@DATABASE GfxCon
@AUTHOR Dirk Farin
@NODE MAIN "contents"
GfxCon - image format converter
=================================
V1.7 (24 June 1995)
Dirk Farin
Introduction
------------
@{" what is GfxCon ? " link FEATURE_POSTER}
@{" what's new ? " link NEWS }
@{" Remarks " link REMARKS }
@{" bugs ? " link BUGS }
installation
------------
@{" requirements " link REQUIREMENTS}
@{" installation " link INSTALLATION}
usage
-----
@{" entering the filenames " link FILENAMES}
@{" selection the output-format " link SELECT_OUT_FORMAT }
@{" modifying size and orientation " link GEOMETRY }
@{" modifying colors " link COLORS}
@{" the conversion-process " link CONVERT_PROCESS}
@{" showing information about the input image " link INPUT_INFORMATION}
@{" using GfxCon as shell-command NEW" link SHELL}
reference
---------
@{" supported formats " link SUPPORTED_FORMATS}
@{" problems that may occur " link PROBLEMS }
@{" about the program " link PRG_INFOS}
@{" copyright " link COPYRIGHT}
@{" (registration) " link REGISTER}
@ENDNODE
@REMARK =========================================================================
@NODE FEATURE_POSTER "the purpose of GfxCon"
what is GfxCon ?
================
GfxCon makes it possible to convert images to the most popular image
formats in a user-friendly way. Simplicity of usage and a high quality
have been the primary goals while developing GfxCon.
Besides this it is possible to modify the image in many ways like
changeing the resolution, contrast of the image or numerous color-effects.
not implemented:
- showing the picture
@ENDNODE
@NODE NEWS "news"
news
----
version 1.7
- RGB-Raw input, output and brightness, contrast may be used with
the Shell-interface
- output format = input format by default in the shell
- GIF89a transparent color is now saved and loaded too
(not only background color)
- Guru with bad input files removed
- progress-information in the Shell may be switched off
version 1.6
- two ways to proportionally resize an image (shell only)
- it's now possible to query the image-infos from shell
version 1.5
- may now be used as shell-command
- saves background-color if known (ILBM<->GIF)
can only be changed by hand with shell-parameter
- runs under Kick 1.2 :) (shell only)
- bug fixed: changing image-geometry and saving as GIF at once was faulty
version 1.4
- bug in PCX-loader fixed (images with odd number of bytes per row)
- bug in PCX-saver fixed (BytesPerRow in PCX-header was set incorrectly)
- 68020-version doesn't need 68881 any more (uses IEEE-libraries now)
version 1.3
- HAM6 is working again
- virtual memory improved (new Tooltypes)
- TIFF-RGB-images and bug-fix
version 1.2
- HAM8-mode (finally)
- Targa-CLUT-images
@ENDNODE
@REMARK =========================================================================
@NODE REMARKS "remarks"
I've written GfxCon a long time ago (before the release of WB2.1).
Therefore some things like localisation have been solved in a different
way (with tooltypes instead of using the WB2.1-catalogs). But this allows
even users of WB1.2 to use the feature of localisation :)
For the same reason GfxCon does not support the use of datatypes
(introduces in Kick3.0).
Since upgrading the program to use all these would exceed my free time at
the moment, it is unlikely that an update will be released. But a
completely new program (of course much better than this) may be possible
in the near future.
@ENDNODE
@REMARK =========================================================================
@NODE BUGS "bugs"
known bugs:
- sometimes the font in one of the button-gadgets changes
to a smaller one "without any sensible reason". But
this does not affect functionality in any way.
If you discover a bug, please let me know. I will try to fix it, but I
cannot promise to succeed, as this program is nearly 2 years old by
now (!) and has not been looked at in the meantime.
@{"my address" link PRG_INFOS}
@ENDNODE
@REMARK =========================================================================
@NODE REQUIREMENTS "requirements"
requirements
============
This program requires to run:
Kickstart 1.2
20000 Bytes stack
Recommended:
At LEAST a 68020 !
@{"4 MB memory" link MEMORYUSAGE}
@{"hard-disk " link VIRTUAL_MEMORY}
@ENDNODE
@REMARK =========================================================================
@NODE INSTALLATION "installation"
installation
============
To install this program simply move the drawer containing
the program (or only the program) to a place of your choice.
To configure the program to your system, it is recommended to
change the @{"tooltypes" link TOOLTYPES} to your needs.
@{B}Thats all.@{UB}
@ENDNODE
@REMARK =========================================================================
@NODE SHELL "using as shell-command"
GfxCon can now (since version 1.5) be used as shell-command. This works even
with Kickstart 1.2 (or lower :-)
usage:
GfxCon inputfile [TO outputfile] [FORMAT f] [COLORS c] [SIZE x y] [FLIPX] [...]
If you don't specify an outputfile, the old suffix will be replaced with the
suffix of the output-format.
parameter:
TO destname set destination filename
FORMAT output format { ILBM,GIF,PCX,JPEG,Postscript,RGB-Raw }
COLORS n use 'n' colors max.
DITHER use floyd-steinberg dithering
FASTDITHER use fast floyd dithering
BACKGROUND n use color 'n' as background color (only ILBM and GIF)
SIZE x y resize image
RESIZE f proportionally resize image with factor f (>0)
BOXFIT x y proportionally resize image to fit into the box specified
QUALITY q set JPEG-quality
CLUT select IFF-CLUT mode\
HAM IFF-HAM mode \_ use only with 'FORMAT ILBM'
HAM8 IFF-HAM8 mode /
24BIT IFF-24bit mode/
INVERS invert colors
BW create black and white only
GRAYSCALE produce a grayscale (not simply BW!) output
NORED remove red
NOGREEN remove green
NOBLUE remove blue
BRIGHTNESS n change brightness by n (255 to shift black to white)
CONTRAST n change contrast by n (percent)
FLIPX flip horizontally
FLIPY flip vertically
ROTATELEFT rotate left (90 degrees)
ROTATERIGHT rotate right (90 degrees)
GREEN n set name of green-RGB-raw-file to n (RGB-Raw only)
BLUE n set name of blue -RGB-raw-file to n (RGB-Raw only)
INPUTWIDTH n force input-width to be 'n' pixels (RGB-Raw only)
INFO show image-information only
NOPROGRESS don't show how much the conversion has progressed
HELP / -help / -h / ? / -? this help-page :-)
remarks:
The shell-interface becomes much more powerful if combined with the 'foreach'-
program (by me; available e.g. on aminet 'util/cli/foreach.lha'.
Examples:
Convert all IFFs to GIFs: foreach #?.iff GfxCon & FORMAT GIF
Produce a series of JPEGs with different quality-settings:
foreach {30,50,70,90} GfxCon Pic.iff TO Pic&.jpg FORMAT JPEG QUALITY &
The shell-functions and the Kick1.2-support have been implemented after
the program has been finished with a lot of dirty tricks and is therefore
everything but perfect. It is mainly to offer these capabilities until
GfxCon2.0 is finished and to finally give a graphics-converter to those
poor Kick1.2 people :)
NEW (v1.7) How to load RGB-Raw-files: type the red-file as input file and
add the two other files with the GREEN and BLUE parameter.
Example: To load the files "pic.red", "pic.green", "pic.blue", type
GfxCon pic.red GREEN pic.green BLUE pic.blue ...
If you have to override the width, GfxCon guesses for the width,
use the INPUTWIDTH parameter.
The suffices ".red", ".green" and ".blue" are used for RGB-Raw
file output. If you specify an output filename with "TO", these
suffices are appended too.
caution:
Kick1.3 users CANNOT use the Postscript-output. Sorry!
@ENDNODE
@REMARK =========================================================================
@NODE FILENAMES "entering filenames"
entering filenames
==================
general case
------------
There is a box 'Load' at the top left and a box 'Save' at the right to
enter the filenames for the input- and output-filename respective. Both
boxes can be used in the same way.
To enter a filename simply type it into the string-gadget or choose a file
using the filerequester which can be opened by pressing the button to the
left of each string-gadget.
If you entered the input-filename (and have pressed RETURN or the TAB-key
or you choose it with the filerequester, GfxCon tries to determine the
format of the input-file and enters the name of the format into the
popup-gadget below the string-gadgets.
RGB-files (raw)
---------------
If GfxCon doesn't recognise the input-format, it supposes that it is a raw
@{"RGB-file" link RGB} (consisting of three files for red,green and blue).
As three input-files are needed to load a RGB-raw-image, GfxCon enables
the two string-gadgets below the normal one to let you enter the filenames
for the green and blue part of the image. The character at the right of
each string-gadget shows which filename to enter (the red-part at R, the
green at G and the blue at B of course).
@ENDNODE
@REMARK =========================================================================
@NODE SELECT_OUT_FORMAT "selection the output format"
selection the output format
===========================
selecting the file-format
-------------------------
You can select the format using the "popup-gadget" in the 'Save'-box.
selecting the color mode (only ILBM)
------------------------------------
As the @{"ILBM-format" link ILBM} supports several color modes (colortable, 24 bit,
HAM6, HAM8), you may choose the color mode with the popup-gadget 'colormode'.
setting number of colors
------------------------
The number of colors allowed in the output-file may be specified with the
gadget 'colors'. You may also enter unusual numbers like 28 or 53. As most
image-formats only allow numbers that are a power of 2, GfxCon rounds this
number up to the next possible number of colors and leaves the remaining
colors unused. (If you specify more colors than the format allows, like
512 für PCX, GfxCon uses the maximum colors possible with that format).
@ENDNODE
@REMARK =========================================================================
@NODE GEOMETRY "modifying size and orientation"
modifying size and orientation
==============================
resizing
--------
Switch the button 'custom' in the box 'size' to on. The two gadgets
'width' and 'height' will become enabled with the original size of the
image in pixel already filled in. Modify these values to the desired ones.
If you enlarge the image and don't want that "big pixels" are created, set
the switch 'interpolated'. The conversion will take longer but the result
will be a smoother image.
NOTE: Interpolating an image creates new colors. It makes no sense to
interpolate a image when the output will only have 2 colors for
example.
NOTE: To enlarge the picture, GfxCon need more @{"memory" link MEMORYUSAGE} than normal.
flipping the image
------------------
you may flip the image horizontally or vertically (or both, what results
in rotating the image 180 degrees) using the gadget 'flip' in the
'transform' box.
rotating the image
------------------
you may rotate the image 90 degrees (counterclockwise) or -90 degress
(clockwise) using the gadget 'rotate' in the 'transform' box.
NOTE: for rotating the image GfxCon needs more @{"memory" link MEMORYUSAGE} than normal.
@ENDNODE
@REMARK =========================================================================
@NODE COLORS "modifying colors"
modifying colors
================
brightness and contrast
-----------------------
At the bottom of the box 'colors' there are two gadgets to change the
brightness and contrast of the whole image. The value for the brightness
must be in the range between -255 and +255 inclusive. The value for
contrast should of course be at least 1%. A value between 1% and 99%
reduces the contrast, a value above 100% enhances contrast.
Brightness is always changed first.
I.e.: output = (input+brightness)*contrast
color effects
-------------
There are several special color effects, which you can choose using the
popup-gadget 'effects' in the 'colors' box:
convert to black-white : choose 'black/white'
invert image : 'inverse'
black-white-negative : 'inserse & bw'
Besides this you can select filters. In the 'effects'-popup-gadget there
are filters in these colors:
red, green, blue, yellow, magenta and cyan.
grayscales
----------
If you want the output image to be colored using a linear grayscale (or
images using only the colors black and white as needed for printing)
choose the item 'grayscale' from the 'effects'-gadget. But remember that
this only works if you are converting to a image with a format using a
CLUT (colortable).
dithering
---------
If you are reducing the number of colors of a picture with many colors
(like JPEG) it is very important to use dithering. If you don't, awful
looking "wrong" colors will stay in the output. The principle of dithering
is to use several pixels of different colors to achieve the effect that
these colors look like the missing one if looked at from a distance.
GfxCon uses at the moment only the 'Floyd-Steinberg'-dithering, that is
very time-consuming but produces very good images.
Select the Floyd-Steinberg-dithering with the gadget 'dithering'.
@ENDNODE
@REMARK =========================================================================
@NODE CONVERT_PROCESS "the conversion-process"
the conversion-process
======================
The conversion can be started by clicking the big gadget 'convert' at the
bottom left. A window will open telling you what GfxCon is doing at the
moment (rotating, saving...) in the lower part and showing how much of
this action is finished with the bar in the upper part of the window.
It may be that another window will open, where you may enter additional
parameters (like the quality of @{"JPEG-output" link JPEG} or the @{"size of the image" link POST_JUSTIFY}
with @{"Postscript-output" link POST}).
@ENDNODE
@REMARK =========================================================================
@NODE INPUT_INFORMATION "showing information about the input-image"
showing information about the input-image
=========================================
With the big gadget 'infos' at the bottom, you can display some details
about the input-image. What is displayed depends on the format of the
input-image and what is really stored in the image. This information can
be very complex (like TIFF) or very scarce (like IMG).
@ENDNODE
@REMARK =========================================================================
@NODE SUPPORTED_FORMATS "supported formats"
supported formats
=================
input formats output formats
------------- --------------
@{" ILBM " link ILBM} @{" ILBM " link ILBM}
@{" LBM " link ILBM} @{" PCX " link PCX }
@{" RGB8 / RGBN " link RGB8} @{" JPEG " link JPEG}
@{" PCX " link PCX } @{" RGB-Raw " link RGB }
@{" IMG " link IMG } @{" Postscript " link POST}
@{" BMP / RLE4 / RLE8 " link RGB8} @{" GIF " link GIF }
@{" GIF " link GIF }
@{" TIFF " link TIFF}
@{" JPEG " link JPEG}
@{" Targa " link TGA }
@{" RGB-Raw " link RGB }
@ENDNODE
@REMARK =========================================================================
@NODE PROBLEMS "problemes"
@NEXT PROBLEMS
any problems ?
==============
There are files called 'GFXCON...' in a directory of my harddisc. Have
they any use?
These files are created if the program needs to save parts of the image
it is working on onto the hard-disc (see @{"virtual memory" link VIRTUAL_MEMORY}).
Normally, these files are automatically deleted after use. But it you
reset your computer while GfxCon is working it doesn't have a change to
do this.
Please delete these files by hand or start GfxCon again to delete these
files automatically.
-----------------------------------------------------------------------------
After starting the program I only get the message 'window does not fit on the
screen'.
GfxCon is font-sensitive. You have selected too big fonts for the
gadgets, which would require a window that is too large to fit on the
screen.
Please select smaller fonts in the @{"Tooltypes" link TOOLTYPES}.
@ENDNODE
@REMARK =========================================================================
@REMARK =========================================================================
@REMARK =========================================================================
@NODE TOOLTYPES "setting the tooltypes"
@PREV TOOLTYPES
@NEXT TOOLTYPES
These tooltypes are used by GfxCon:
LANGUAGE - lets you select the language of GfxCon. Currently
possible values are: 'english', 'german' or
'englisch', 'deutsch'
TEMPDIR - As GfxCon uses a big amount of memory, it swaps part
of the image to hard-disc if more memory is needed
than available (see @{"virtual memory" link MEMORYUSAGE}) to let you even
convert pictures that are too large to fit in the
memory.
With this tooltype you specify the directory into
which GfxCon writes these temporary files.
NOTE: This string will be used as prefix to the
filename. Therefore do NOT enter 'sys:t' but
'sys:t/'.
HINT: Insert a command in your s:user-startup to clear
this directory during startup, as files my stay
in there if a strange error occurs or the system
crashes.
STDLOADPATH - Most people have most of their image in the same
directory. To prevent you from entering the path to
this directory each time you start GfxCon, you may
specify a default directory which will automatically
be used if you use the filerequester.
STDSAVEPATH - The same as STDLOADPATH for the output-file.
GADGETFONT - With this tooltype you may specify the font for most
of the gadgets and nearly all output. The format of
this tooltype is: '<name> <size>'. For example: 'topaz
8' or 'helvetica 18'.
TITLEFONT - The font for titles and big gadgets. The format is the
same as that of 'GADGETFONT'.
MINMEM - How much memory (in bytes) GfxCon may not use for its
image-data.
MINMEMBLOCK - GfxCon only allocates memory if there is a mem-block
with at least "MINMEMBLOCK" bytes.
@ENDNODE
@NODE FORMAT_RECOGNITION "automatic format-recognition"
@PREV FORMAT_RECOGNITION
@NEXT FORMAT_RECOGNITION
GfxCon determines the input-format my itself by scanning the input
for unequivocal characteristics. It does @{B}not@{UB} look at the suffix of
the filename (as most programs do).
The format-recognition will be called when you leave to topmost
string-gadget in the 'load'-box.
The algorithm works fine in almost every case. But theoretically it
may be, that the wrong format is guessed. Then you have to set
it manually using the popup-gadget.
@ENDNODE
@NODE MEMORYUSAGE "memory usage"
@PREV MEMORYUSAGE
@NEXT MEMORYUSAGE
This program needs @{B}very@{UB} much memory, as all images are
stored internally using 24 bit. Additional the program needs one
more byte for each pixel for special data.
You may estimate the memory-usage using this equation:
+---------------------------------------------------------------+
| |
| memory usage = height in pixels * width in pixel * 4 Bytes |
| |
+---------------------------------------------------------------+
For example: a 320x 256 image needs 320 kByte
640x 512 1.25 MB
800x 600 ca. 1.80 MB
1024x 768 3 MB
1280x1024 5 MB
exceptions
----------
Some effects (resize, rotate) need two buffers at the same
time. But the memory that is needed is not the memory of
these two buffers added together, as GfxCon knows what parts
of the image it does not need any more. These parts will
be freed when it is not needed any more.
virtual memory
--------------
To make it possible to convert images that do not fit into the
memory, parts of the image, that are not needed at the moment
are written to hard-disc (see @{"virtual memory" link VIRTUAL_MEMORY}).
@ENDNODE
@NODE VIRTUAL_MEMORY "virtual memory"
@PREV VIRTUAL_MEMORY
@NEXT VIRTUAL_MEMORY
If more memory is needed to process an image than available, parts
of the image are saved to hard-disc to use the memory to do other
calculations there.
If a image would for example @{"require" link MEMORYUSAGE} 5 MB to be processed, but you
only have 3 MB free memory, GfxCon will swap at least 2 MB to
hard-disc during processing the image.
These files are of course deleted just when they are not needed any
more, not to fill the hard-disc with useless files.
NOTE: If you reset the computer while GfxCon is using such temporary
files, GfxCon cannot delete this files. Beginning with version 1.3,
GfxCon deletes this files automatically at start.
DO NOT delete any file, while GfxCon is using them!!!
@ENDNODE
@NODE ILBM "IFF - ILBM Format"
@PREV SUPPORTED_FORMATS
IFF - ILBM format
-----------------
load
----
- images with any number of colors, 24 bit, HAM6/8 and ExtraHalfBrite
- the PC-version of the ILBM-format with 16 or 256 colors
- compressed and uncompressed
save
----
- images with any number of colors (even more than 256)
- images with 24 Bit or HAM6/8
- all images are compressed
- not supported at the moment is ExtraHalfbrite
@ENDNODE
@NODE PCX "PCX - format"
PCX - format
------------
load
----
- all PCX-images
- compressed and uncompressed
save
----
- images with 2, 16 or 256 (all possibilities)
- all images are compressed
- PCX-version 3.0 is used
@ENDNODE
@NODE JPEG "JPEG - format"
JPEG format
-----------
load
----
- all JPEG/JFIF images
save
----
- the quality of the output can be altered in the range between 25% and 100%
@ENDNODE
@NODE POST "Postscript"
Postscript - format
-------------------
load
----
- not supported at the moment
save
----
- output is intended to drive a Postscript-printer
- it is not intended as exchange-format (EPS)
- the @{"size and position" link POST_JUSTIFY} of the picture in the output can be specified in 'cm'.
- only black/white output (color images are automatically converted to
black/white)
@ENDNODE
@NODE RGB "RGB - raw - files"
RGB - raw - files ( 3 files with the red, green and blue parts of the image)
-----------------
load
----
- As the size of the picture cannot be determined by looking at the
input-file. Therefore GfxCon guesses the width of the picture. If
it guesses wrong, you can correct the width in pixel.
save
----
- yes
@ENDNODE
@NODE RGB8 "RGB8 / RGBN - format"
IFF - RGB8 / RGBN - format (Imagine)
-------------------------------------
load
----
- RGBN (4096 colors) and RGB8 (truecolor, 24 bit) images
save
----
- no
@ENDNODE
@NODE IMG "IMG - format"
IMG (GEM Image File Format)
----------------------------
load
----
- all IMG-files
- IMG files do @{B}not@{UB} contain a colortable. GfxCon uses the PC
default colors.
save
----
- no
@ENDNODE
@NODE BMP "BMP/RLE4/RLE8 - formats"
BMP / RLE4 / RLE8 (MS-Windows)
---------------------------------
load
----
- all files except 24bit BMP-files (I don't have a 24bit image to test
my routines)
- compression is supported
save
----
- no
@ENDNODE
@NODE GIF "GIF - format"
GIF - format
------------
load
----
- all images (even those with fewer than 256 colors
- raster and interlace format
- only the first picture of the file is read
save
----
- always saves GIF87a / raster-images
@ENDNODE
@NODE TIFF "TIFF - format"
@NEXT TIFF
TIFF - format
-------------
load
----
- uncompressed: black/white (2 colors)
16 colors (with and without bitplanes)
256 colors (with and without bitplanes)
RGB-images (without bitplanes)
- CCITT3 (FAX) compressed: black/white
- RunLength black/white
- LZW and JPEG are not supported.
save
----
- no
@ENDNODE
@NODE TGA "Targa - format"
@NEXT TGA
Targa (TGA) - format
--------------------
load
----
- only truecolor pictures
- 15,16,24 or 32 bit
- uncompressed or compressed with RLE
save
----
- no
@ENDNODE
@NODE POST_JUSTIFY "position and size of Postscript output"
@PREV POST_JUSTIFY
@NEXT POST_JUSTIFY
position and size of Postscript output
======================================
The window to enter the position and size of the Postscript-output
opens automatically. You must specify the position of the top left
corner of the picture measured from the bottom left corner of the
paper in portrait-orientation (may depend on your printer) and the
width and height of the image (in cm).
The values that are by default filled into the gadgets produce a
picture @{B}18cm@{UB} wide and a height according to the aspect of the picture.
@ENDNODE
@NODE PRG_INFOS "about GfxCon"
@PREV PRG_INFOS
@NEXT PRG_INFOS
author
======
Dirk Farin
Kapellenweg 15
72070 Tübingen
Germany
EMail: farindk@trick.informatik.uni-stuttgart.de
@{"copyright" link COPYRIGHT}
@{"(registration)" link REGISTER}
I would be happy to receive any proposals and bug-reports (EMail
preferred). I will answer to all of these if possible.
If you miss your favourite format (and probably do have any
documentation about it), I will try to implement this format too in
future versions.
This program has been written on an Amiga3000T/25Mhz with 5MB,
Kick2.04 and graphics-card. It has been compiled with SAS/C V6.55.
It has been successfully tested on these systems:
Amiga 3000T / OS2.04
Amiga 3000 / OS2.04
Amiga 2000 / OS2.04
Amiga 500 / OS2.04
Amiga 4000 / OS3.0
Amiga 1200 / OS3.0
... :-))
thanx go to:
Fred Fish for his Fish-discs / -CDs.
Matthew Dillon for his DME-editor.
The maintainers of Aminet
All users of GfxCon who did register or/and sent-bug reports !!!
@ENDNODE
@NODE REGISTER "registration"
@PREV REGISTER
@NEXT REGISTER
(registration)
--------------
This program is Cardware / EMailware.
As this program is very old and not up to date, I think it would be
unfair to sell it as Shareware.
But if you use this program sometimes I would be very happy to
receive a postcard or even an EMail, in which you can tell me, if you
like this program or what could be done better.
Of course I will answer any questions concerning the program.
@{"my address" link PRG_INFOS}
@ENDNODE
@NODE COPYRIGHT "copyright"
@PREV COPYRIGHT
@NEXT COPYRIGHT
copyright:
----------
GfxCon is @{"Cardware / EMailware." link REGISTER}
It may be included in Public-Domain-discs, AmiNet and spread in any
other noncommercial way. No part of GfxCon may be used for commercial
purpose without my permission.
The Graphics Interchange Format(c) is the Copyright property of
CompuServe Incorporated. GIF(sm) is a Service Mark property of
CompuServe Incorporated.
@ENDNODE