Re: [Phys-L] Rolling Shutters

[John Denker <jsd@av8n.com>]

On 01/26/2015 11:50 AM, Bob Sciamanda wrote:
> Be aware of the “Rolling Shutter” effect when using CMOS camera
> generated video data  ==>
>
> http://www.diyphotography.net/everything-you-wanted-to-know-about-rolling-shutter/

As pointed out in one of the comments on that web page,
the guitar string video does not exhibit rolling shutter
effects to any great degree.  Instead it exhibits mostly
plain old frame-by-frame time-domain aliasing.

  Rolling shutter is more like line-by-line aliasing,
  as opposed to frame-by-frame.

Either way, it can be formalized and quantified in
terms of a convolution.

At the cost of increased CPU load, the camera "could"
get rid of field-by-field aliasing by taking multiple 
images and applying straightforward digital filter 
technology.

=======

Sometimes you want to get rid of the aliasing, and
sometimes you don't.

Sometimes-constructive suggestion:  if/when you want
to get rid of aliasing (frame-by-frame *or* line-by-
line), you can lower the lighting level and use a 
longer exposure time.

  Low-end cameras that don't have an iris and don't
  have much CPU power are particularly at risk of
  aliasing problems.

If you can't reduce the lighting (e.g. outdoors)
and/or you don't have an iris and/or you want
separate control of brightness and depth of field,
you can put a neutral gray filter over the lens.