All image stabilization created equal?
in [Digital Cameras]
|
Prev: ugly bugs
Next: SX120 as point and shoot
From: Dave Cohen on 4 Aug 2010 12:10 Eric wrote: > On Tue, 03 Aug 2010 23:28:36 +0200, Ofnuts <o.f.n.u.t.s(a)la.poste.net> > wrote: > >> On 03/08/2010 22:25, Eric wrote: >>> I've noticed that most Canon cameras use their "Digic 4" processor. >>> Presumably that is powering their image stabilization. If so, it seems >>> logical that the IS in all their digital cameras would be equally >>> effective (or ineffective, as the case may be). >> No, the IS is done in the lens, without help from the camera. A recent >> lens will du as much good on an old camera body with a "lesser" >> processor than on the shiny new ones with the latest processing marvel. >> >>> Does anyone know more about this? Any other brands have more >>> effective IS systems? >> There are basically two kinds of stabilization: lens-based (Canon, >> Nikon, Panasonic, plus Tamron and Sigma for lenses) as described above >> and sensor-based (sensor is moved behind the lens)(Pentax, Olympus...). >> In currently available cameras, the so-called "software" stabilization >> is likely pushing the ISOs to crank up the shutter speed (there are >> de-shake algorithms, including some that are based on acceleration >> measurement) but I doubt they can run on the processors found in >> entry-level cameras (and the other cameras already have one of the two >> "hardware" IS). >> >>> BTW, one of the reasons that I ask: I've had a Canon SLR zoom lens for >>> quite a while, and the image stabilization was great. Seems to be some >>> kind of intertial-sensing mechanism though--almost a gyro feel to it. >>> Obviously the P&S digitals are doing this in software. >> No, they use lens-based or sensor-based IS. But the accelerometers are >> really tiny chips, not gyroscopes. The "gyro feel" is really from the IS >> lens mechanism, when it's big. >> >>> My first test >>> of their digital IS was with an SD1200. It seemed to do almost >>> nothing. >> Check the specs, it's a lens-based IS (but on a tiny lens) > > That makes sense. I got the original info from two different Canon > techs who said that the IS system was related to their Digic 4 chip, > and that therefore the IS in all their digital cameras would be the > same. In fact, one of the very knowledgable SLR guys had me on hold > for a while to confirm that with another (a third) tech. > > Your comments sound logical, and that would make some sense of why the > IS systems seemed different (ie nonexistent) in the SD1200. > > The next logical question is: How much difference is there between the > IS in the SX120 and the G11? Those are the two cameras I had been > considering. I just got to try them briefly, and it's difficult to > discern the finer points in a short test in a camera store. > I can't speak for the SD1200, but the IS on the A3100 definitely works. Any stabilization is better than nothing and most of us spent many years without such a luxury. I wouldn't worry too much in selecting between the models you're looking at.
From: Peter on 4 Aug 2010 14:09
"SneakyP" <48umofa02(a)WHITELISTONLYsneakemail.com> wrote in message news:Xns9DCA84D35EF048umofa02sneakemailc(a)127.0.0.1... > Eric <Eric(a)sorry---nospam---.com> wrote in > news:10fh56d5hq335mbm5qneo2e19ucsr5cf5s(a)4ax.com: > > >> The next logical question is: How much difference is there between the >> IS in the SX120 and the G11? Those are the two cameras I had been >> considering. I just got to try them briefly, and it's difficult to >> discern the finer points in a short test in a camera store. >> >> > > Take your own memory storage device and use that in those two cameras, or > use two separate memories to test the cameras for whatever you want to > test > for. > > > Then go home and compare the photos you took with both cameras. > Good suggestion: One test I like to do is shoot a brick wall at various zoom lengths and ISOs. Then shoot some known solid colors with defined edges. -- Peter |