View Full Version : Non-native lens IBIS issues

David Morison
28th May 2013, 07:46 AM
As I often use non-native and legacy lenses on the E-M5 with the IBIS I obviously have to adjust the settings for the focal length of the lens in use (a pain when I need to change lenses between m4/3 or 4/3 and others on a frequent basis). However the settings don't correspond well with the actual focal length and this can be seen when using maximum magnification on the EVF for MF. For instance the best setting for a 12-24mm Tokina is 10mm, the best for Canon 400mm is as expected 400 but 600mm doesn't work well when adding a 1.4x - 400 is still the best option. Similarly with the Pentax 500mm with a 1.4x the most effective setting is 600mm - sometimes!. It seems a bit of a lottery at the moment, perhaps the firmware is not quite as sophisticated as we would like to think. Anyone got any ideas?


28th May 2013, 09:55 AM
Hi David,
I have been entering closest values as prescribed (600 for 600mm, 800 for 850mm and 1000 for 1200mm) and as far as I can see they provide good results. I haven't verified in detail though.

28th May 2013, 11:28 AM
I had mixed result too. I am not certain about the how and why though. The one idea I have is, that legacy glass tends to be more heavy and you may hold the camera more stable therefore, which in turn may irritate the IS and/or lead to overcorrection. Similar to using a tripod with IS on, so to say.

I guess using IS mid range* on a legacy zoom lens is a bit to much. I would not have tried that one ...

* edit: misread this one. 10mm, okay - but anyhow a gamble.

Can we deduce from this, that a low correction is better then none?! I often thought about it and this seems to confirm it. But I am not at all sure.

28th May 2013, 11:43 AM
My natural tendency when setting for a focal length not specifically covered would be to go for the next length up, eg, for a 55mm lens, set to 60mm. But I've never tested to see if my theory is valid.

28th May 2013, 11:57 AM
Although the details of the IBIS algorithm is not known, it involves a constant that is used when transforming the angles detected in any of the yaw, pictch and roll axises into the resulting (user perceived) shake blur in the viewfinder/LCD.

The relationship is inverse proportional to the tangent of the angle of view through the lens. For small angles (telephoto), focal length, tangent and angle can be approximated to have a linear relationship.

This means that a lens with double focal length requires twice as much compensation to be performed by the IBIS to neutralize the angular shake.

28th May 2013, 04:38 PM
Which 1.4 converter are you using.
If it's the Olympus one, I wonder if the camera is able to see it, and consequently requires the correct focal length setting for the manual lens.

28th May 2013, 05:19 PM
Yes, the camera "sees" it since it is elecrtically connected. But since there is no lens attached to it electrically the camera interprets it as unknown focal length/F number.

If you use a legacy lens adapter fitted with an AF confirm chip/dandelion chip that you program to the focal length of the legacy lens then the EC14 will report the focal length/F number multiplied by square root of 2. The EC20 will report times 2.

Cheers Tord