Suggestions for testing R6/R5 IBIS?

YuengLinger

Sufficiently Pixelated
CR Pro
Dec 20, 2012
3,377
1,740
USA
To get straight to the point, I just have to get the doubts out of my mind. We've waited a long time for IBIS. Going through these tests will hopefully show me that everything is fine with the R6. If not, then I just report to Canon and hope others will too!

Cheers!
 

Mt Spokane Photography

I post too Much on Here!!
CR Pro
Mar 25, 2011
16,534
1,487
When using a mechanical shutter, it is known that there are certain shutter speeds where a resonance softens images slightly. I think its a slower shutter speed though. With electronic shutter, its high speeds that can cause issues. That's covered in the manual and is a issue with very fast lenses as I recall. There was quite a bit of discussion and testing in the Fred Miranda forum but it was several weeks ago. The answer was to use electronic shutter for slower shutter speeds, mechanical for higher speeds to get the utmost performance. I've forgotten the affected shutter speeds.
 

YuengLinger

Sufficiently Pixelated
CR Pro
Dec 20, 2012
3,377
1,740
USA
Based on a careful viewing of 1139 images taken with the R6 + Rf 50mm 1.2, I am seeing only negligible differences between IBIS on and off at shutter speeds from 1/250th to 1/1000th. (And, yes, I'm seeing big help from IBIS at 1/60th to 1/125th.)

I was standing four feet from the target. All shots at f/5.6.

Using only EFCS, I took four main sets of shots for shutter speeds from 1/60th to 1/2000th: 1) IS on, Servo AF 2) IS off, Servo AF 3) IS on, One Shot AF 4) IS off, One Shot AF, on a tripod

Within each main set of shots, I took four subsets of shots: Single, Low Speed, High Speed, and High Speed+. I was curious about whether IBIS was behaving differently when mashing the shutter button for each shot than for holding it down for bursts. I saw no significant differences. The few I did see, and they were very subtle, suggest that IBIS, for my own brand of hand shake, seems to improve slightly near the end of a burst, but I did not take nearly enough shots to be sure, and, in my opinion, this has to do with my technique (or lack of!).

How many burst shots did I take? Good question. Of course with Single and Low Speed, I could just count to five. With High Speed I had to time it, and I was probably getting 5-6 shots most bursts. With High Speed+, no way could I do anything but guess that I had passed five shots--and then try to let go asap.

Since I don't see any significant difference in any of the drive modes, and since I can't find a way in DPP to sort by drive mode, I'm not going to take the time to use ratings or colors, etc. for this. If I had seen differences, I'd make the effort.

One thing I've confirmed regarding my own style of shooting: When I am standing still, and the subject is still, I do better with Servo AF enabled. I don't think this has anything to do with IBIS. One reason I plunged into mirrorless after buying the R was because of how amazing Servo AF is for very shallow depth of field, for consistently nailing the sharpest shots I've ever gotten with handheld, and I just assumed it was because I tend to sway a bit as I shoot, and the Servo AF was making micro compensations. Any thoughts?

So where does this leave me? I am definitely going to take some shots with an RF zoom that has lens IS. I've done a lot of shots with one lens, but not nearly enough for statisticians. If I had seen any kind of a trend with the prime, I'd plan to do the exact same sets with an Rf zoom, but now I'm balancing what I might learn against work and family life. (Very busy, as are many CR members, I imagine, with things other than the gear aspects of life!)

I'm planning to use the Rf 70-200mm at a longer focal length next. Should I just go with 200mm because it needs more stabilization than shorter focal lengths? I don't really have time, unless I see an issue, to try at various focal lengths. And I also plan to discard two main sets: 3) IS on, One Shot, and 4) IS Off, One Shot, on a tripod. This is because what I really want to see is how IBIS is working with handheld shots, and, as mentioned above, I just leave AF Servo enabled when handholding. On a tripod, I go with one shot, and I turn off IBIS. Just habit.

I do have three ef lenses to choose from: ef 100mm 2.8L IS, ef 35mm f/1.4 II, and ef 135mm f/2.

I've learned a few things so far for sure: Don't jump to conclusions, but be open minded when somebody is describing an issue. Also, it is very time consuming to do proper tests, and they take some careful thought to set up. And presenting the findings in a clear, convincing way is also a challenge.

And what do I do with all my sample shots? Cherry-pick a few to show my target? As there is, to my eyes, no significant differences with the rf 50mm f/1.2, the shots reveal nothing beyond what I've stated here.

Attached are just a couple as examples, both taken at 1/400th, what I was apparently imagining to be in the problematic range. :oops:

Now I understand this is a bit of an anticlimax, and some here are going to want full and better tests for more lenses, but anybody with an R6 (or an R5) would learn a lot by setting up their own tests! And I welcome any suggestions regarding streamlining!

I'll see what happens with the zoom with lens IS. And I will also do one of the ef lenses.
 

Attachments

Last edited:
  • Like
Reactions: jd7 and Joules

YuengLinger

Sufficiently Pixelated
CR Pro
Dec 20, 2012
3,377
1,740
USA
Here's the worksheet I used for the 50mm to make sure I took all the shots intended. I was surprised too that it was over 1000! The colors referred to at the bottom were what I used in LR CC to group according to my "main sets."

Today I'll take more shots at 200mm (but not as many!!!), and will try tomorrow to take shots with the ef 35mm 1.4L II, a lens that is so wonderful and so well balanced on the R6 that I imagine other photographers will be keeping it with mirrorless for years to come. I intend to!

And I MIGHT take a few shots with the ef 100mm 2.8L IS Macro. It is the only ef lens I have with IS.

From now on, much greater respect to the reviewers and those here who go through the time and effort to provide test results!

In the meantime, have some pretzels. (R6 + ef 35mm 1.4L II, ISO1250, f/5.6, 1/80th sec, IBIS definitely on).

Cheers!
 

Attachments

Last edited:

snappy604

EOS RP
CR Pro
Jan 25, 2017
480
318
Wouldn't it be loverly if DPP shows whether we had IBIS on or off? :rolleyes:

was about to ask and ask about results and about whether there was a way to tell if IBIS was or wasn't on.

I was taking a sunset pano of the city today on a tripod and noticing some blur on random images. using R5, firmware 1.2 and I think latest firmware on my RF24-70IS 2.8L. I believe I had manual focus for all tests and I.S (and therefore IBIS I think) off. Even tried variations where I used mechanical and electric shutter.... even tried to account for maybe my strap was moving things so I'd let it rest a second or two and I also used the live view touch screen to take the picture to reduce movement. In same series of settings for the multiple pano shots some sharp and others seemed to have a softness and some had a more blur....

I was using 1/8th of second shutter speeds but been able to do much lower speeds without issues on a tripod.

not sure how that can happen on a tripod with autofocus off and I think IS / IBIS off.. again this isn't hte first time I'm getting weird results and this is with a fully RF set up!

I don't recall seeing that prior to 1.2 firmware but I didn't have the camera long before updating to 1.2 so weird.

I can attach 2 RAWs but not sure how to upload those files here... taken split second after each other... one more or less sharp (could be better) and the one right after is a WTF fuzzy. No wind or movement.
 
Last edited:
Feb 15, 2020
398
285
was about to ask and ask about results and about whether there was a way to tell if IBIS was or wasn't on.

I was taking a sunset pano of the city today on a tripod and noticing some blur on random images. using R5, firmware 1.2 and I think latest firmware on my RF24-70IS 2.8L. I believe I had manual focus for all tests and I.S (and therefore IBIS I think) off. Even tried variations where I used mechanical and electric shutter.... even tried to account for maybe my strap was moving things so I'd let it rest a second or two and I also used the live view touch screen to take the picture to reduce movement. In same series of settings for the multiple pano shots some sharp and others seemed to have a softness and some had a more blur....

I was using 1/8th of second shutter speeds but been able to do much lower speeds without issues on a tripod.

not sure how that can happen on a tripod with autofocus off and I think IS / IBIS off.. again this isn't hte first time I'm getting weird results and this is with a fully RF set up!

I don't recall seeing that prior to 1.2 firmware but I didn't have the camera long before updating to 1.2 so weird.

I can attach 2 RAWs but not sure how to upload those files here... taken split second after each other... one more or less sharp (could be better) and the one right after is a WTF fuzzy. No wind or movement.
Very strange. Did you use a countdown timer or a remote to trigger the shutter?
 

Joules

EOS R
CR Pro
Jul 16, 2017
1,359
1,602
Hamburg, Germany
nope, just a very mild touch on the live view.. works fine on my 80d, no shake.
If you want a consistent chance if getting sharp images, you absolutely have to use the 2 second timer. Especially with such slow shutter speeds.

There are a lot of things beyond the camera that can account for random blur. Is the blur uniform across the image? If so, did you stand in a bridge with cars passing by, or a wooden platform of any sorts? Was it maybe windy and your tripod not extremely solid?

Or does the blur differ in different parts of the image? In that case, is the city scape you shot across a body of water and perhaps located in warm climate?

With you touching the camera, that's of course the most likely source of blur, if we are just talking about a mild softening here.
 

YuengLinger

Sufficiently Pixelated
CR Pro
Dec 20, 2012
3,377
1,740
USA
These recent posts are reminding me of something that I did notice during my test shots. I have the camera set to allow manual focus of electronic lenses, but it's possible that with both those and with ef mechanical lenses, I have been slightly changing Focus while the shutter button is depressed. I am not sure about this, but certainly on the 50 mm 1.2, the big Focus ring is very, very easy to turn while am holding the lens in the way I hold most lenses. So I might be back to questions of technique.

Since macro photography is usually the only time I choose manual focus, I'm going to disable the option to manual focus while using autofocus.

Last night I did too sets of test shots th the RF 24-70, and today I'm going to try EF 100 2.8 IS macro.
 
  • Like
Reactions: Chris.Chapterten

YuengLinger

Sufficiently Pixelated
CR Pro
Dec 20, 2012
3,377
1,740
USA
Conclusions!

After having tested the Rf 50mm f/1.2L IS, the Rf 24-70mm f/2.8L IS, and the Ef 100mm f/2.8L IS Macro, I am NOT seeing anything unexpected with IBIS and sharp images. The IBIS is doing an excellent job with both non-IS and IS lenses, including an older Ef lens with IS.

Images get shaper at higher shutter speeds with IBIS on and IBIS off, but they are very sharp, completely as hoped, with IBIS on from 1/60th sec. through 1/2000th sec. for me.

One thing I realized is that I must be very careful to avoid inadvertently moving the focus ring on the Rf 50mm 1.2. I only had one focus-by-wire lens in the past, the ef 85mm f/1.2L II, and I rarely used it for manual focus, so I never enabled "Lens Electronic MF" on my 5DIV (when I still that body and lens). But with the much better MF features on the R bodies, I did enable it. Now I realize it carries more risks than benefits for most situations other than still-lifes, landscapes, and macro, so I'll just keep it on my custom menu for special occasions.

So, my conclusion is that, under controlled testing, IBIS is performing as expected. But I am not discounting what I and others seem to be seeing with our own eyes when reviewing images. Is it possible that when the camera is being moved quickly in real-life use the IBIS "jumps" during a shot? I have no idea, that's just a brainstorm, but I am not going to say that it can only be user error. I will say that things seem to be 100% as expected when taking a series of test shots.

Since "real-world" use situations are difficult for Canon techs to reproduce, I wouldn't expect any of this to be addressed unless many photographers are reporting a pattern of problems that can be traced to a few common factors.

Personally, I'll continue to review my images as carefully as ever, but also be more relaxed (and happy!) with my R6, and be paying more attention to my shooting techniques generally.

Attached, just as an example of my ef 100mm f/2.8L IS Macro test shots, is one image taken from about five feet back, 1/640th, f/5, ISO 800, with lens IS on. It is representative of the whole series of shots from 1/60th-1/2000th.

I hope Canon continues to offer tweaks and improvements with firmware updates, but for now, my R6 seems to have IBIS performing as expected.

PS This test image was considerably sharper on my computer. Anybody have the optimal resolution numbers for posting here? Thanks!
 

Attachments

Last edited:
  • Like
Reactions: jd7 and Joules

snappy604

EOS RP
CR Pro
Jan 25, 2017
480
318
Try a countdown timer and see if the results are the same?
[/QUOTE
If you want a consistent chance if getting sharp images, you absolutely have to use the 2 second timer. Especially with such slow shutter speeds.

There are a lot of things beyond the camera that can account for random blur. Is the blur uniform across the image? If so, did you stand in a bridge with cars passing by, or a wooden platform of any sorts? Was it maybe windy and your tripod not extremely solid?

Or does the blur differ in different parts of the image? In that case, is the city scape you shot across a body of water and perhaps located in warm climate?

With you touching the camera, that's of course the most likely source of blur, if we are just talking about a mild softening here.
I've done this hundreds of times with my 80D without these kinds of results.

Solid ground, aperture was 6.3 so its not playing with nasty shallow depth of field, 1/8th of a second is fast on a tripod... while I didn't leave it for a 2 second delay, I did let it rest and did the lightest touch on the view finder.. also tried mechanical and tried electronic shutter.. manual focus with what I think is IBIS off (can't find a way to confirm, but I know I tried to turn it off). It was not in the slightest windy. The image can confirm that as there is still water in front.

Started taking multiple images with delays in between and not moving the tripod... I've taken images in all sorts of conditions and well aware of image quality change with temperature differences (really noticeable on runways when you plane spot), but the images I'm talking about the first was sharp, without any movement... second was not.

was at first thinking maybe I was introducing shake somehow when touching it or left focus or ibis or any number of things on and most turned out but had some time so over the span of about 45 mins took about 60 pics and it was occuring at about 1/15th of a sec... seemed to have about 1 in 8 or so rate where this occurred.. hence bugging me.

Can't outright upload the RAWs here I think (would if you show me how) and due to size, I've exported the first and second, identical crop/processing .. its not great processing, just a rough preset I use to start via On1. And its not the editor I checked on the camera screen and on DPP. This is to give a rough idea of what I'm seeing. Its also not the first time... I seem to almost get better results handheld with IBIS on. Keep in mind the first image was sharp, I was only giving the lightest of touches via live view and using electronic shutter by this point (was trying to eliminate vibration)... its weird the second has more blur than the first without changing any settings or posture on the tripod.
8U4A3999.jpg
8U4A4000.jpg
 

snappy604

EOS RP
CR Pro
Jan 25, 2017
480
318
Earthquake tremors? Whoa!

Any chance your balcony is picking up vibrations? That's horrible.
Looks like an earthquake, but its on solid ground, not a balcony.. . I do this a lot with 80D and don't see that kind of result. I only had the R5 for a couple of weeks prior to upgrading the firmware to 1.2 and don't recall seeing those results, but hadn't tried tripod work.
 

snappy604

EOS RP
CR Pro
Jan 25, 2017
480
318
Have you contacted Canon support yet? Sent these images?

nope.. never had to contact them. certainly thinking on it, but not sure if there is additional info I should gather.. right now I think I'm eliminating most variables and I'm trying to see if others having similar issues before going down that path. Other times the damn things takes wonderful shots.... so peeved.
 

YuengLinger

Sufficiently Pixelated
CR Pro
Dec 20, 2012
3,377
1,740
USA
nope.. never had to contact them. certainly thinking on it, but not sure if there is additional info I should gather.. right now I think I'm eliminating most variables and I'm trying to see if others having similar issues before going down that path. Other times the damn things takes wonderful shots.... so peeved.
I think you have to try with a delayed or remote release. Canon is likely to say this too, regardless of your good luck with the 80D. It's just standard practice at longer shutter speeds to use a delay or remote!
 

snappy604

EOS RP
CR Pro
Jan 25, 2017
480
318
I think you have to try with a delayed or remote release. Canon is likely to say this too, regardless of your good luck with the 80D. It's just standard practice at longer shutter speeds to use a delay or remote!
its kind of why I don't want to call them.. call centres primary purpose is to avoid blame and often will blame a hundred things first... it always a painful and irritating experience. I've had success on these kinds of settings and even slower with 80D, 7D, point and shoots, phones... so needing to use a timer is... silly.. but it's also why I was trying electronic to reduce change of vibration. Pre-firmware I was (with same combo with IS and IBIS) able to handhold and get clear shots at 1/4th of a second (attached).. this is why I don't think it's hardware.. I just had not tried tripod work with IS/IBIS disabled pre-firmware
8U4A0700.jpg
so really puzzled by all this.

its why I wanted to see what others found... and I also wish I could roll firmware back (like another person mentioned).. I've worked in IT for 30+ yrs and often find in-intended bugs in firmware upgrades. Its quite possible and also acceptable to me as long as I know it will be corrected in near future (within a few months)

what I am suspecting is that its not properly shutting down IBIS even though it is supposed to.
 
Last edited:

Joules

EOS R
CR Pro
Jul 16, 2017
1,359
1,602
Hamburg, Germany
I've had success on these kinds of settings and even slower with 80D, 7D, point and shoots, phones...
I'm not trying to imply that the issue is defintively user error. But not seeing as much impact from potential sources of camera shake (touching the camera, shutter shock, resonant frequencies due to these two or wind, ...) on a 80D or 7D compared to the R5 is to be expected. After all, the latter has almost or more than double the resolution of these older models, and also features a much finer low pass (AA) filter.

What you show in your two images does look like some pretty hefty camera shake though. Looking at the light point sources, it looks like motion blur, almost exclusively in the vertical. If you hadn't also encountered this issue with the electronic shutter, I would have guessed shutter shock.

If you want to rule out a potential variable for future assesment of the issue, I would still recommend using the 2 second timer. I do it all the time on my 80D and am sure it helps.
 

YuengLinger

Sufficiently Pixelated
CR Pro
Dec 20, 2012
3,377
1,740
USA
its kind of why I don't want to call them.. call centres primary purpose is to avoid blame and often will blame a hundred things first... it always a painful and irritating experience. I've had success on these kinds of settings and even slower with 80D, 7D, point and shoots, phones... so needing to use a timer is... silly.. but it's also why I was trying electronic to reduce change of vibration. Pre-firmware I was (with same combo with IS and IBIS) able to handhold and get clear shots at 1/4th of a second (attached).. this is why I don't think it's hardware.. I just had not tried tripod work with IS/IBIS disabled pre-firmwareView attachment 194296 so really puzzled by all this.

its why I wanted to see what others found... and I also wish I could roll firmware back (like another person mentioned).. I've worked in IT for 30+ yrs and often find in-intended bugs in firmware upgrades. Its quite possible and also acceptable to me as long as I know it will be corrected in near future (within a few months)

what I am suspecting is that its not properly shutting down IBIS even though it is supposed to.
If I were in your shoes and found that using a delay or remote resolved the issue, I would feel very relieved. If it doesn't resolve the issue, then it does sound like something Canon should look at. I will try exactly what you are doing with mine later. Too busy during the day, so this evening I can get the same basic lighting in our Suburban neighborhood.
 

snappy604

EOS RP
CR Pro
Jan 25, 2017
480
318
I'm not trying to imply that the issue is defintively user error. But not seeing as much impact from potential sources of camera shake (touching the camera, shutter shock, resonant frequencies due to these two or wind, ...) on a 80D or 7D compared to the R5 is to be expected. After all, the latter has almost or more than double the resolution of these older models, and also features a much finer low pass (AA) filter.

What you show in your two images does look like some pretty hefty camera shake though. Looking at the light point sources, it looks like motion blur, almost exclusively in the vertical. If you hadn't also encountered this issue with the electronic shutter, I would have guessed shutter shock.

If you want to rule out a potential variable for future assesment of the issue, I would still recommend using the 2 second timer. I do it all the time on my 80D and am sure it helps.

yep I will probably try quite a few more variables.. mostly I'm trying to see if others are having any issues.. also why I was interested in this thread. I noticed it in daylight as well when shooting bald eagles... 1/400th of a second but wasn't sure at time if it was because I was panning by hand with a 3rd party lens (sigma)... there are times where it just seems to decide it doesn't want to cooperate, but haven't found a trigger I can reproduce and sadly not a lot of cycles to do so.. a proper test is time consuming :) but yes I'll likely do 2 sec pause.. I've used many times in past but this was the barest of touches on the liveview screen on a tripod.. hence the concern, but yep 45 MP vs 18MP can produce other results, but that seems drastic... more so given how much success I had pre-firmware handheld at even slower speeds.
 

snappy604

EOS RP
CR Pro
Jan 25, 2017
480
318
If I were in your shoes and found that using a delay or remote resolved the issue, I would feel very relieved. If it doesn't resolve the issue, then it does sound like something Canon should look at. I will try exactly what you are doing with mine later. Too busy during the day, so this evening I can get the same basic lighting in our Suburban neighborhood.
cool, more or less looking for feedback from others if they've seen similar. If you have the time it'd be great, but not expected :) just posted because I saw your concerns as well, lucky your tests showed otherwise. I might just get annoyed enough to stay on phone with them... might. My day is incredibly busy usually and prefer to spend my free time out actually taking pics :)

there is a lot of new tech and now the sensor does move thanks to IBIS and it's first gen IBIS, but it seemed fine pre-firmware. Wish I'd taken more time or not upgraded :) so I could rule it in or out.. intermittent issues are really the most irritating ones to troubleshoot (as you saw I also had clear images in prior by about a second)