3-axis IBIS comparisons

clengman · Feb 7, 2015

I'm wondering if anyone has any data comparing early IBIS implementations (e.g. E-PL1) vs. later Pen models (E-PL3) vs current 3-axis (E-M10). Have there been any improvements here or is the only difference between the 3-axis and 5-axis systems? (and apparently between first generation 5-axis and the new 5-axis in the E-M5 II.)

Steven Wandy · Feb 7, 2015

Data - nope. General impressions - yep.

I had several versions of the earlier 3-Axis (various Pens and EPL models) and now have the EM10 and the general impressions I get - again, no data - is it works better. Have I done a direct comparison of shots, no, just a general feeling of the way it works the the end results.

Also, I had the EP5 and EM5 (older 5-Axis) before getting the EM10 and while I don't do a lot of video - which from what I have read the "missing" two axises are more important for than stills - I have not noticed any decrease in the quality of the usage or results going from the EM5 and then EP5 to the EM10.

clengman · Feb 7, 2015

Steven Wandy said:
Data - nope. General impressions - yep.

I had several versions of the earlier 3-Axis (various Pens and EPL models) and now have the EM10 and the general impressions I get - again, no data - is it works better. Have I done a direct comparison of shots, no, just a general feeling of the way it works the the end results.

Also, I had the EP5 and EM5 (older 5-Axis) before getting the EM10 and while I don't do a lot of video - which from what I have read the "missing" two axises are more important for than stills - I have not noticed any decrease in the quality of the usage or results going from the EM5 and then EP5 to the EM10.

Thanks! General impressions are good, too!

I'm generally pretty happy with the IBIS on the E-PL1, just curious about whether I might expect some improvement when I upgrade cameras. Probably won't be choosing any of the cameras with 5-axis.

I believe 5-axis good for macro shooting as well as video. I don't do a lot of video, but I do like macro. I don't generally rely on IBIS for that though. Usually set up on a tripod or use a flash if I'm looking at slow shutter speed with available light.

I do really like stabilization for use with my telephoto lenses for wildlife, kids playing etc.

drj3 · Feb 7, 2015

I believe the E-M10 was the first Olympus to have 3 axis image stabilization. At least for the DSLR's E510 and E5, I believe it was 2 axis. The E-M10 is much better than either of the DSLR's, I would estimate 2 stops for the E510 & E5, three plus for the E-M10 and four plus for the E-M1 for me. I don't have any of the other mFTs cameras, so I don't know how effective they would be for me and how effective any of the systems are probably also depends greatly on the user.

Guy Parsons · Feb 7, 2015

clengman said:
I'm wondering if anyone has any data comparing early IBIS implementations (e.g. E-PL1) vs. later Pen models (E-PL3) vs current 3-axis (E-M10). Have there been any improvements here or is the only difference between the 3-axis and 5-axis systems? (and apparently between first generation 5-axis and the new 5-axis in the E-M5 II.)

First, I have a Timeline page which shows which body has which IBIS http://homepages.ihug.com.au/~parsog/olyepl1/80-Oly-timeline.html

Through my hands have passed E-PL1, E-P3, E-PL5, E-P5. (Still got them all).

Comments.....

The 2 axis IBIS (E-PL1, E-P3, E-PL5) adds blur to images due to the way it works, turn it off for best results if the shutter speeds allow it. For me the 2 axis allowed the same 3 stops slower shutter speeds as the Panasonic Mega OIS delivers. Must turn 2 axis IBIS off when on a tripod.

The 5 axis IBIS on my E-P5 is unbelievably good, I have at times achieved over 5 stops slower shutter speeds than the old hand-held rule of shutter speed = 1/(35mm equivalent focal length). But never bothered to do a real test, all I know is that I can use the 75-300mm lens with total confidence with the 5 axis IBIS at sometimes silly slow shutter speeds. I leave the 5 axis IBIS on all the time, even on a tripod with no ill effects seen so far.

The 3 axis IBIS of the E-M10 and E-PL7 should be practically as good as the 5 axis IBIS except for macro and closeups where the extra 2 movements matter more.

The improved 5 axis IBIS of the E-M5 Mk2 should be marginally better again than the regular 5 axis IBIS of the E-P5, E-M5 and E-M1.

And then there is the 0 second anti-shock that uses electronic first curtain shutter in the critical range below 1/320 sec - that is definitely the icing on the cake allowing sharper and more reliable results at all times in the critical shutter speed range. Only available on later bodies as per my Timeline chart linked above.

Summary.

2 axis IBIS used carefully when appropriate is better than no IBIS.
3 & 5 axis IBIS much better, sharper results when compared to 2 axis IBIS.
0 sec anti-shock and 3 & 5 axis IBIS = very good indeed.

Regards.... Guy

travellighter · Feb 7, 2015

Very informative timeline table. Bookmarked!

baxters · Feb 7, 2015

Look at the last video sample from the DPR EM10 review. What do you think?

http://www.dpreview.com/reviews/olympus-om-d-e-m10/8

I think it's awful. The 5-axis IS does better.

clengman · Feb 7, 2015

Thanks Guy, that's great information. Looking forward to getting another 1 or 2 stops of stabilization. I have to say though, that my experience with the E-PL1 doesn't exactly match yours. After I started using the 1/8 anti-shock setting I really haven't had too much trouble with shutter shock. It'd be really unfortunate if it was still a problem. I really noticed it when using my 40-150 on overcast days when my shutter speeds came down into the 1/100 - 1/200 range. Just a little too slow for handholding reliably without stabilization, but in the range where shutter shock is worst. I tried the 1/8 sec antishock and I like the result. I keep IBIS on most of the time now unless I'm using a tripod, or if I know I'm not going to need it.

I also wonder about 2-axis vs 3-axis. I know that in-lens stabilization has always been 2-axis (x and y rotation). That's all that in-lens systems are capable of correcting as far as I know. I thought one of the selling points of oly's IBIS, right from the beginning, is that it can add stabilization in a 3rd axis (z rotation) that isn't possible with in-lens systems.

Guy Parsons said:
clengman said:

I'm wondering if anyone has any data comparing early IBIS implementations (e.g. E-PL1) vs. later Pen models (E-PL3) vs current 3-axis (E-M10). Have there been any improvements here or is the only difference between the 3-axis and 5-axis systems? (and apparently between first generation 5-axis and the new 5-axis in the E-M5 II.)

Click to expand...

First, I have a Timeline page which shows which body has which IBIS http://homepages.ihug.com.au/~parsog/olyepl1/80-Oly-timeline.html

Through my hands have passed E-PL1, E-P3, E-PL5, E-P5. (Still got them all).

Comments.....

The 2 axis IBIS (E-PL1, E-P3, E-PL5) adds blur to images due to the way it works, turn it off for best results if the shutter speeds allow it. For me the 2 axis allowed the same 3 stops slower shutter speeds as the Panasonic Mega OIS delivers. Must turn 2 axis IBIS off when on a tripod.

The 5 axis IBIS on my E-P5 is unbelievably good, I have at times achieved over 5 stops slower shutter speeds than the old hand-held rule of shutter speed = 1/(35mm equivalent focal length). But never bothered to do a real test, all I know is that I can use the 75-300mm lens with total confidence with the 5 axis IBIS at sometimes silly slow shutter speeds. I leave the 5 axis IBIS on all the time, even on a tripod with no ill effects seen so far.

The 3 axis IBIS of the E-M10 and E-PL7 should be practically as good as the 5 axis IBIS except for macro and closeups where the extra 2 movements matter more.

The improved 5 axis IBIS of the E-M5 Mk2 should be marginally better again than the regular 5 axis IBIS of the E-P5, E-M5 and E-M1.

And then there is the 0 second anti-shock that uses electronic first curtain shutter in the critical range below 1/320 sec - that is definitely the icing on the cake allowing sharper and more reliable results at all times in the critical shutter speed range. Only available on later bodies as per my Timeline chart linked above.

Summary.

2 axis IBIS used carefully when appropriate is better than no IBIS.

3 & 5 axis IBIS much better, sharper results when compared to 2 axis IBIS.

0 sec anti-shock and 3 & 5 axis IBIS = very good indeed.

Regards.... Guy

--
....................................................................................
Every day I learn something new.
And every day I also manage to forget two things that I learned earlier.

Guy Parsons · Feb 7, 2015

clengman said:
Thanks Guy, that's great information. Looking forward to getting another 1 or 2 stops of stabilization. I have to say though, that my experience with the E-PL1 doesn't exactly match yours. After I started using the 1/8 anti-shock setting I really haven't had too much trouble with shutter shock. It'd be really unfortunate if it was still a problem. I really noticed it when using my 40-150 on overcast days when my shutter speeds came down into the 1/100 - 1/200 range. Just a little too slow for handholding reliably without stabilization, but in the range where shutter shock is worst. I tried the 1/8 sec antishock and I like the result. I keep IBIS on most of the time now unless I'm using a tripod, or if I know I'm not going to need it.

Variations in the human component makes the most difference.

Yes, always 1/8 sec anti-shock for me on E-PL1, E-P3, E-PL5. Seems to keep me out of trouble.

clengman said:
I also wonder about 2-axis vs 3-axis. I know that in-lens stabilization has always been 2-axis (x and y rotation). That's all that in-lens systems are capable of correcting as far as I know. I thought one of the selling points of oly's IBIS, right from the beginning, is that it can add stabilization in a 3rd axis (z rotation) that isn't possible with in-lens systems.

Yes, the optical systems are only 2 axis but maybe there's lens 3 axis out there somewhere by now with some lens maker?

The 2 axis in lens or body equals purely X-Y sensor movements to compensate for up/down tilt and sideways tilt of the lens by the hand-holding human.

The 3 axis adds the sensor rotate to compensate for the roll about the lens axis, too enthusiastic pressing of the shutter button can do that. A walking video also does that.

The 5 axis adds to the 3 axis with the parallel up/down and parallel sideways movement correction, most likely to happen in macro and closeups. I suspect that it would need distance information from the lens, so adapted legacy lenses would only be handled in a 3 axis way.

Now it seems with the E-M5 Mk2 they have improved both the hardware and the firmware to make the 5 axis IBIS work even better.Meanwhile, I have no complaints about the 5 axis of the E-P5, hard to improve on that. That one definitely needed the 0 second anti-shock as I has shocking results when I turned it off for the first time after some months of use.

Regards...... Guy

clengman · Feb 7, 2015

Guy Parsons said:
clengman said:

Thanks Guy, that's great information. Looking forward to getting another 1 or 2 stops of stabilization. I have to say though, that my experience with the E-PL1 doesn't exactly match yours. After I started using the 1/8 anti-shock setting I really haven't had too much trouble with shutter shock. It'd be really unfortunate if it was still a problem. I really noticed it when using my 40-150 on overcast days when my shutter speeds came down into the 1/100 - 1/200 range. Just a little too slow for handholding reliably without stabilization, but in the range where shutter shock is worst. I tried the 1/8 sec antishock and I like the result. I keep IBIS on most of the time now unless I'm using a tripod, or if I know I'm not going to need it.

Click to expand...

Variations in the human component makes the most difference.

Yes, always 1/8 sec anti-shock for me on E-PL1, E-P3, E-PL5. Seems to keep me out of trouble.

Guy Parsons said:

I also wonder about 2-axis vs 3-axis. I know that in-lens stabilization has always been 2-axis (x and y rotation). That's all that in-lens systems are capable of correcting as far as I know. I thought one of the selling points of oly's IBIS, right from the beginning, is that it can add stabilization in a 3rd axis (z rotation) that isn't possible with in-lens systems.

Click to expand...

Yes, the optical systems are only 2 axis but maybe there's lens 3 axis out there somewhere by now with some lens maker?

I don't think it's possible because rotating the lens isn't going to do the trick.

Guy Parsons said:
The 2 axis in lens or body equals purely X-Y sensor movements to compensate for up/down tilt and sideways tilt of the lens by the hand-holding human.

The 3 axis adds the sensor rotate to compensate for the roll about the lens axis, too enthusiastic pressing of the shutter button can do that. A walking video also does that.

Yeah, I was sure I saw some talk about Olympus IBIS correcting for exactly this action before the E-M5 came along. I never saw any discussion of "2-axis" or "3-axis" IBIS until people began trying to name the "IBIS system that is not 5-axis." I can't find where I read about IBIS in the early PEN models correcting for shutter jerk. I may have been mistaken all this time. (but honestly?... I don't think I am.)

Guy Parsons said:
The 5 axis adds to the 3 axis with the parallel up/down and parallel sideways movement correction, most likely to happen in macro and closeups. I suspect that it would need distance information from the lens, so adapted legacy lenses would only be handled in a 3 axis way.

Now it seems with the E-M5 Mk2 they have improved both the hardware and the firmware to make the 5 axis IBIS work even better.Meanwhile, I have no complaints about the 5 axis of the E-P5, hard to improve on that. That one definitely needed the 0 second anti-shock as I has shocking results when I turned it off for the first time after some months of use.

Regards...... Guy

ristar · Feb 7, 2015

Another advantage of the 3- and 5-axis stabilization systems, compared to the older 2-axis, is that the live view can be stabilized when you half-press the shutter to focus. At least for me, this is a huge deal when using telephoto lenses, though it is also useful for shorter focal lengths.

Guy Parsons · Feb 7, 2015

clengman said:
clengman said:

Yes, the optical systems are only 2 axis but maybe there's lens 3 axis out there somewhere by now with some lens maker?

Click to expand...

I don't think it's possible because rotating the lens isn't going to do the trick.

Somewhere, somehow, somebody will invent a way, but it's more likely to be a hybrid part lens part body system.

Casio pocket cameras get "5 axis" by taking multiple shots at 40 fps and merging.

clengman said:
clengman said:

The 2 axis in lens or body equals purely X-Y sensor movements to compensate for up/down tilt and sideways tilt of the lens by the hand-holding human.

The 3 axis adds the sensor rotate to compensate for the roll about the lens axis, too enthusiastic pressing of the shutter button can do that. A walking video also does that.

Click to expand...

Yeah, I was sure I saw some talk about Olympus IBIS correcting for exactly this action before the E-M5 came along. I never saw any discussion of "2-axis" or "3-axis" IBIS until people began trying to name the "IBIS system that is not 5-axis." I can't find where I read about IBIS in the early PEN models correcting for shutter jerk. I may have been mistaken all this time. (but honestly?... I don't think I am.)

Shutter jerk - that the name for the human holding the camera

Shutter shock is what the camera contributes.

The IBIS systems cannot correct for shutter shock as they are tuned for human movement, not for high frequency shock inside the camera.

The E-PL1 for one does not mention about correcting for shutter shock directly apart from this on page 75......

It's only when you get to the 0 second anti-shock on a few of the later cameras that you get this on the screen....

That is the first time that the effect of "shutter impact" has been acknowledged by Olympus.

Regards.... Guy

clengman · Feb 7, 2015

Guy Parsons said:
clengman said:

Guy Parsons said:

Yes, the optical systems are only 2 axis but maybe there's lens 3 axis out there somewhere by now with some lens maker?

Click to expand...

I don't think it's possible because rotating the lens isn't going to do the trick.

Click to expand...

Somewhere, somehow, somebody will invent a way, but it's more likely to be a hybrid part lens part body system.

Casio pocket cameras get "5 axis" by taking multiple shots at 40 fps and merging.

Guy Parsons said:

Guy Parsons said:

The 2 axis in lens or body equals purely X-Y sensor movements to compensate for up/down tilt and sideways tilt of the lens by the hand-holding human.

The 3 axis adds the sensor rotate to compensate for the roll about the lens axis, too enthusiastic pressing of the shutter button can do that. A walking video also does that.

Click to expand...

Yeah, I was sure I saw some talk about Olympus IBIS correcting for exactly this action before the E-M5 came along. I never saw any discussion of "2-axis" or "3-axis" IBIS until people began trying to name the "IBIS system that is not 5-axis." I can't find where I read about IBIS in the early PEN models correcting for shutter jerk. I may have been mistaken all this time. (but honestly?... I don't think I am.)

Click to expand...

Shutter jerk - that the name for the human holding the camera Shutter shock is what the camera contributes.

Sorry, maybe you misunderstood. I was talking about jerking the shutter. Rotating the camera around the optical axis when pressing the shutter release. I was pretty certain I read something about early PEN IBIS correcting for that type of camera motion. It was well before anyone began talking about the E-M5s "5-axis" system. No one really talked about 2-axis or 3-axis at that point. There was also a lot of confusion about what was added when olympus moved from the early system to the 5-axis system. A lot of people (wrongly) assumed that the sensor was now able to tilt in the x and y axes and that now the camera could correct for vertical tilt and yaw. It's still a common misconception of the way IBIS works.

Like I said, I could be wrong, but I was pretty sure that the early PENs were able to correct for rotation about the optical axis by rotating the sensor. It's something I "learned" (perhaps incorrectly) before so called "5-axis" IBIS was introduced.

Guy Parsons said:
The IBIS systems cannot correct for shutter shock as they are tuned for human movement, not for high frequency shock inside the camera.

The E-PL1 for one does not mention about correcting for shutter shock directly apart from this on page 75......

It's only when you get to the 0 second anti-shock on a few of the later cameras that you get this on the screen....

That is the first time that the effect of "shutter impact" has been acknowledged by Olympus.

Regards.... Guy

EFCS will be a great thing to have.

clengman · Feb 7, 2015

ristar said:
Another advantage of the 3- and 5-axis stabilization systems, compared to the older 2-axis, is that the live view can be stabilized when you half-press the shutter to focus. At least for me, this is a huge deal when using telephoto lenses, though it is also useful for shorter focal lengths.

This will definitely be nice. Looking forward to it.

I don't think it's a result of which camera movements can be corrected, though. In other words, I think this could be implemented in a system that corrects for two types of movement just as easily as it could in one that corrects for 3 or 5. It's just a matter of when the correction is activated. (Though the 2 movement system clearly wouldn't be as effective as a 3 or a 5 movement system) I can imagine that prolonged stabilization while shooting video or using the stabilized view in the VF would really drain the battery quickly. Does it?

drj3 · Feb 7, 2015

clengman said:
ristar said:

Another advantage of the 3- and 5-axis stabilization systems, compared to the older 2-axis, is that the live view can be stabilized when you half-press the shutter to focus. At least for me, this is a huge deal when using telephoto lenses, though it is also useful for shorter focal lengths.

Click to expand...

This will definitely be nice. Looking forward to it.

I don't think it's a result of which camera movements can be corrected, though. In other words, I think this could be implemented in a system that corrects for two types of movement just as easily as it could in one that corrects for 3 or 5. It's just a matter of when the correction is activated. (Though the 2 movement system clearly wouldn't be as effective as a 3 or a 5 movement system) I can imagine that prolonged stabilization while shooting video or using the stabilized view in the VF would really drain the battery quickly. Does it?

I don't think the 5 axis stabilization would have much of an effect on battery drain, since there should be little difference in the power required to hold the sensor in place or move it with stabilization . I have no idea where it would affect the battery on the 3 axis stabilization. However, any time you are shooting video or have the LCD or EVF on, there is definitely a battery drain.

Guy Parsons · Feb 7, 2015

clengman said:
Sorry, maybe you misunderstood. I was talking about jerking the shutter. Rotating the camera around the optical axis when pressing the shutter release. I was pretty certain I read something about early PEN IBIS correcting for that type of camera motion.

Yes, sorry, I did substitute "shock" for "jerk".

No way that 2 axis IBIS or OIS can fix that possible rotation about the lens axis if the camera tilts sideways in a twist motion when a shutter is pressed in a jerky fashion on the right side. If the whole camera+lens does a nose dive towards the ground and not a twist about the lens axis motion, then 2 axis can help, but there are limits. People need to learn how to gently squeeze the button and be slow to let it go. Stabbing does not do well.

clengman said:
It was well before anyone began talking about the E-M5s "5-axis" system. No one really talked about 2-axis or 3-axis at that point. There was also a lot of confusion about what was added when olympus moved from the early system to the 5-axis system. A lot of people (wrongly) assumed that the sensor was now able to tilt in the x and y axes and that now the camera could correct for vertical tilt and yaw. It's still a common misconception of the way IBIS works.

Yes, some still think that it can tilt. It certainly can only move in the same plane, no tilt, no forwards or backwards motion. The early videos of 5 axis in action definitely shows it only doing its dance in the one plane.

clengman said:
Like I said, I could be wrong, but I was pretty sure that the early PENs were able to correct for rotation about the optical axis by rotating the sensor. It's something I "learned" (perhaps incorrectly) before so called "5-axis" IBIS was introduced.

No, not any chance that the old 2 axis could do that, it only has simple X-Y movements. Sensor rotation only came with the E-M5 in early 2012 and on all 3/5 axis IBIS cameras that came later.

clengman said:
EFCS will be a great thing to have.

Yes, the 0 sec anti-shock certainly saves the E-P5 from the scrap-heap, useless without it.

But the holy grail of course is the global shutter, one day when the camera industry is about totally dead they may come up with it.

Regards....... Guy

Henry Richardson · Feb 7, 2015

clengman said:
Sorry, maybe you misunderstood. I was talking about jerking the shutter. Rotating the camera around the optical axis when pressing the shutter release. I was pretty certain I read something about early PEN IBIS correcting for that type of camera motion. It was well before anyone began talking about the E-M5s "5-axis" system. No one really talked about 2-axis or 3-axis at that point. There was also a lot of confusion about what was added when olympus moved from the early system to the 5-axis system. A lot of people (wrongly) assumed that the sensor was now able to tilt in the x and y axes and that now the camera could correct for vertical tilt and yaw. It's still a common misconception of the way IBIS works.

Like I said, I could be wrong, but I was pretty sure that the early PENs were able to correct for rotation about the optical axis by rotating the sensor. It's something I "learned" (perhaps incorrectly) before so called "5-axis" IBIS was introduced.

Pentax said a few years ago that their IBIS could do this.

Guy Parsons · Feb 7, 2015

Henry Richardson said:
Member said:

[rotation around lens axis]

Click to expand...

Pentax said a few years ago that their IBIS could do this.

I had a search and found no mention of this at all amongst the Ricoh/Pentax information. I have to assume from their pages that they have simple 2 axis correction like early Pens and OIS has.

Digging further in DPReview I found reference to "horizon correction" to get a tilted image level and they implied that the sensor does that, but no mention of a dynamic roll correction, it sounds more like a simple levelling action. Other models may or may not have this horizon correction. Only a Pentax owner who has tested this function may be of help here. The DPR review of the K-7 wasn't a lot of use in regard to their simple mention of horizon correction.

Meanwhile on this page http://www.pentaxforums.com/articles/tutorial-videos/pentax-shake-reduction-video-tutorial.html the video is quite good, but one main point to me was that you must hold the shutter button half pressed for at least a second for the stabilisation to kick in for the exposure, just a quick press through will not use stabilisation effectively. Plus they imply that all systems work the same.

Hmmmm, no evidence of any problems like that with my E-P5 where 99.9% of the time I have preview IBIS turned off and only use IBIS for the actual exposure period. I can't detect any "1 second warm-up" delays like the Pentax seems to have.

Regards...... Guy

Rod Becker · Feb 8, 2015

ristar said:
Another advantage of the 3- and 5-axis stabilization systems, compared to the older 2-axis, is that the live view can be stabilized when you half-press the shutter to focus. At least for me, this is a huge deal when using telephoto lenses, though it is also useful for shorter focal lengths.

Yes, I have the E-M10 and it works that way. But I also notice that when I then release the shutter, the image stays stabilized - at least in the EVF. So, question - when releasing the shutter while IS continues operating (in the EVF), does it provide the IS in the taken pic?

Thanks, Rod

ristar · Feb 8, 2015

That is not what I meant to imply. I meant Olympus first introduced live view stabilization with their 5-axis system (and it was never introduced to the older 2-axis system). The 3-axis system appears to be a stripped down version of the 5-axis system, but I am not sure anyone knows for certain. In any case, it appears that instead of updating the 2-axis system to provide live view, Olympus will be using the 3- or 5-axis systems in all new models (e.g. E-P5 and E-PL7). Who knows, they might renew the E-PM line and keep the 2-axis system, though it appears unlikely to me.

I have never tested the effect of using IBIS on battery life since I always leave it on unless I am using a tripod. I rarely shoot video so I cannot answer that question.

3-axis IBIS comparisons

Senior Member

Veteran Member

Senior Member

Forum Pro

Forum Pro

Well-known member

Veteran Member

Senior Member

Forum Pro

Senior Member

Forum Enthusiast

Forum Pro

Senior Member

Senior Member

Forum Pro

Forum Pro

Henry Richardson

Guest

Forum Pro

Leading Member

Forum Enthusiast

Similar threads

Keyboard shortcuts