Is it still 5-axis with EF lenses or downgrades to 3-axis?

Some used R6's are popping up here and there and I am contemplating if its going to benefit my STILL images in any way - (No interest in video)?

Thanks in advance

I just made the switch to R6 from 5D4. I find IS working better than with my 5D including Sigma 150-600 and 70-300

R5/R6 IBIS only adds roll and X/Y translation correction to IS lenses.  If you seem to be getting better pitch/yaw stability, that may be due to the loss of the large slapping FF mirror, rather than the added IBIS.  "Roll" correction has nothing to do with magnification or focal length, and is for when you rotate the camera on the lens' axis during a long exposure time, not something that you're likely to benefit from with the relatively short exposures that many people would use at 600mm.  X/Y translation has nothing to do with focal length, but depends on how close you are to the subject, and the closer you are, and the longer the exposure (absolutely; not relative to the focal length), the more you are likely to benefit from it.  These 3 axes have questionable value for shooting telephoto at a distance.

You also now have larger pixels, so when you inspect at 100% pixel view, you are less likely to see the same amount of blur, because you're magnifying less.  For that reason, R6 users may be happier with the added IBIS than R5 users in general (and people with monitors with smaller pixels, as well), even if it turns out that as some of us suspect, that these roll and X/Y IBIS implementations actually add a little bit of blur if they are not actually needed.  In my experience with my R5, there seems to be waves of added blur that come in any sequence of images, about 1/2 pixel in size on average with the R5.  Out of about every 3 images, one will have less than 0.1 pixels of blur (good), one about 1/4 (OK), and one about 1/2 pixel (loss of fine detail), over a wide range of shutter speeds, which would be rock solid with my previous APS-C DSLRs (90D and 7D2), despite higher magnification at 100% pixel view with the 90D.  Unfortunately, such added blur can still make the system qualify for the claimed stops of stability by the CIPA standard, which is based on postcard-sized image viewing at 450mm, so Canon may never feel compelled to address the issue, even if it slightly ruins more possible otherwise-perfect keepers.  I have no confidence that my R5 will nail stability with a single shot with IBIS/IS enabled, unless the shutter speed is 1/1600 or faster, so even for non-action, I take a burst of at least 3 shots if the shutter speed is not fairly high, which creates a lot of extra culling work, and wears the shutter unnecessarily when using mechanical shutter or electronic first curtain.  With full-electronic shutter, the overhead is still there in culling time, even if I can get more near-perfect-stability keepers per second.