Apple exploring accelerometer, gyro stabilization for iPhone video recording

2»

Comments

  • Reply 21 of 27
    Quote:
    Originally Posted by melgross View Post


    Sony, and others, use a method in their still camera's where the sensor itself moves around. The disadvantage to this is that they need a sensor with more sensing sites to cover the extra area needed at the edges. It wasn't thought they could do this for full frame, but they did.



    Canon invented optical stabilization which they and Sony, who licensed the technology, use in their video cameras. Canon and Nikon, who also licenses the tech from Canon, use it in their still lenses as well. It's a very expensive technology, but it's the best. It uses sensors and a module centered at the nodal point in the lens that contains lenses and a prism. This vibrates to counter the movement from handholding.



    I'm not certain as to how Apple intends to do this even though I read the patent application.



    It seems as though the software pulls the image around somewhat like vibrating the sensor in cameras. I would think it would require a bigger sensor as well. But I'm missing something.



    It would seem to me that this could be done easily if the sensor was larger, and the camera just detected where the image was, and used those pixels. But that's likely an oversimplification, as it would only work pixel to pixel and wouldn't be able to account for smaller shake.



    So, If you had a larger sensor -- wouldn't most people want the option of shooting a larger, unstabilized frame, in addition to shooting a smaller, stabilized frame.



    I normally use a tripod when filming things like soccer games. That eliminates shake. But I frequently have problems when shooting midrange shots, where the center of action changes suddenly -- a pass or a long kick, for instance. Then, I have rapid side-to-side movement, followed by a a second (or so) to align the new center of action. The only way I have found to compensate for this is to add a dissolve.
  • Reply 22 of 27
    melgrossmelgross Posts: 31,961member
    Quote:
    Originally Posted by Dick Applebaum View Post


    So, If you had a larger sensor -- wouldn't most people want the option of shooting a larger, unstabilized frame, in addition to shooting a smaller, stabilized frame.



    I normally use a tripod when filming things like soccer games. That eliminates shake. But I frequently have problems when shooting midrange shots, where the center of action changes suddenly -- a pass or a long kick, for instance. Then, I have rapid side-to-side movement, followed by a a second (or so) to align the new center of action. The only way I have found to compensate for this is to add a dissolve.



    It's not a whole lot larger, perhaps 10%. just enough to cover the most extreme shake. With a fair number of cameras, you need to turn the anti shake off when on a tripod. There are some cameras that will compensate for horizontal shake when on a tripod, whether still or video. You would need to read the reviews to see which ones.
  • Reply 23 of 27
    addaboxaddabox Posts: 12,660member
    A larger sensor would work, but then you get into non-standard resolutions and what I assume would be higher prices. I don't think video functionality is central enough to the iPhone experience for Apple to invest in their own sensor tech, but who knows?



    This is all more critical on tiny sensors with so so lenses where you're barely eking out an acceptable image to begin with. FCP can get good results because you're often working with material generated by a much larger sensor, or higher initial resolution, or both. One of the big advantages of the higher res pro motion cameras is that you can reframe a shot via cropping without any IQ loss, since your sensor is already overshooting the resolution of your deliverables. No pad like that on a phone camera, though.



    I wonder if might be possible to sufficiently miniaturize the gimbal and actuators to move the sensor itself ala the scheme Mel mentioned. Again, it might not be important enough to the all over iPhone experience to merit the investment, but if anyone is going to improve on things like this it's going to be Apple.



    Apple has a real motivation for this kind of thing because they're the only ones making enough money on their hardware to invest in basic improvements like these. It's where they can differentiate from Android handsets. A stabilized high quality video camera would be the sort of thing that Samsung, just to pick a name out of a hat, would be hard pressed to copy.
  • Reply 24 of 27
    Thanks Guys...



    I learned something(s).



    That's why I visit and participate in these forums.
  • Reply 25 of 27
    melgrossmelgross Posts: 31,961member
    Quote:
    Originally Posted by addabox View Post


    A larger sensor would work, but then you get into non-standard resolutions and what I assume would be higher prices. I don't think video functionality is central enough to the iPhone experience for Apple to invest in their own sensor tech, but who knows?



    This is all more critical on tiny sensors with so so lenses where you're barely eking out an acceptable image to begin with. FCP can get good results because you're often working with material generated by a much larger sensor, or higher initial resolution, or both. One of the big advantages of the higher res pro motion cameras is that you can reframe a shot via cropping without any IQ loss, since your sensor is already overshooting the resolution of your deliverables. No pad like that on a phone camera, though.



    I wonder if might be possible to sufficiently miniaturize the gimbal and actuators to move the sensor itself ala the scheme Mel mentioned. Again, it might not be important enough to the all over iPhone experience to merit the investment, but if anyone is going to improve on things like this it's going to be Apple.



    Apple has a real motivation for this kind of thing because they're the only ones making enough money on their hardware to invest in basic improvements like these. It's where they can differentiate from Android handsets. A stabilized high quality video camera would be the sort of thing that Samsung, just to pick a name out of a hat, would be hard pressed to copy.



    All sensors have more pixels than are used. A few rows on the outside aren't. One reason is because of alignment. If the sensor is off by one row, then it will use one row on that edge. All sensors used for anti-shake have even more rows that normally wouldn't be in the image if stabilization was turned off. But this doesn't mean that non standard resolutions are used.



    Two things I'd love to see. One is stabilization of some kind that is effective with such a small and lightweight device. And two is an optical zoom. Lack of optical zoom is the last thing preventing phone cameras from challenging compact cameras.



    There have been, for years now, tiny lenses developed that are plastic in the sense that they change shape with the application of a small current. These lenses change focus that way. Phillips developed the first one a good ten years ago, and I have a prototype of that. But this concept could also be used for making a zoom. If Apple could just get a 2x zoom, it would change everything.
  • Reply 26 of 27
    Quote:
    Originally Posted by cvaldes1831 View Post


    The Internet has these things called "search engines."



    If you type in "iPhone steadicam" into one of these fancy search engines, you'll end up on a page showing a genuine Steadicam for the iPhone 4 (and priced at what one might expect from Steadicam).



    Same day, you should try these search engines. They can be quite helpful.



    A few suggestions are: Google, Bing, and Yahoo!



    Cool (and expensive) as that is, there is a simple poor man's solution.

    The very smallness that makes small cameras so handy is what also makes them shaky. The small distance from the lens to the points of support (hands) amplifies any small movement. The fact that the lens is not in the center of mass of the phone makes it worse (in fact it's really bad since it's out on the corner of the phone.) Finally, the small mass and therefor inertia of the camera also makes it very susceptible to shake and jitter.



    If a handy person were to simply take an iPhone case and attach handles to it in such a way as to place the iPhone lens at the geometric center of the support points (handles) while keeping the shutter buttons convenient, one would have a very steady and superior point and shoot camera.
  • Reply 27 of 27
    addaboxaddabox Posts: 12,660member
    Quote:
    Originally Posted by DESuserIGN View Post


    Cool (and expensive) as that is, there is a simple poor man's solution.

    The very smallness that makes small cameras so handy is what also makes them shaky. The small distance from the lens to the points of support (hands) amplifies any small movement. The fact that the lens is not in the center of mass of the phone makes it worse (in fact it's really bad since it's out on the corner of the phone.) Finally, the small mass and therefor inertia of the camera also makes it very susceptible to shake and jitter.



    If a handy person were to simply take an iPhone case and attach handles to it in such a way as to place the iPhone lens at the geometric center of the support points (handles) while keeping the shutter buttons convenient, one would have a very steady and superior point and shoot camera.



    There are some pretty bad ass iPhone mounting systems available, up to and including an actual Steadicam:



Sign In or Register to comment.