Image Stabilisation.


paulbroad 12 131 1288 United Kingdom
20 Dec 2016 8:26AM
I have a bit of a comment on this subject, and am interested in opinions. I have several stabilised lenses from Canon, Sigma and Fuji. The manufacturers claim anything from 2 to 5 stops improvement in sharpness depending on lens age and technology at the time.

Improvement over what?

The systems do work to varying degrees. My new 100/400 Fuji is great for hand holding, but 5 stops better than what? The old rule, one over the focal length is just a guide, not a scientific fact.

In order to be truthful with the claims there needs to be published scientific data, and any experiment requires basic datum points. In this case, what is the safe hand held shutter speed for the lens under test with the stabilisation switched off? How do you establish that fixed point?

Impossible, because the point cannot be fixed. It must depend on the degree of magnification, the focal length, the focusing distance, (All obviously related), the size and weight of the lens. Then comes the person holding it. How are they standing, on what are they standing, is it moving or vibrating, is a strong wind blowing, how healthy is the photographer, how old are they etc. etc.

How can a manufacturer make statements which appear as fact without the basic starting data?

Paul

Join ePHOTOzine for free and remove these adverts.

Chris_L 6 5.5k United Kingdom
20 Dec 2016 9:32AM

Quote:Improvement over what?

The same lens in the same conditions with the image stabilisation switched off
20 Dec 2016 9:56AM

Quote:The same lens in the same conditions with the image stabilisation switched off



yep, totally agree
Dave_Canon Plus
13 1.6k United Kingdom
20 Dec 2016 10:08AM
A few years ago when I first bought a couple of stabilised lenses, I also wonder if they really had any practical value for me. I did a few tests by taking photos of a newspaper. I would take a shot with the camera on a tripod then repeat the shot hand held at different shutter speeds with stabiliser on and off. The tripod shot was the reference and for higher speeds there was little or no difference with the hand held shots. As the shutter slowed, signs of camera movement gradually became obvious for stabiliser off. In fact the one over the focal length appeared to be a useful guide for me though this would not necessarily apply to someone less steady. The shots with the stabiliser on were still excellent at the one over the focal length and I could continue to lengthen the shutter speed to the point claimed by the lens manufacturer which was 3 stops. Interestingly while the non-stabilised images gradually worsen as the shutter speed is lengthened, I found that there was a complete collapse at speeds longer that this additional 3 stops for stabiliser on (i.e. much worse than stabiliser off). In practical terms, this means that for my 24-105mm at 100mm, I should use at least 1/100 Sec with stabilisation off but could go down to about 1/15 Sec with it on. In practice, I would probably not push it more than say 1/20 for such hand held shots.

I have not seen any scientific tests but I am sure the manufacturers have such tests.

Dave
themak 6 1.0k Scotland
20 Dec 2016 11:27AM
Personally, I think it's a miracle that I can get sharp shots hand-held at 1/30s at 400mm, so I don't spend much time worrying about what the base point is or whether it's 3.5, 4 or 5 stops worth.
ChrisV 13 2.3k 26 United Kingdom
20 Dec 2016 11:46AM
If you look at lens tests on this very site, they publish data on the number of acceptably sharp shots at given focal lengths and decreasing shutter speeds to demonstrate to what degree IS is effective. The data points will only apply to the specific user [the tester] as each of us represents a variable and it's impossible to make an accurate prediction of how we might perform under varying circumstances [the weather, stance and so on]. It is reasonable to extrapolate however that an average 3-4 stops improvement for the tester for hand-held sharpness, would equate to a similar improvement for any given user, albeit the start and end points would not necessarily be the same.

Relatively long exposures [by any method] always need to be approached with caution anyway. Subject movement needs to be taken into account [and this is pertinent to multi-shot exposure too] even for landscapes. People photography (unless you have very still subjects) is not advisable at under 1/30th in any case, so that IS at shorter focal lengths is generally of less practical use than at longer focal lengths - where it really comes into its own. It seems to be at longer telephoto ranges where lens-based stabilisation is more effective than sensor-based IS - which is why the king of IBIS, Olympus, has started to include it in its longer range optics, offering dual IS.

At telephoto ranges where there are typically smaller apertures, more weight and of course more magnification of subject [and therefore its movement as well as your own], IS is such a big advantage it's become an indispensable feature of modern optics.
riddell 15 89 United Kingdom
20 Dec 2016 3:27PM
It basically enables you to use settings that would give you x number of shots of extra light by slowing down the shutter speed.

Of course you may not want to do that, you may not want a slower shutter speed or are using a tripod, in which case IS of no use.
paulbroad 12 131 1288 United Kingdom
20 Dec 2016 3:52PM
It still depends on who is holding the camera and in what conditions. It all depends how good the photographer is at holding the camera still and if he/she has a thumping great diesel engine under there feet. Surely the point is that, given exactly the same gear, in exactly the same conditions, one photographer might get sharp results at significantly slower shutter speeds than another.

There is no base line. So when the manufacturer says 5 stops better, it still does not say better than what? It means better than YOU can normally manage, but not necessarily the bloke next to you. So the statement has no real clearly defined meaning in general because the performance has so many variables.

Paul
themak 6 1.0k Scotland
20 Dec 2016 4:22PM
Manufacturers usually claim 'up to' x stops which is bound to be under optimal conditions. Testers usually try to take into account the variables by testing with and without IS at a range of settings and quoting % ages of steady images at each. They frequently find the makers claims a stop or so optimistic.
Since you seem to have a good grasp of the difficulties, I'm a little surprised you seem to think there's a better way to do it.
petebfrance 8 2.9k France
20 Dec 2016 5:04PM
There is a CIPA standard which, for better or for worse I think manufacturers are leaning towards using. http://www.cipa.jp/image-stabilization/index_e.html
This may help a bit. I admit; I haven't read the Pdf attachments yet, but hopefully you can glean something from them.
LenShepherd 11 4.1k United Kingdom
20 Dec 2016 6:52PM
I have experience of Nikon.
The "of what" query can be very relevant.
The CIPA standard involves firmly attaching a lens/body to a vibrating platform, which is not an exact comparison to hand holding.
Some suggest CIPA is about 1 stop better than hand holding, though a lot depends on hand holding ability with or without VR.
Nikon literature in lens brochures implies a maximum gain of 4 stops with recent lens at around 1/8th second, around 1.5 to 2 stops at 1/250 and nothing at 1/750.
With a 24x36 body, 2 stops gain at 1/250 is useful but 4 stops gain at 1/8 is still likely to result in some camera shake with say a 400mm unless using a monopod.
Nikon VR certainly works - but not always with maximum CIPA amount at all shutter speeds.
More than 1x focal length, multiplied by any crop factor relevant to 24x36 format, is general considered as a safer hand holding speed for critical quality with recent high resolution sensors.
Chris_L 6 5.5k United Kingdom
20 Dec 2016 7:32PM

Quote:It still depends on who is holding the camera and in what conditions

How does it?

In windy conditions and shaky hangover hands Tom might manage 1/100th second without IS but three stops better with IS

In windy conditions and shaky hangover hands **** might manage 1/125th second without IS but three stops better with IS

Then on a calm day with steady hands Tom might manage 1/60th second without IS but three stops better with IS and **** 1/100th without, three stops better with!

Which way do you think it should be expressed that would enable you to understand what sort of improvement you can expect from IS?
It seems to me that Chris_L has, in his two posts, answered the query.

Sign In

You must be a member to leave a comment.

ePHOTOzine, the web's friendliest photography community.

Join For Free

Upload photos, chat with photographers, win prizes and much more.