fredmiranda.com Login
This topic is in the 2015 archive. You can read it and like its posts here. It is read only.

Archive 2015a7R II IBIS effectiveness

  
 
Matt Grum
Offline
Details
p.2 #1a7R II IBIS effectiveness
exdeejjjaaaa wrote:
m43 users have perfectly working IBIS for 7mm (=14mm FF eq FOV)... like in 7-14mm zooms there


No they don't - it may be good enough, but not perfect. All mechanical stabilisation systems suffer from exactly the same problem. It's the nature of projective geometry.

alundeb wrote:
Thanks, Matt. I hadn't thought of that. I will also keep my expensive Canon IS lenses


The problem is also not unique to sensor-shift stabilisation it also affects lens based IS.
Sep 03, 2015 at 01:58 PM
philber
Offline
Details
p.2 #2a7R II IBIS effectiveness
I feel I get more stops with a longer lens than with shorter ones. With my ZM 35 f:1.4, I get around 2 stops, possibly 2.1/2 to 3 with my Otus 55, and, to my surprise, definitely more than 3 with a 135APO. I get some sharp shots around 1/15s with that one, and my hands aren't exactly steady
Sep 03, 2015 at 02:26 PM
Big Cheese
Offline
Details
p.2 #3a7R II IBIS effectiveness
Matt Grum wrote:
exdeejjjaaaa wrotem43 users have perfectly working IBIS for 7mm (=14mm FF eq FOV)... like in 7-14mm zooms there
No they don't - it may be good enough, but not perfect. All mechanical stabilisation systems suffer from exactly the same problem. It's the nature of projective geometry.

alundeb wroteThanks, Matt. I hadn't thought of that. I will also keep my expensive Canon IS lenses
The problem is also not unique to sensor-shift stabilisation it also affects lens based IS.


I am not sure what this adds. Of course there is no perfect IS system. That is, except a tripod. With IS if it is good enough then it is good enough. I've never noticed degradation in the corners of wideangle lenses due to ineffective IS as being a particular problem.
Sep 03, 2015 at 11:39 PM
dass101
Offline
Details
p.2 #4a7R II IBIS effectiveness
I read the Cameralabs review and what I took from regarding IBIS is that with the high resolution the benchmark has also moved. So while we're talking 3-4 stops improvement is compared to what? To 1/focal? Or to best razor sharp? What I'm saying is maybe the camera provides 3 stops over razor sharp, which translates to 2 stops over 1/f, because at this resolution you can't consistently get razor sharp at 1/f.
Sep 04, 2015 at 12:17 AM
Steve Spencer
Online
Details
p.2 #5a7R II IBIS effectiveness
dass101 wrote:
I read the Cameralabs review and what I took from regarding IBIS is that with the high resolution the benchmark has also moved. So while we're talking 3-4 stops improvement is compared to what? To 1/focal? Or to best razor sharp? What I'm saying is maybe the camera provides 3 stops over razor sharp, which translates to 2 stops over 1/f, because at this resolution you can't consistently get razor sharp at 1/f.


Ultimately what matters is what each person gets out of IBIS. It will vary from person to person. Some people have shakier hands than others. You actually get more out of IBIS if you have shakier hands (but steadier hands will let you go to slower shutter speeds even if you get less from IBIS). So, use the camera and compare you shutter speeds for hand held shots with and without IBIS and you can get a good sense of how many stops you get. This is one test that is pretty easy to do on one's own.
Sep 04, 2015 at 04:53 AM
exdeejjjaaaa
Offline
Details
p.2 #6a7R II IBIS effectiveness
Big Cheese wrote:
Matt Grum wrote
exdeejjjaaaa wrotem43 users have perfectly working IBIS for 7mm (=14mm FF eq FOV)... like in 7-14mm zooms there
No they don't - it may be good enough, but not perfect. All mechanical stabilisation systems suffer from exactly the same problem. It's the nature of projective geometry.

alundeb wroteThanks, Matt. I hadn't thought of that. I will also keep my expensive Canon IS lenses
The problem is also not unique to sensor-shift stabilisation it also affects lens based IS.
I am not sure what this adds. Of course there is no perfect IS system. That is, except a tripod. With IS if it is good enough then it is good enough. I've never noticed degradation in the corners of wideangle lenses due to ineffective IS as being a particular problem.


anyways I could perfectly get same moire with IBIS both in corner and center of the frame stopped down to 5.6 shooting regular pattern while handholding camera (E-M1, when I owned it).... projective geometry did not blur anything to avoid a moire
Sep 04, 2015 at 10:55 AM
Matt Grum
Offline
Details
p.2 #7a7R II IBIS effectiveness
Big Cheese wrote:
I am not sure what this adds. Of course there is no perfect IS system. That is, except a tripod. With IS if it is good enough then it is good enough. I've never noticed degradation in the corners of wideangle lenses due to ineffective IS as being a particular problem.


It adds facts about IS systems and their limitations. Maybe I shouldn't have used the word perfect. The point is rotation about the camera axis can be corrected, whilst pitch/yaw motions can only be approximated by moving the sensor and that approximation becomes less and less valid as your field of view increases.

You can correct for pitch and yaw accurately in software, provided you sample the image several times. You can also correct for all types of motion exactly by moving the sensor and lens as one (i.e. using a steadycam).

You might not have noticed it, but others have:

http://diglloyd.com/blog/index-60.html#20150823_0800-SonyA7R_II-IBIS

The point is, if you're using a wideangle lens and your corners are in focus you shouldn't expect your IBIS images to be sharp in the corners.
Sep 04, 2015 at 11:02 AM
GabrielPhoto
Offline
Details
p.2 #8a7R II IBIS effectiveness
How do you guys test the stabilization exactly to determine how many stops?
Just trying to see how much I can get for my legacy lenses using the 3 axis one.
Thanks!
Sep 09, 2015 at 03:36 PM
philber
Offline
Details
p.2 #9a7R II IBIS effectiveness
I don't do a scientific test, but more a "common use" test. I go down in shutter speed to where I am getting around 1 good shot in 4. This is what i consider my minimum acceptable shutter speed (for stationary targets). Then I do the same with the stabliser on, and note the difefrence in shutter speed.
My experience so far is a bit unexpected. With a shorter FL (Zeiss ZM 35 f:1.4), I get about 2 stops. With my 55mm (Zeiss Otus 55 f:1.4), I get around 2 1/2 to 3 stops. Now I am using a Zeiss 135 f:2.0, and I am getting even better results, at least 3 stops, maybe 3 1/2.
Sep 09, 2015 at 03:49 PM
GabrielPhoto
Offline
Details
p.2 #10a7R II IBIS effectiveness
Interesting. I will try that tonight for kicks. I always thought of it more along the "rule" of for example setting 1/100 for a 100mm lens and using that as my base but of course it makes more sense to test how much I can improve compared to no Stabilization.
Sep 09, 2015 at 04:03 PM
Charlie N
Offline
Details
p.2 #11a7R II IBIS effectiveness
philber wrote:
I don't do a scientific test, but more a "common use" test. I go down in shutter speed to where I am getting around 1 good shot in 4. This is what i consider my minimum acceptable shutter speed (for stationary targets). Then I do the same with the stabliser on, and note the difefrence in shutter speed.
My experience so far is a bit unexpected. With a shorter FL (Zeiss ZM 35 f:1.4), I get about 2 stops. With my 55mm (Zeiss Otus 55 f:1.4), I get around 2 1/2 to 3 stops. Now I am using a Zeiss 135
...Show more →

my guess would be the weight and size of the lens to be a big factor. For instance, with a lightweight 135mm F2.8, I handheld with generally the same shutter speeds as a 200mm F2, around 1/FL. Something like the OM 200 F5, I have to go 1/2FL to get anything close to usable, mostly 1/3FL is where I try to keep it, that's 1/600 for as consistent results as the 200F2 @ 1/200.
Sep 09, 2015 at 05:09 PM
rainer-v
Offline
Details
p.2 #12a7R II IBIS effectiveness
i am not so optimistic here. I think i can gain between 1-2 stops, rarely more. Sometimes, if i can get more benefit, this is rarely consistent, so i think i can go down to 1/2 /f.length if i want consistant sharp images.
Sep 11, 2015 at 02:46 AM
charles.K
Offline
Details
p.2 #13a7R II IBIS effectiveness
With good technique, I am still seeing about 3 to 3 1/2 stops. I do think we have to be mindful that we have 42MP's.

A shot in the almost darkness in the corridors of the convent at El Escorial, Madrid.
A7rII and WATE 21mm. 6400 ISO, f/5.6 1/5sec. There would have been no way I could have ever captured this shot without IBIS, maybe of course the A7s, but with very high ISO. I did have another shot at 1/2 sec which also was fine.




Sep 11, 2015 at 06:28 AM
InFocus2014
Offline
Details
p.2 #14a7R II IBIS effectiveness
When I was shooting the A7II, I always marveled at the 4 stops in stabilization I was actually experiencing with the FE 35mm f1.4. Sometimes I would almost pinch myself when I would get tack-sharp shots at 1/8 second.

When I started shooting the A7RII, I found I was getting about 2.5 to 3 stops in effective stabilization when comparing directly with the A7II. This was not unexpected, however.

The rule of thumb for shutter speed in the film days was 1/FL. When digital came along the rule still seemed to apply up to about 16mp. When we started shooting 22 to 24mp, the formula was likely a bit light - 1/(1.5*FL) might have been a better rule-of-thumb.

When I started shooting the D800 (then later the A7R), it became immediately apparent that the formula needed to be modified, once again. My experience seemed to dictate something closer to 1/(2.5*FL) to avoid pixel blurring under most conditions (1/(2*FL) minimum).

So, for me, that explains the difference I am seeing between the 24mp A7II and the 42mp A7RII. Even on the A7RII, the IBIS performance is amazing, however.
Sep 11, 2015 at 11:13 AM
FM ForumsSony ForumJoin Upload & Sell
  
 

You are not logged inLog inRegister