Re: paleonet Digital Images and Depth of Field

["Atlas" <r_johnson@xxxxxxxxxxx>]

Hi,

This technique was used to great effect by the recently deceased Jim Craig.
Jim would take a number of images with the point of focus at differing
levels and the super-impose on on top of the other and deleting the out-of
focus area on Paintshop Pro. The resulting images gave truly impressive
results.

To see some of Jim's images, they can be viewed at www.ammonite.ws

Regards


Rick
----- Original Message -----
From: <TomDeVrie@aol.com>
To: <paleonet@nhm.ac.uk>
Sent: Saturday, October 27, 2001 5:40 PM
Subject: paleonet Digital Images and Depth of Field


> Following the discussion of limitations in the depth of field of digital
> cameras, I wondered:
>
> Can the photographer with a digital camera overcome this problem by taking
> two pictures of the same object, but with slightly different distances to
the
> object (say, a 1 cm difference), and then blend the two images in
Photoshop
> to take advantage of the portions of each image that are best focused?
>
> I've started experimenting with this process and initial results are
> encouraging for a uniformly curved gastropod about 3 cm long.  Two
pictures
> were taken with distances to obect varying by one cm.  Each image included
a
> portion in focus and out of focus. One of the two images was re-scaled to
a
> slightly larger size (e.g., 2048 to 2058 pixels) so that image dimensions
> matched.  An airbrush-eraser was used on two Photoshop layers, each layer
> having one of the images.
>
> The final images need to be 'flattened' to create a single image.
>
> Fortunately, there seemed to be no seam between the airbrushing done on
the
> two components of the final image.  Airbrushing the same area produces a
> white patch, but the error can be caught when the airbrushing is first
being
> done and remedied with an "undo" command.
>
> It may be that a small degree of poor resolution near the boundaries of
the
> airbrushed areas is an acceptable price to pay for much better focus along
> the entire vertical distance of the fossil.
>
> Are there others on the listserve who have experimented more extensively
with
> this approach?  Are there parameters beyond which the technique does not
work?
>
> Tom DeVries
>
>