by Ian Burley
How good is the E-510's moving sensor IS system?
We've been testing the new Olympus E-510 in the test lab to find out just how effectively its new Supersonic Wave driven moving sensor image stabilisation system (IS) performs. 100s of test exposures later, we have our verdict. We've also done a cursory comparison of the effectiveness of Panasonic's in-lens Mega OIS (optical image stabilisation) system featured in its high-end standard zoom, the Leica Vario Elmarit 14-50mm f/2.8-3.5, when fitted to an E-510.
Olympus moving sensor IS, it's a supersonic thing
Moving sensor IS, which Olympus calls 'mechanical' IS, as opposed to 'optical' IS in lenses, is not new. It has been used in some compact cameras for years and such systems have been developed by Ricoh, (Konica) Minolta, Nikon, and Olympus. Konica Minolta was the first to introduce moving sensor IS onto the DSLR stage with the introduction of the Dynax (Maxxum, in some markets) 7D. Pentax later followed and now Olympus joins the fray, with the E-510.
All these competing systems are inherently different, say the manufacturers and the main reason why is because none want to pay licensing fees to the other, assuming that these would be granted, of course. The Olympus system is interesting because it uses in-house Supersonic Wave Drive (SWD) technology to actuate the sensor motion. SWD is a development of the Supersonic Wave filter system that keeps Four Thirds DSLRs sensors clean. Very high frequency vibrations can be turned into mechanical force for precision motorised control; in this case the camera sensor. SWD is also being developed for use in future Olympus lenses for faster, quieter and more precise auto focusing.
OIS v.s. moving sensor IS - which is better?
The above question is like asking if a diesel engine is better than a petrol one. There can be many positive and negative answers. This article doesn't attempt to provide a universal answer. Instead, I'm trying to find out if the Olympus E-510's IS actually works and, to a degree, how well. The quick answer is, yes, most IS systems really do work. I recently tested four compacts with OIS and several DSLRs with both OIS and moving sensor IS systems and all proved beneficial. The best performers, by some margin, happened to be two OIS lenses from Nikon (the 55-200 and 75-300 Zoom Nikkor VR models).
But some other lenses I looked at, like the Canon 70-200 f/4 IS L, a very expensive pro-spec lens, was not as impressive, IS-wise, as the Nikkors. My point is, there is considerable variability in performance. If you read on, you will find that the E510 IS system puts up a very good showing compared to the Panasonic Mega OIS system in the Leica Vario Elmarit 14-50mm that serves as the Panasonic Lumix DMC-L1 'kit lens'.
The testing, the results
How do you test IS systems? Some testers have built oscillating machinery onto which cameras can be mounted. I don't doubt this methodology is useful, but human beings aren't machines - at least, not like that. I have employed machine-validation of the test results by using DXO Analyzer 2.0 software to evaluate test images taken in the test lab, but all the IS test photography was done manually, not by machine.
The DXO system uses a special test target that the Analyzer software can intelligently analyse to provide information about optical distortion, vignetting, chromatic aberration and sharpness, or in DXO-speak, 'blur', or the Blur eXperience Unit (BXU). It's basically similar to an MTF measurement, but in proprietary units. The lower the BXU number, the better. 1.0 is extremely good, 2.0 is OK, while higher figures mean your lens is showing visible softness.
I tested the IS systems by taking test images, both on a tripod for reference, and hand-held using no IS and the different modes available. I generally took at least 20 test images for each mode being evaluated in order to build up a statistically viable result. And here they are in graphical form.
I decided to use 1/15th second shutter speed as the starting point, as this is a speed I consciously know I am going to have great problems keeping steady using the usual standard zoom range. 1/30th seems to be my personal threshold; anything below and sharpness really is a lottery. First of all, I tested the E-510's two modes while the camera was tripod mounted. Conventional wisdom says you should never use IS when the camera is on a tripod. Admittedly, I was using a hefty Manfrotto 058 tripod on a concrete floor, but I did release the shutter manually on the camera. Anyway, the results show that I didn't have major problems. The variations in the chart above are highly magnified and so statistically insignificant. The reference performance of the camera and lens, on a tripod but with IS switched off, is 1.56.
In the chart above, we compare the E-510's Mode 1 IS operation on the tripod and hand held. At 1/15th second, the IS keeps things well in hand.
Suddenly, the chart (above) has become much more complicated. Here we compare the E-510, shooting hand held at 1/15th second in Mode 1, in Mode 2 and with IS switched off. The Mode 1 smoothness is very reassuring. Mode 2, which only operates vertical stabilisation, is on all sorts of bother here. The evidence clearly shows that Mode 2, which is designed for stabilising while panning the camera, can be unstable when the camera is not panned - you have been warned! And look at the trace for operation without IS; the sharpness variability is huge. At 1/15th, at least, the E-510's Mode 1 IS works very well.
Now you can see the relative performance of the E-510 in Mode 1 at 1/15th sec, 1/8th sec and even 1/4 sec. The really wild trace is with IS off at 1/15th sec. As you would expect, the stability erodes progressively as the exposure time is increased, but even at 1/4 sec you are likely to get better results than hand held without IS at 1/15th second.
And what about the OIS Leica Vario Elmarit 14-50mm fitted to the E-510?
First of all, you can only use the in-lens OIS system in Mode 1 (always on). The Leica's Mode 2 isn't a panning mode like the E-510's, but a mode that only triggers at the time of exposure, minimising the possibility of the correction optic in the lens being extremely off-axis. Mode 2 is slightly more consistent, but you can't use it on a non-Leica or Panasonic Lumix body. Nor can you use in-lens OIS at the same time as the E-510's IS being active. If you try, the lens OIS switches off when you press the shutter.
So does the Leica lens OIS work better than the E-510's in-camera moving sensor IS? The chart above suggests that the E-510 has the edge, though statistically there is little real difference. What's more important is that the results prove that both systems are definitely beneficial.
It's not a gimmick - IS really does work and it's a very worthwhile feature to include in your photographic armoury. And the Olympus E-510 moving sensor IS is competitive with the OIS served up by Leica's 14-50mm standard zoom. But don't leave the camera's Mode 2 on unless you need it for panning! And don't expect to combine the forces of the lens OIS and in-camera IS; it won't work. More testing is needed to form a bigger picture and I'll be publishing more results in the future using longer lenses, for example.
| Check the latest price or purchase options for the Olympus E510 via
|Amazon Germany is currently the only European Amazon outlet that supplies electronic goods like cameras, printers and scanners priced in Euros.|
|Each Amazon outlet can ship to most countries in the world.|
Discuss this story: