For restoring a ripped photo we need:
to scan all scraps at once (this is a prerequisite condition to realize
a good work)
to create as many objects as scraps, on different layers, with mask tool,
to move close as much as possible and with caution the image's strips to
reassemble the draw's lines. |
|
If there are some missing pieces, fill them as follows:
select, with mask tool again, areas close to scraps, that is alike,
create one or more duplicates (mask/create from object's)
place their layers below the main image
combine all these objects and tha main image to background. |
|
| Use the clone tool to delete the rip's lines and to give the
image back its completeness |
|
Use the clone tool also to delete lines created by scratches, rucks and,
at large, all the imperfections of the photos.
The clone brush named 'detail', that changes to one pixel at any one time,
is very usefull in the operations on images enlarged also to 300/400 times. |
  |
The default configauration plans it works pixels at any time.
It's main:
to operate at low definition, 96 dpi is the optimum condition (that is the
screen setting)
to modify frequently the place of pixels to be traced(mouse right button
lets to change the place of pixels to be traced cursor, that you can notice
for the cross inside)
to follow the lines direction in the image (i.e. in the pleats above: the
action will be up and down, down and up and so on, in shadow area and light
area one by one)
at the end of work, use clone-soft brush on the image, visualized with low
enlargement, to soften eventual sharpness that have been created. |
|
A satisfactory work requires much patience and it's a good idea to change
often modality of visualization to keep a right overview.
When you think your work is completed, increase the photo definition (dpi),
image/resampling. The program increases the number of the pixels through
a process called interpolation.
This is the same process a scanner uses when scanning at resolutions much
higher than its optics can resolve.
PhotoPaint resamples by examining every pixel throughout the entire image,
comparing pairs of adjacent pixels and creating pixels that represent the
average tonal value.
This method will make softer the colours' tone in the image to make invisible
your editing. |
|
