R5/R6 IBIS effectiveness ratings seem to be all over the board

Nov 12, 2016
910
615
I know that rating cameras and lenses with stabilization to a certain number of stops of effectiveness is kind of a dubious pseudoscience to begin with. But there are a lot of wild claims going around about how many stops the IBIS claims to provide.

In the unveiling video, 8 stops was mentioned repeatedly, but I assumed anywhere near this would only be achieved when using a lens that also had stabilization. But this video offhandedly mentions that the 50mm f1.2 should provide 7 stops, and the 28-70mm f2 should provide 8 stops because of the "image circle" of those lenses, even though those lenses do not have IS. I assume they mean the image circle is large enough to allow the sensor a lot of room to move. (Link goes right to the spot in the video that mentions this.)


Claiming anything higher than 5 stops of stabilization with IBIS only and no IS in the lens sounds to me like something between overly ambitious and flat-out unheard of. Could the IBIS really be this good?
 

koenkooi

CR Pro
Feb 25, 2015
3,575
4,110
The Netherlands
Looking at the patents Canonnews.com has written about and interviews, Canon has been working on stabilizing different things, from obvious things like shaky hands, to non-obvious things like low-frequency movements like breathing or slow waves on a big boat.

So I can see a case where Canon adds up the stops, e.g. 4 stops high frequency shake, 4 stops low frequency shake, done, 8 stops! On the other hand, there is an actual CIPA standard (DC-011-2015) for it, so I hope they are talking about that instead of the funny math.
 
  • Like
Reactions: 1 user
Upvote 0

rosw

EOS R
Sep 13, 2012
35
8
Hi

Really confused about these 8 stop corrections

how does Canon determine which lens can "achieve" 8 stop of "anti-shake" ? (does not seems like L lens are "better")

how about third party camera lens (tamron & sigma) or even the older EF lens (both L & No-L), will they be able to achieve 8 stop?

how will we know?

please advice


from dpreview

dpreview.JPG
 
Upvote 0

koenkooi

CR Pro
Feb 25, 2015
3,575
4,110
The Netherlands
Really confused about these 8 stop corrections

how does Canon determine which lens can "achieve" 8 stop of "anti-shake" ? (does not seems like L lens are "better")
[..]
The answer is literally in the picture you posted: "Canon says the large image circle [..] allows [..] to hit this maximum figure"
 
  • Like
Reactions: 1 user
Upvote 0
Nov 12, 2016
910
615
I'll take a wild stab in the dark here - but perhaps the 50 F1.2 and 28-80 F2 still feature the shake detecting gryos built into the lenses to communicate with a [future] IBIS system even if they don't actually feature a floating IS element in the lens itself?
I highly doubt this, and I see absolutely no reason for it. The lens is affixed firmly to the camera body. Any shake or movement within the lens would also be detectable in the camera itself. If your lens is moving independently at all from the camera body, IBIS performance is the least of your worries.

Edit: Actually, after reading the blurb above from Canon, it kind of suggests this. So, maybe. I dunno. I don't see why you would need separate sensors within the lens and the camera. I would think that any shake, tilt, or movement in the lens would be detectable in the camera since they are fixed together. :unsure:
 
Upvote 0

Jack Douglas

CR for the Humour
Apr 10, 2013
6,980
2,602
Alberta, Canada
I'll take a wild stab in the dark here - but perhaps the 50 F1.2 and 28-80 F2 still feature the shake detecting gryos built into the lenses to communicate with a [future] IBIS system even if they don't actually feature a floating IS element in the lens itself?
I might as well guess too. If you think about the effect of movement the greatest sensitivity to it is going to be furthest from the sensor which seems to me is out in the lens but movements are both linear and rotational as well so ... ?

Jack
 
Upvote 0
Apr 25, 2011
2,510
1,885
I highly doubt this, and I see absolutely no reason for it. The lens is affixed firmly to the camera body. Any shake or movement within the lens would also be detectable in the camera itself. If your lens is moving independently at all from the camera body, IBIS performance is the least of your worries.

Edit: Actually, after reading the blurb above from Canon, it kind of suggests this. So, maybe. I dunno. I don't see why you would need separate sensors within the lens and the camera. I would think that any shake, tilt, or movement in the lens would be detectable in the camera since they are fixed together. :unsure:
I can speculate that if Canon uses help of linear acceleration sensors to increase the accuracy of angular movement detection, then having an additional sensor in the lens may improve pitch axis rotation recognition.
 
Upvote 0
I highly doubt this, and I see absolutely no reason for it. The lens is affixed firmly to the camera body. Any shake or movement within the lens would also be detectable in the camera itself. If your lens is moving independently at all from the camera body, IBIS performance is the least of your worries.
My theory is that the best way to measure multi axis movement is to space multiple sensors as far apart as possible.
 
Upvote 0
Mar 25, 2011
16,848
1,835
I highly doubt this, and I see absolutely no reason for it. The lens is affixed firmly to the camera body. Any shake or movement within the lens would also be detectable in the camera itself. If your lens is moving independently at all from the camera body, IBIS performance is the least of your worries.

Edit: Actually, after reading the blurb above from Canon, it kind of suggests this. So, maybe. I dunno. I don't see why you would need separate sensors within the lens and the camera. I would think that any shake, tilt, or movement in the lens would be detectable in the camera since they are fixed together. :unsure:
The shake or movement is likely magnified in the lens since you may be holding the camera body relatively still, but the lens is still moving, particularly toward the end. So there may be a combination of various motions that can be computed and nullified.
 
Upvote 0
Nov 12, 2016
910
615
hi guys,

so this 8-stop is it "auto" or do I need to manually need to adjust the "digital image stabilization" with the setting ?
The higher IBIS rating due to a larger lens circle almost sounds like some sort of digital stabilization, but that does not make much sense since you need a larger sensor to still put out a full size image.
But if the lens is capable of providing a larger image circle, and the sensor is capable of making large movements inside that large circle, doesn't that mean the whole system can compensate for more shake?
 
  • Like
Reactions: 1 users
Upvote 0

cayenne

CR Pro
Mar 28, 2012
2,868
796
Canon claims certain lens has 8 stop, while others have 7 stop,

Q1) is there any way to test / prove?


Q2) how about the older EF lens ? Can they achieve 8 stop?

It potentially sounds like it, if the EF (and maybe even other 3rd party lenses). have a large enough image circle for the sensor to move around inside of...?

cayenne
 
Upvote 0

koenkooi

CR Pro
Feb 25, 2015
3,575
4,110
The Netherlands
Canon claims certain lens has 8 stop, while others have 7 stop,

Q1) is there any way to test / prove?


Q2) how about the older EF lens ? Can they achieve 8 stop?
In the B&H Q&A panel Rudy Winston from Canon said the following:
  • The stops are measured by the CIPA standard
  • EF and RF non-IS lenses will get a menu entry for IBIS
  • EF and RF IS lenses won’t get a menu entry, IBIS is controlled by the IS switch on the lens
  • EF IS will work ‘simultaneously’ with IBIS
  • RF IS will work ‘in conjunction’ with IBIS
One of the Canon USA promo clips had a disclaimer saying any IS lens announced before July 9th will need a firmware update.

My interpretation is that for EF IS lenses, IBIS will only do rotation compensation, not shake compensation. For some select EF lenses you’ll get a firmware update to enable full IBIS, my guess would be the 2018 400mm and 600mm, since those seem to have RF style electronics already.
I hope the 100mm macro also gets a firmware update and full IBIS :)
 
  • Like
Reactions: 4 users
Upvote 0