Belgian Amiga Club - ADF Collection/BS1 part 60.zip/BS1 part 60/Imagemaster d4.adf/apa.lzh/mrf_hlp

Text File | 1993-08-30 | 13KB | 369 lines

\ApAssist
\: - This is the Quick Help for the process filters panel -
\: ---------------------------------------------------------------------
\:   Written by David E. Patterson
\: ---------------------------------------------------------------------

\font topaz.font 8
\tc 1
\wr
\lj
\dd "imh_descriptor"

\index "hlp_index/index"
\toc "hlp_toc/toc"
\help "aa_help/help

\node "mrf_sharp1"
\title "Sharpen I"
\next "mrf_sharp2"
\prev ""
This function works by detecting edges, and then increasing the rate at which
those edges occur.  Edges are defined as changes in brightness.  The control
allows you to specify the amount of the effect.  If it is set to zero there
will be no effect.  If it is set to 100 then the maximum amount of Sharpen I
will occur in the specified area.  Sharpen I uses a 3 by 3 array of pixels to
detect changes (edges) in the image.

For example, this function would take a scan line that goes from black to
white in three pixels, and change the line so that the transition happens in
two to two and a half pixels.  Your eye would percieve a quicker change, and
thus, you percieve a sharper image.

See Also: \{ \nw \ts b \tc 15 \link "Sharpen II " "mrf_sharp2"
          \link "Sharpen III" "mrf_sharp3"
\}

\image "clips/sharpen.iff" c
\flushimage
\ce \{ \ts b Sharpen I Example \}

\endnode

\node "mrf_sharp2"
\title "Sharpen II"
\next "mrf_sharp3"
\prev "mrf_sharp1"
This function works by detecting edges, and then increasing the rate at which
those edges occur.  Edges are defined as changes in brightness.  The control
allows you to specify the amount of the effect.  If it is set to zero, there
will be no effect.  If it is set to 100, then the maximum amount of Sharpen II
will occur in the specified area.  Sharpen II uses a 5 by 5 array of pixels to
detect changes (edges) in the image.

For example, this function would take a scan line that goes from black to
white in five pixels, and change the line so that the transition happens in
two to two and a half pixels.  Your eye would percieve a quicker change, and
thus, you percieve a sharper image.

See Also: \{ \nw \ts b \tc 15 \link "Sharpen I  " "mrf_sharp1"
          \link "Sharpen III" "mrf_sharp3"
\}

\endnode

\node "mrf_sharp3"
\title "Sharpen III"
\next "mrf_decontour"
\prev "mrf_sharp2"
This is a broader sharpening filter.  This is particularly well suited for
higher resolution images, and will have a more dramatic effect than Sharpen I
or Sharpen II.

See Also: \{ \nw \ts b \tc 15 \link "Sharpen I " "mrf_sharp1"
          \link "Sharpen II" "mrf_sharp2"
\}

\endnode

\node "mrf_decontour"
\title "DeContour"
\next "mrf_contour"
\prev "mrf_sharp3"
This tool attempts to increase the number of colors in the image.  DeContour
measures the distance from one color to the next in the specified direction.
When it recognises a \{ \ts i step \} in color, it will replace the \{ \ts i
step \} with a \{ \ts i ramp \} (smooth change) from the first color to the
second color.  For this function to operate there are two requirements, they
are:

\{ \ra \image "Clips/Bullet2.pic" l 10 \cap
The difference in color between the two pixels, in the selected direction,
must be larger than the smallest step possible. \}

\{ \ra \image "Clips/Bullet2.pic" l 10 \cap
The two pixels must have at least one pixel between them that is the same
color as one or the other of the end pixels. \}
\endnode

\node "mrf_contour"
\title "Contour"
\next "mrf_remsmear"
\prev "mrf_decontour"
This tool allows you to reduce the number of colors in each pixel.  The
number of independent color levels will be reduced to the number you select.
If the selection slide gadget is all the way to the left, there will be no
effect.  As the slide is moved to the right the number of colors in each
pixel will decrease.  This will not have the same effect as reducing the
number of bit planes, because it is still possible to have 16 million values
in a pixel.

The effect created here is often called \{ \ts b Mach Bands \} .  Mach Bands
are the \{ \ts i steps \} of colors that you can see when an image is not
displayed in enough colors to cause a smooth transition between colors.

\endnode

\node "mrf_remsmear"
\title "Remove Smear"
\next "mrf_dedith"
\prev "mrf_contour"
This will cause a \{ \ts i smeared \} look in the selected region.  This tool
replaces each pixel, in the selected area, with the average of its self, and
all of it's nearest neighbors.  All data for the smear effect will come from
the \{ \ts i un-smeared \} image.  If the pixel is all ready smeared it will
not be used, but the original value will be used for the smear calculation.

\endnode

\node "mrf_dedith"
\title "DeDither"
\next "mrf_rempix"
\prev "mrf_remsmear"
This will eliminate \{ \ts i some \} of the dithering effect in the selected
region.

\endnode

\node "mrf_rempix"
\title "Remove Pixel"
\next "mrf_remstreak"
\prev "mrf_dedith"
This function works to remove isolated pixels.  It will examine the
eight surrounding pixels for each affected pixel.  In this eight pixel group
the minmumum (darkest) and maximum (brightest) pixel will be determined.
Then the affected pixel will be examined to determine if it is in the range.
If it is not in the range the pixel will be removed.

The provided slide gadget allows you to select how the pixel will be
replaced.  When the slide is set to 0%, the far left, the pixel will be
replaced with it's original value.  If it is set to 100%, the far right, the
pixel will be replace with an average value of the surrounding eight pixels.
If the control is set to 50%, then the pixel will be replaced with 50% from
the average, and 50% from the original.

\endnode

\node "mrf_remstreak"
\title "Remove Streak"
\next "mrf_remchunk"
\prev "mrf_rempix"
This function is very similar to the \{ \ts i \tc 10 Remove Pixel \} tool.
This will remove streaks instead of pixels.

See Also: \{ \nw \ts b \tc 15 \link "Remove Pixel" "mrf_rempix"
          \link "Remove Chunk" "mrf_remchunk"
\}

\endnode

\node "mrf_remchunk"
\title "Remove Chunk"
\next "mrf_remfeature"
\prev "mrf_remstreak"
This function is very similar to the \{ \ts i \tc 10 Remove Pixel \} tool.
This will remove chunks (groups of pixels) instead of pixels.

See Also: \{ \nw \ts b \tc 15 \link "Remove Pixel" "mrf_rempix"
          \link "Remove Smear" "mrf_remsmear"
\}

\endnode

\node "mrf_remfeature"
\title "Remove Feature"
\next "mrf_thin"
\prev "mrf_remchunk"
This tool attempts to remove a feature completely from the image.  The
feature \{ \ts i must \} be totally surronded by colors that would be desired
for the interior of the region.

For example, you could remove a bird from the sky, if the bird was surrounded
by blues.  You would not wish to remove any object that is \{ \ts i connected
\} to another object.

\endnode

\node "mrf_thin"
\title "Thin"
\next "mrf_thicken"
\prev "mrf_remfeature"
This tool will actually take elements of the image and make them narrower.
Pixels will be removed at the edges of the region.

\endnode

\node "mrf_thicken"
\title "Thicken"
\next "mrf_lowpass"
\prev "mrf_thin"
This tool will actually take elements of the image and make them wider.
Pixels will be added at the edges of the region.

\endnode

\node "mrf_lowpass"
\title "Low Pass"
\next "mrf_highpass"
\prev "mrf_thicken"
This tool processes the image as if an electronic filter were used on a
camera.  This generally produces a smoother effect.

\endnode

\node "mrf_highpass"
\title "High Pass"
\next "mrf_antialias"
\prev "mrf_lowpass"
This tool processes the image as if an electronic filter were used on a
camera.  The transformation is based on a middle grey color.  Any area that
is darker than the middle grey are changes in the darker direction in the
original image.  Any area that is brighter than the middle grey are changes
in the brighter direction in the original image.  The darker or brighter the
new pixels, the more change there was in the original image.

\endnode

\node "mrf_antialias"
\title "Anti-Alias"
\next "mrf_connect"
\prev "mrf_highpass"
This will take the selected region and smooth it so that sharp color changes
become smooth ones.  The method used is a large matrix convolution (8x8).
This, in general, will produce a better effect.

\endnode

\node "mrf_connect"
\title "Connect"
\next "mrf_disconnect"
\prev "mrf_antialias"
This will take areas that are lighter than their surroundings, and connects
them. These areas must be close together, and nearly the same color.

\endnode

\node "mrf_disconnect"
\title "Disconnect"
\next "mrf_ntsclim"
\prev "mrf_connect"
This will take areas that are darker than their surroundings, and connects
them. These areas must be close together, and nearly the same color.


\endnode

\node "mrf_ntsclim"
\title "NTSC Limit"
\next "mrf_ntscfilt"
\prev "mrf_disconnect"
This will take the specified region and modifie it to conform to the NTSC
Chrominance modulation specification.  This tool will look for areas of high
saturation, and any area that is luma will not be involved in this step of
the evaluation.  After the area is evaluated the maximum saturation of any
color is reduced to 82%.  Any luma portions of greater amplitude will be
scaled down proportionally, but you could have values of 100% dependent on
the selected area.  This tool will ensure that the image can be encoded
without exceeding the ability of the NTSC signal to carry the color
information, and without color distortions.

It is important to note that this \{ \ts i does not \} filter transitions.
You could still have a pixel that goes from red to blue, which would not be
properly encoded by NTSC.  In order to produce a completely NTSC compatible
image, you will need to use the \{ \ts i \tc 10 NTSC Filter \} tool on the
same region.

\endnode

\node "mrf_ntscfilt"
\title "NTSC Filter"
\next "mrf_reduceglare"
\prev "mrf_ntsclim"
This tool will take the specified region and modify it to conform to the
limited bandwidths of both color and luma signals in an NTSC composite
signal.  This process will modify only the region you specify, but the entire
length of the image scan line is used for computation.  For this reason, you
should only apply this effect to an \{ \ts i entire \} region.

The NTSC signal can only change brightness (luma) about 300 times per scan
line, and can only change colors about 100 times per scan line.  It is
important to note that the color and brightness changes are independent of
the number of RGB pixels in the scan line.  Due to this factor, the NTSC
filter will have differing effects on images of different horizontal
resolution.

For example, if your image had 320 horizontal pixels, it would only be
smoothed to 300 changes.  The net effect on the image is less than 10%.  Now,
let's say your image has 786 horizontal pixels.  This process would smooth
the image a great deal.  The image would go from 728 to 300, that amont to
over 50% change in the image.  The same is true for colors, the only
difference is that there can only be about 100 color changes in a scan line
for a NTSC signal.

It is also important to note that the destination software or hardware may
perform this function as well as the \{ \ts i \tc 10 NTSC Limit \}
automatically.  For this reason you may wish to experiment with both the
destination software/hardware and \{ \ts b Imagemaster R/t \} to find the one
that best suits your needs, and will produce the best quality image on your
output device.

\endnode

\node "mrf_reduceglare"
\title "Reduce Glare"
\next "mrf_ziprem"
\prev "mrf_ntscfilt"
This tool locates regions that are extremely white in an image.  The color
satuation is then increased for that area.  This will help to restore color
to those regions.

\endnode

\node "mrf_ziprem"
\title "Zipper Removal"
\next "mrf_userconvalue"
\prev "mrf_reduceglare"
This function is designed to remove inherent errors in the DCTV format.  To
use this tool, you will need to load a non-scaled DCTV image.  Scaled images
\{ \ts b can not \} be used.  A scaled image will have the zipper errors
spread out over several scan lines.  Once you have the image, apply this tool
where the zippers occur.

It is important to remember that some color errors may occur, due to the
image's origin.  These errors will occur where there are sharp edges in the
image.  These images can \{ \ts i generally \} be repaired.

\endnode

\node "mrf_userconvalue"
\title "User Convolve"
\next ""
\prev "mrf_ziprem"
This function allows you to set the values for the nine convolve matrice
positions.  These can be set manually to a random pattern, or you could
select one of the preset shapes.  As an example, if you selected the \{ \tc
15 \ts b - \} (horizontal) shape, horizontal lines would be enhanced in the
image.  A understanding of what convolution means is necessary for optimal
use of this function.

\endnode