sebboh wrote:
if the sensor stays still while the camera and lens move around, it means that the sensor is moving in relation to the camera. or another way to look at it is that the image circle is moving around on the sensor. this means that the sensor must be large enough so that there are are always pixels under a 4/3 crop of the image circle.
Hmm I don't think you are right about the need for an oversized sensor for IS. You kind of have it backwards I believe. Regardless of whether the lens moves or the sensor moves the idea is that the "right portion of the image circle" has to fall on the sensor. So basically that requires that the image circle has to be big enough to accommodate the sensor motion and not the other way around.
Jman13 wrote:
Um, that wouldn't stabilize an image. That would just avoid any initial shock. Trust me...the sensor moves...you can WATCH it move with the lens off.
Agreed. This would not stabilize the image. This can be done for video stabilization where you need to align successive frames with each other but that's a totally different topic. Image stabilization requires movement during the capture to avoid blur.
Jman13 wrote:
Um, that wouldn't stabilize an image. That would just avoid any initial shock. Trust me...the sensor moves...you can WATCH it move with the lens off.
why do you think i'm saying the sensor doesn't move.
I frankly have no idea what you're saying because it really doesn't make sense. There is absolutely no need for there to be extra pixels on a larger sensor for sensor shift IS...the sensor moves in the opposite direction of the lens, which keeps the light focused on the same pixel throughout.
Why would you need extra pixels if the whole sensor is moving to keep the rays of light aligned on the same spot on the sensor? If the image moved outside the main sensor area, that would mean the entire image shifted off the sensor, which would result in blur in the final image...the sensor moves to ensure that doesn't happen.
Jman13 wrote:
I frankly have no idea what you're saying because it really doesn't make sense. There is absolutely no need for there to be extra pixels on a larger sensor for sensor shift IS...the sensor moves in the opposite direction of the lens, which keeps the light focused on the same pixel throughout.
Why would you need extra pixels if the whole sensor is moving to keep the rays of light aligned on the same spot on the sensor? If the image moved outside the main sensor area, that would mean the entire image shifted off the sensor, which would result in blur in the final image...the sensor moves to ensure that doesn't happen....Show more →
i feel like a diagram might be necessary, but don't care that much. the sensor moves to keep the light focused on the same pixel throughout by moving the sensor in the opposite direction of the lens. check. we agree on this. the advantage of a larger sensor is that when the sensor has moved (which it starts doing before the shutter opens) there are still pixels filling the image circle of the lens even though sensor is not centered. this means the camera can choose to take the most centered crop of the pixels. this could be done by using only pixels that are centered at the start of capture time or another selection method.
No, framing is always pretty much exactly where it appeared (when I pressed the shutter). Do you have any factual basis for this theory at all? This is literally the very first time I have ever heard of anything even remotely like this, and since we know the sensor is exactly the same size as non-IS sensors, and smaller than the GH2 sensor, I think it's pretty darn likely they don't do what you are saying (and I think any perceived benefit would be extremely marginal). It sure as heck isn't required for IBIS to work.
Anyway, we've wasted enough time on this sideshow, and it is rather irrelevant anyway. We all know the IBIS works well, and that's pretty much all that's important, right?
sebboh wrote:
i feel like a diagram might be necessary, but don't care that much. the sensor moves to keep the light focused on the same pixel throughout by moving the sensor in the opposite direction of the lens. check. we agree on this. the advantage of a larger sensor is that when the sensor has moved (which it starts doing before the shutter opens) there are still pixels filling the image circle of the lens even though sensor is not centered. this means the camera can choose to take the most centered crop of the pixels. this could be done by using only pixels that are centered at the start of capture time or another selection method....Show more →
K5 and a few other Pentax DSLRs allow you to manually shift the sensor to shift the captured scene. That is only possible if you select different parts of image circle as you shift the sensor. Which clearly means that they assume that the image circle is big enough to accommodate sensor movements.
Also if they were using a lager sensor, it would surely be noted in sensor specs. However Pentax and Sony sensors are just as big as Nikon who doesn't use sensor based shake reduction.
The problem that you have described of the sensor moving out of the optimal image circle region is real but I am pretty sure companies don't use over-sized sensors to fix it. Nikon for example just re-centers its VR system right before start of the exposure to ensure that the sensor is in the center of the image circle.
Jman13 wrote:
No, framing is always pretty much exactly where it appeared (when I pressed the shutter). Do you have any factual basis for this theory at all? This is literally the very first time I have ever heard of anything even remotely like this, and since we know the sensor is exactly the same size as non-IS sensors, and smaller than the GH2 sensor, I think it's pretty darn likely they don't do what you are saying (and I think any perceived benefit would be extremely marginal). It sure as heck isn't required for IBIS to work.
Anyway, we've wasted enough time on this sideshow, and it is rather irrelevant anyway. We all know the IBIS works well, and that's pretty much all that's important, right? ...Show more →
as i said that is what an olympus rep told me, whether he knew what he was talking about is another story. i haven't shot at all with the om-d but i certainly saw framing change from what i saw on the viewfinder or lcd with the e520 and e-p2. i agree this is way off track, and who really cares how it works, IBIS (at least oly's) is awesome.
^ I don't think it works as you're describing sebboh, though I know what you're saying. That really sounds almost like the method used in video stabilization via software.
Like described earlier, a physical movement of some sorts is required for a single sharp frame, but a software shift wouldn't be able to accomodate this because the sensor's CPU/logic would have to intelligently blend what would essentially be a "sensor-sized+X" area of exposed pixels into just a "sensor-sized" area and be able to know which photons belonged to which electron bin - hence physical movement only (either lens elements or sensor), and no electronic stabilization that I know of.
You can do this electronic shifting with video because of the nature of the media (constant motion vs frozen time).
cputeq wrote:
^ I don't think it works as you're describing sebboh, though I know what you're saying. That really sounds almost like the method used in video stabilization via software.
Like described earlier, a physical movement of some sorts is required for a single sharp frame, but a software shift wouldn't be able to accomodate this because the sensor's CPU/logic would have to intelligently blend what would essentially be a "sensor-sized+X" area of exposed pixels into just a "sensor-sized" area and be able to know which photons belonged to which electron bin - hence physical movement only (either lens elements or sensor), and no electronic stabilization that I know of.
You can do this electronic shifting with video because of the nature of the media (constant motion vs frozen time)....Show more →
i don't think i described it very well, because what i'm talking about involves moving the sensor not using software to simulate moving the sensor. for what i'm talking about the larger sensor area is just used to maximize the probability of the moving sensor staying in the sweet spot of the lens.
To me everything comes down to output. The OMD files look way too digital compared with film like files from the X100 though the IS in the Oly is amazing.
I disagree on the 'digital look' for the OM-D. Some other m4/3 cams can occasionally get this look, but I don't feel it's that way for the OM-D. Of course 'looks digital' is a nebulous term. In many cases I could say that translates to 'looks better' since digital is producing higher resolution images with much lower noise than pretty much any 35mm film. But I understand a bit what you mean. A lot of that 'look' comes down to how you process the images.
I was dead set on the X100 about a month ago, but never found one for what I thought was a good price (given the age/coming replacement).
Now I've got funds just burning for an X200 if it delivers next month. Or a nice, cheap X100 if it doesn't.
I don't see what the argument is about the X100 and OM-D. For some people one works better, for other people the other works better. They're different cameras, with some overlapping features/uses-- so it's not likely that everyone will agree on which is best for his needs. But they both seem like great systems for certain uses.
When you have a leaf shutter and 35mm equiv. lens, I don't think IBIS is all that necessary. I can handhold the X100 very slow. Plus, there are weakness areas in some IBIS systems. If I remember correctly from a lot of the reading I did a few months back, a bunch of users were getting random blurriness with the OM-D's IBIS when shooting around the 1/80 shutter range. With my Sonys, I usually turned off the IBIS at higher shutter speeds.
ReneMurea wrote:
To me everything comes down to output. The OMD files look way too digital compared with film like files from the X100 though the IS in the Oly is amazing.
Actually, I do agree with this.
But the OM-D files look better to me even though it looks more digital. The OM-D files don't have as much character but they appear to have more detail at least when using my 14-45, there just appears to be more micro contrast and more resolved structure.
