[Bf-vfx] Footage for Tracking
mierle at gmail.com
Tue Jun 14 20:32:18 CEST 2011
On Tue, Jun 14, 2011 at 10:55 AM, Andrew Hunter <andrew at aehunter.net> wrote:
> Hey everyone,
> On Tue, Jun 14, 2011 at 12:27 PM, Andrew Hunter <andrew at aehunter.net>
> > Hey Everyone,
> I had another though that perhaps Keir can answer. Does subject
> distance affect the calibration? Should the calibration charts be
> taken from the same distance.
What matters is that the calibration target fills the screes as much as
possible. I can take some "good" calibration shots and post them sometime in
the next few days.
GOOD: Calibration target fills most of view.
GOOD: Calibration target is shown at different angles (needed for accurate
focal length). Perspective should be obvious (converging lines).
GOOD: Calibration target shows the distortion (fronto-parallel view) This is
needed for accurate distortion.
BAD: Calibration target only fills a tiny bit of the image
> The reason I ask is that with wide focal lengths, where distortion is
> most likely to occur, the subject distance greatly affects the
> persption of depth.
For wide focal lengths, you'll have to get closer to the lens.
Also, in order for the chart to fill the frame on different focal
> lengths with the same sized chart (say an 18mm vs an 85mm) the subject
> distance would vary greatly.
Yes, the distance to the calibration grid will vary. I think it's easier to
have one size than to have multiple targets, and just vary the distance to
> Would it make sense to default to a larger chart size like 11x14 or
> offer several different sizes for different ranges of focal lengths?
11x14 is probably fine. It's important that the target is *REALLY* flat. I
sent a PDF earlier in this thread that should do nicely.
> Bf-vfx mailing list
> Bf-vfx at blender.org
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Bf-vfx