Strange case of blur images. Canon RF 24-70 f2.8

jd7

CR Pro
Feb 3, 2013
1,064
418
Without wishing to take this thread off topic ...

In recent months I have been trying out live view with face tracking AF on my 6DII, for portraits with fast prime lenses. When it works well, the AF is very accurate even at wide apertures, and I do see the potential advantage in that regard over traditional DSLR AF. However, I get a surprising number of blurry photos, and they remind me of the sample blurry image YuengLinger nitsuga has posted in this thread. The photos don't seem have the plane of focus in the wrong place, they seem to be blurry all over. I thought at first it must be user error, but it has kept happening despite me being more careful about holding the camera stable. Now I'm wondering if I'm seeing the same issue that this thread is about?

Of course, I'm not using an R5 or R6 and there is no IBIS on my 6DII, so maybe my issue is completely different.

If relevant, I am seeing the issue with Canon lenses and Sigma lenses, with my zooms as well as my fast primes, and with lenses with and without IS.

Below are three examples, taken with a Canon 70-200/2.8L IS II, Sigma 35 Art and Canon 24-70/4L IS. As mentioned above, not all shots I take with live view are like this (some are great), but I get patches where I seem to get a group of blurry shots like these ones.
 

Attachments

  • IMG_1381.jpg
    IMG_1381.jpg
    348 KB · Views: 209
  • IMG_1406.jpg
    IMG_1406.jpg
    484.5 KB · Views: 195
  • IMG_1638.jpg
    IMG_1638.jpg
    231.9 KB · Views: 195
Last edited:
Upvote 0

Joules

doom
CR Pro
Jul 16, 2017
1,801
2,247
Hamburg, Germany
Of course, I'm not using an R5 or R6 and there is no IBIS on my 6DII, so maybe my issue is completely different.
Your problem may be related in the sense that it could have to do with shutter shock. Using LiveView, I guess you are holding the camera differently than when using the OVF, right? Which may cause the forces acting on the shutter to impact the image.

I believe silent mode in the 6D II uses electronic first curtain shutter? If so, you could try that and see if the problem persists. It shouldn't, if my guess is correct.

Regardless, only shutter speeds of 1/100, but not lower, being affected by the issue in this thread, it seems likely the IBIS is at least in part to blame. And your 6D II doesn't have that, so it is a different issue.

*alternative guess: Make sure to use focus priority if you do not need the highest possible FPS.
 
Upvote 0

jd7

CR Pro
Feb 3, 2013
1,064
418
Your problem may be related in the sense that it could have to do with shutter shock. Using LiveView, I guess you are holding the camera differently than when using the OVF, right? Which may cause the forces acting on the shutter to impact the image.

I believe silent mode in the 6D II uses electronic first curtain shutter? If so, you could try that and see if the problem persists. It shouldn't, if my guess is correct.

Regardless, only shutter speeds of 1/100, but not lower, being affected by the issue in this thread, it seems likely the IBIS is at least in part to blame. And your 6D II doesn't have that, so it is a different issue.

*alternative guess: Make sure to use focus priority if you do not need the highest possible FPS.
Thanks Joules. I will try silent mode as soon as I get chance. I will also check whether I have the camera set on release priority or focus priority.

And yes, using LiveView I do hold the camera differently compared with when using the OVF. My initial thought was that was the reason I was getting blurry images, ie just not holding the camera steady enough. However, on more recent occasions I have tried to be very careful about holding the camera steady and I have still had the issue. I guess it's possible I'm still not holding it steady enough and the answer is as simple as that, but I would be a bit surprised if that is the explanation.

It's frustrating because when it works well, LiveView gives me images which are perfectly in focus.
 
Last edited:
  • Like
Reactions: 1 user
Upvote 0

Joules

doom
CR Pro
Jul 16, 2017
1,801
2,247
Hamburg, Germany
However, on more recent occasions I have tried to be very careful about holding the camera steady and I have still had the issue.
As far as I'm aware, in LiveView, the shutter closes, before then opening in order to take the picture and closing again to stop the exposure. With EFCS, this initial closing is avoided and hence, no shake is introduced before the image capture begins.

I don't think just gripping the camera more firmly is sufficient to deal with this kind of shake.

Take a look at this jn depth article, if you want to know more:

 
  • Like
Reactions: 1 users
Upvote 0

YuengLinger

Print the ones you love.
CR Pro
Dec 20, 2012
3,751
2,269
USA
Without wishing to take this thread off topic ...

In recent months I have been trying out live view with face tracking AF on my 6DII, for portraits with fast prime lenses. When it works well, the AF is very accurate even at wide apertures, and I do see the potential advantage in that regard over traditional DSLR AF. However, I get a surprising number of blurry photos, and they remind me of the sample blurry image YuengLinger has posted in this thread...
I can't take credit for any photos posted here yet. By tonight I will post some shots showing mechanical shutter vs electronic first curtain, stabilisation on and stabilization off. Got to go!
 
Upvote 0

jd7

CR Pro
Feb 3, 2013
1,064
418
I can't take credit for any photos posted here yet. By tonight I will post some shots showing mechanical shutter vs electronic first curtain, stabilisation on and stabilization off. Got to go!
Ah, apologies, the photos I was referring to were actually posted by nitsuga. I should have paid more attention.
 
Upvote 0

Joules

doom
CR Pro
Jul 16, 2017
1,801
2,247
Hamburg, Germany
Below are three examples, taken with a Canon 70-200/2.8L IS II, Sigma 35 Art and Canon 24-70/4L IS.
I missed those images previously. They look like defocus blur to me, not shake. Especially the man with the baby, his shoulder is much sharper than the more distant parts of the image.

Have you made any adjustments to AFMA in the camera or perhaps with the dock in case of the sigma lenses?
 
Upvote 0

jd7

CR Pro
Feb 3, 2013
1,064
418
I missed those images previously. They look like defocus blur to me, not shake. Especially the man with the baby, his shoulder is much sharper than the more distant parts of the image.

Have you made any adjustments to AFMA in the camera or perhaps with the dock in case of the sigma lenses?
I know I haven't made any changes via the Sigma dock, but I'd have to check AFMA adjustments ... but I didn't think AFMA was relevant when shooting in LiveView?
 
Upvote 0

YuengLinger

Print the ones you love.
CR Pro
Dec 20, 2012
3,751
2,269
USA
I've attached four shots taken at 1/100th of a second, f/2.8, ISO 640. Window light. The file names indicate which ones were taken with EFCS or Mechanical, IBIS or no IBIS.

They were all taken with an Rf 24-70mm f/2.8L IS at 35mm. I found the same results at 24mm, but didn't test at a longer Focal Length. I also found similar results, but not as pronounced using an ef 35mm f/1.4L at the same settings.

I took the same shots at 1/200th and found that, at 100%, the shutter shock was barely detectable, not a real world factor, but if I zoomed in to 200%, I could see the tiniest bit of difference. Really not any issue, imo.

I did the same tests at 1/30th of a second (ISO 160), and I've attached two of those to show that there is still visible shutter shock, though not as bad as at 1/100th. Both were taken with NO IBIS because the results were pretty much the same with it on, and because I think most photographers at 1/30th are going to choose no IBIS on a tripod. I do know that 1/30th is in the range dSLR shooters would use mirror lock-up, but I will admit I thought the blur at around 1/30th in dSLR's was caused by the mirror. I guess the mechanical shutter is involved there too.

For myself, I have to put some perspective on this. The R5 shows shutter shock at 1/100th of a second with mechanical shutter. This is aggravating because I tend to shoot quite a lot indoors from 1/60th to 1/160th, and I'm USED to mech shutter, I trust it. But I have been taking photos under LED with EFCS, and I'm not seeing any drawbacks in my images. So all I have to do is remember to switch to mechanical outdoors, bright light, wide apertures to avoid the bokeh problems. Am I missing any EFCS drawbacks here that might bite me someday?

Another perspective that helps me deal with a less than absolutely perfect camera, I know that shutter shock has been a quiet issue for decades, and that we learned to work around it when shooting in that slower shutter-speed danger zone. I can still work around it easily here, but I just have to remember which Shutter Mode to use in various situations. I think for the most part, as a portrait and events photographer, I can use EFCS most of the time. Bright sunshine and higher shutter speeds, go to mechanical.

Can this be fixed with a firmware update? If not, would Canon make some change in production to improve the vibration issue? If so, where would that leave us early adopters?

(I'm really getting behind on yardwork this week, and it's only Tuesday! Cheers!)
 

Attachments

  • EFCS and IBIS.jpg
    EFCS and IBIS.jpg
    1.4 MB · Views: 209
  • Mech and IBIS.jpg
    Mech and IBIS.jpg
    1.2 MB · Views: 204
  • EFCS no IBIS.jpg
    EFCS no IBIS.jpg
    1.3 MB · Views: 215
  • Mech no IBIS.jpg
    Mech no IBIS.jpg
    1.2 MB · Views: 206
  • 30th EFCS.jpg
    30th EFCS.jpg
    1.1 MB · Views: 195
  • 30th Mech.jpg
    30th Mech.jpg
    1 MB · Views: 208
Last edited:
Upvote 0

nitsuga

Canon R5
Nov 3, 2020
13
7
I've attached four shots taken at 1/100th of a second, f/2.8, ISO 640. Window light. The file names indicate which ones were taken with EFCS or Mechanical, IBIS or no IBIS.

They were all taken with an Rf 24-70mm f/2.8L IS at 35mm. I found the same results at 24mm, but didn't test at a longer Focal Length. I also found similar results, but not as pronounced using an ef 35mm f/1.4L at the same settings.

I took the same shots at 1/200th and found that, at 100%, the shutter shock was barely detectable, not a real world factor, but if I zoomed in to 200%, I could see the tiniest bit of difference. Really not any issue, imo.

I did the same tests at 1/30th of a second (ISO 160), and I've attached two of those to show that there is still visible shutter shock, though not as bad as at 1/100th. Both were taken with NO IBIS because the results were pretty much the same with it on, and because I think most photographers at 1/30th are going to choose no IBIS on a tripod. I do know that 1/30th is in the range dSLR shooters would use mirror lock-up, but I will admit I thought the blur at around 1/30th in dSLR's was caused by the mirror. I guess the mechanical shutter is involved there too.

For myself, I have to put some perspective on this. The R5 shows shutter shock at 1/100th of a second with mechanical shutter. This is aggravating because I tend to shoot quite a lot indoors from 1/60th to 1/160th, and I'm USED to mech shutter, I trust it. But I have been taking photos under LED with EFCS, and I'm not seeing any drawbacks in my images. So all I have to do is remember to switch to mechanical outdoors, bright light, wide apertures to avoid the bokeh problems. Am I missing any EFCS drawbacks here that might bite me someday?

Another perspective that helps me deal with a less than absolutely perfect camera, I know that shutter shock has been a quiet issue for decades, and that we learned to work around it when shooting in that slower shutter-speed danger zone. I can still work around it easily here, but I just have to remember which Shutter Mode to use in various situations. I think for the most part, as a portrait and events photographer, I can use EFCS most of the time. Bright sunshine and higher shutter speeds, go to mechanical.

Can this be fixed with a firmware update? If not, would Canon make some change in production to improve the vibration issue? If so, where would that leave us early adopters?

(I'm really getting behind on yardwork this week, and it's only Tuesday! Cheers!)
Your test confirms what I've found. 1/100 is the speed to avoid using mechanical shutter. i can live with that...i guess
 
  • Like
Reactions: 1 user
Upvote 0
Aug 27, 2019
667
1,414
  • Like
Reactions: 1 user
Upvote 0
I think for owners of the R5 it is very important to know whether this problem affects only a small number of cameras or is a characteristic of the R5.

I found the same shutter shock issues with my R5... reported it on the forums about a month ago. I contacted Canon and let them know about it. I then requested that they add an automatic switch from EFCS to Mechanical that is shutter speed dependent. I believe Nikon and Fuji already have this feature.

I don’t think a firmware update will be able to remove the shutter shock so probably best to contact Canon to request the automatic shutter mode switching feature. It’s our best hope... just need a whole bunch of people to request this feature...
 
  • Like
Reactions: 2 users
Upvote 0

Bdbtoys

R5
CR Pro
Jul 16, 2020
467
331
So when I was looking for shutter shock I was looking for tells such as nitsuga originally shown (as actual blur, which I don't see). However looking at YuengLinger's images and ones from Ramage's link, it appears to be more of a softening of the image (or almost like an un-sharp mask). I went back thru my pictures last night as noticed this softening on my mechanical shutters around the 1/100 mark w/ IBIS on.

I'll be honest... if this wasn't pointed out to me I may have missed it since I don't normally pixel peep to that level.
 
  • Like
Reactions: 1 user
Upvote 0

koenkooi

CR Pro
Feb 25, 2015
3,574
4,109
The Netherlands
So when I was looking for shutter shock I was looking for tells such as nitsuga originally shown (as actual blur, which I don't see). However looking at YuengLinger's images and ones from Ramage's link, it appears to be more of a softening of the image (or almost like an un-sharp mask). I went back thru my pictures last night as noticed this softening on my mechanical shutters around the 1/100 mark w/ IBIS on.

I'll be honest... if this wasn't pointed out to me I may have missed it since I don't normally pixel peep to that level.

I started noticing it 2 years ago when I rented an R and shot with the 100mm L macro next to my RP with a Sigma 150mm, at 100s the antennae and probisces of butterflies showed way more motion blur with the 100mm+R than 150mm+RP. I switched the R to EFCS and it went away. Closer inspection showed motion blur across the whole picture, but the antennae where shot against a blue sky which made the blur stand out more.

I had set the R to MS because I also rented the RF50 f/1.2 and the internet at the time was scared of the bokeh-destroying EFCS :)

On my R5 the EF50 STM at 1/125s is the worst, the lens has trouble focussing and the lens weighs virtually nothing, so no mass to dampen vibrations. I keep forgetting to update C1, every other mode is set to EFCS already since I can't spot any effect on bokeh with f/1.8 lenses.
 
  • Like
Reactions: 1 users
Upvote 0

YuengLinger

Print the ones you love.
CR Pro
Dec 20, 2012
3,751
2,269
USA
I found the same shutter shock issues with my R5... reported it on the forums about a month ago. I contacted Canon and let them know about it. I then requested that they add an automatic switch from EFCS to Mechanical that is shutter speed dependent. I believe Nikon and Fuji already have this feature.

I don’t think a firmware update will be able to remove the shutter shock so probably best to contact Canon to request the automatic shutter mode switching feature. It’s our best hope... just need a whole bunch of people to request this feature...
Ok, I wrote an email to CPS describing the issue, opining that it is a typical characteristic, not a defect in certain bodies, and offering your work around. However, I also said I hoped a firmware fix could improve the issue.

I understand your doubt, though, as this is happening at the moment of the shot, and not related to active IBIS. Brainstorming, what if the sensor could be set by firmware so that when IBIS is not enabled, the sensor is kept in place in such a way as to be less prone to showing the shock? As I speculated earlier, the sensor doesn't seem to be simply locked down when IBIS is off; otherwise, why do we hear a rattle when the body is powered off? So wouldn't that mean that the IBIS function is holding the sensor tightly when IBIS is disabled?

One more thing. CPS told me that Canon has just this week started rolling out a new "Customer's Voice" program. No further info--just that it started in the USA as recently as this week.
 
  • Like
Reactions: 2 users
Upvote 0

YuengLinger

Print the ones you love.
CR Pro
Dec 20, 2012
3,751
2,269
USA
DOH! Real world pain in the butt. Just below freezing here, so decided it was a good time to knock down a very big wasp nest. Taking shots straight up in challenging early morning light, I FORGOT TO CHECK MY SHUTTER MODE! Of course I was in mechanical. And, with some distractions from my little boy, and difficulty seeing much because of glare in the EVF, I also let my shutter speed go lower than necessary, 1/200th on a 70-200 lens. So I did not get a sharp shot of the wasp nest before knocking it down.

If I hadn't done the test shots, I would have just thought all the distractions blew the shot. I know I was steady and still at the moment of capture, and I now believe this was shutter shock. DOH!
 
Last edited:
Upvote 0