To use the internal PDF converter to convert PDF files to TIF, specify a $convert_options keyword of QPDF or include Q in the CVTOPTS environment variable. This renderer handles both portrait and landscape pages and currently converts to TIF without using the document converter print driver (FFCVRT). It is usually faster than either Acrobat or Ghostscript and has more options for rendering documents with images and halftones to produce smaller output TIF files. We recommend testing this option with a range of sample documents because it is possible that very complex PDF documents may be rendered differently from Acrobat.
The Internal PDF converter can also be used for conversion, via PDF, of Microsoft Office documents, image files, and HTML documents to TIF. Please use the links for details.
This conversion method is recommended only for documents of letter, legal, A4 or B4 width (or height, for landscape pages). Other widths will either be truncated (if larger) or scaled depending on the SCALED option.
Documents are first rendered to a color bitmap and then reduced to monochrome for faxing using a standard dithering method. If you are aware of a specific document type that is to converted for a specific job administration job, you can add a QPDF_OPTIONS variable which will be read from the FS file to obtain conversion options to optimize the rendering.
A standard set of QPDF_OPTIONS can also be specified in a configuration file variable, whose value will be overridden by (not augmented by) a value defined in the USR, UJP or FS file.
|A new renderer is in use from build 126.96.36.199, which is implemented in a separate DLL. The DLL is named PDFRendererDLLnnnn.DLL , where nnnn is a version number, and the DLL must be located in the COPIA program files sub-folder. This renderer offers significant quality improvements, particularly in patterns, shading and antialiasing. We have found this renderer to be generally more reliable and consistent than the original, and it produces slightly smaller images, resulting in a faster fax transmission time. If your documents typically contain fine horizontal and vertical lines, or very small text, we recommend darkening the image slightly using for example GAMMA0.8 in QPF_OPTIONS below. The old renderer (which uses GDI+) is still available if you rename the dll or use QPDF_OPTIONS keyword NoDPLR.|
|In addition to the new renderer, the dithering of color and grayscale images to monochrome fax has also been updated, so that the quality and range of options now meets or exceeds that available from the print driver. The DITHERn selection numbers are now as shown below. To obtain the old dithering algorithms, which prioritize smaller size over TIF quality, add 80 to the original dither values as shown below.|
QPDF_OPTIONS can also be specified on the command-line when CVSINGLE is run standalone. The variable can also be read from the system environment for some applications, but this is not recommended because some uses require it defined as a system variable.
Option keywords must be separated by white space. Currently the available QPDF_OPTIONS keywords can be grouped into three separate types:
Fax format options controlling the output format:
|LOWRES||Creates the TIF in low-resolution format. This is advisable only for plain-text PDFs.|
|FAXMH||Creates the TIF file with MH encoding (unless FAXMR is also used). The default is MMR (which is not supported on some TruFax boards).|
|FAXMR||Creates the TIF file with MR encoding. The default is MMR (which is not supported on some TruFax boards).|
|SCALE||Scales the document to the full width of a standard fax page (1728 pixels at 7.7 pels/mm). Because A4 pages would be 1617 pixels wide and Letter pages 1662 pixels wide, this option will also make the resulting fax longer than a standard page (297mm or 11" respectively).|
|SCIMG||For image (JPG, GIF, PNG) conversions only, scales the image to full page in proportion to the image dimensions. The page size used is as specified by environment variable DEFAULT_CONVERTER_PAGE and variable CONVERTER_PAGE.|
|B4A4||Scales a B4 document to produce A4 output. This option also implies the SCALE keyword.|
|A3A4||Scales an A3 document to produce A4 output. This option also implies the SCALE keyword.|
|HEIGHTn||Scales the output image vertically to a scan row height of n (measured in high-res scanlines). If n is omitted, the image is scaled to the same height as the input page, to the next multiple of 16 pixels. Note that the image aspect ratio may not be exactly preserved. For A4 input, use 2292 to make the resulting TIF the height of an A4 page (297mm at 196dpi); for US Letter input, use 2156 to make the resulting TIF the height of a letter page (11" at 196dpi).|
|MARGINSt/l/b/r||For HTML conversions, sets the top, left, bottom and right margins (in 200dpi pixels) to be used when rendering the HTML inside a standard fax width of 1728 (A4/Letter), 2048 (B4), or 2432 (A3). The default is to set the horizontal margins to center the image in the fax width and use a margin of about 6 pixels above that specified in the HTML, if any. Increasing the margin in one direction only will alter the aspect ratio of the image, as will setting the margins greater than the difference between the document width and the standard fax width. The four values for this option may be specified as a default in the Windows environment HTML_MARGINS as for example: set HTML_MARGINS=0/60/0/60.|
|GDIPLUS||For HTML conversions, GDI Plus is used instead of GDI to render the HTML.|
|ANTI||Rotates landscape pages anti-clockwise: the default is clockwise.|
|CHECKBLACK||Inverts the image if the top and bottom few scanlines are predominantly black.|
|NoDPLR||Uses the original renderer in use before release 188.8.131.52. The original renderer is also used if the file PDFRenderer1115.DLL is not present in the COPIA program files sub-folder. Note that the number in this DLL name will change from time to time.|
Image transformation options which may appear more than once in the option string: they are processed in the order in which they appear:
|SHARPENn ||Applies a sharpening algorithm to the whole image. The value n (from 1 to 100) selects the strength of the algorithm.|
|BRIGHTNESSn||Increases (positive values of n) or decreases (negative values of n) the brightness (whiteness) of the image. The value n can range from -100 to 100.|
|GAMMAn ||Increases logarithmically (values of n above 1.0) or decreases (values of n below 1.0) the brightness (whiteness) of the image. The value of n can range from 0.1 to 7.0. If your documents are likely to contain single-pixel or other very thin lines, we recommend darkening images slightly, using for example GAMMA0.8.|
|EDGE||Applies an algorithm to make solid areas more uniform and enhance edges. This will usually produce better results, especially with text, than the SHARPEN algorithm but may take slightly longer to apply.|
|DESPECKLEnm||Despeckles the image. The value of n must be 3, 5 or 7 to specify the size in pixels of the surrounding square to be examined for each pixel, and the value of m must be 1, 2 or 3 to controls how aggressive the despeckling should be. The default is 52. Smaller values of m may be necessary if image elements are removed which should not be. Despeckling is always done last.|
Image dithering options which are used to convert the transformed image to a one-bit-per pixel monochrome fax format. Only one of these options should be specified.
|DITHERn||Selects a specific dithering method (applied after Sharpen, Brightness or Gamma, and Edge processing), where n is a number as follows:|
|1||Nearest color matching without error correction. This just converts lighter colors to white and darker to black.|
|4||Simplified Floyd-Steinberg Error Diffusion dithering|
|5||True Floyd-Steinberg Error Diffusion dithering. This is the default.|
|6||Jarvis, Judice & Ninke Error Diffusion dithering|
|7||Stucki Error Diffusion dithering.|
|8||Burkes Error Diffusion dithering|
|9||Sierra-3 Error Diffusion dithering.|
|10||Sierra-2 Error Diffusion dithering.|
|11||Sierra-2-4A Error Diffusion dithering.|
|12||Atkinson HyperDither Error Diffusion dithering|
| ||To use the original dithering algorithms use 81: nearest, 82: Floyd-Steinberg, 83: Stucki, 84: Sierra, 85: JJN, 86: S&A, 87: Burkes. These use different threshold values and result in reduced quality, along with somewhat smaller file sizes.|
|DITHERLn||This operates as DITHER for values of n from 4 to 12, but also converts to low resolution. This option provides significantly higher quality for graphical and grayscale originals which you want to send at low-resolution, because it converts to low-resolution while dithering. However if your originals are principally text-based, then you should use the option LOWRES, which is optimized for text on portrait-mode originals.|
|PHOTO1||Optimized for color photograph content. Currently this setting selects Burkes Error Diffusion dithering.|
|MAP1||Optimized for street maps with light-colored backgrounds and black text. Most backgrounds are converted to white without dithering, which significantly reduces the transmission time of the resulting TIF file.|
|PARM1=n||As a modifier for the MAP1 option, you may use a value of n to adjust the brightness above which light-colored backgrounds are converted to white. The range of n is 0 to 255 and the default value is 224.|
|PARM2=n||As a modifier for the MAP1 option, you may use a value of n to select the center of a brightness range which is to be forced to black output. The range of n is 0 to 765 and the range size is fixed at 65. This option is ignored if the image contains a large block of this brightness range, and also does not affect yellow/orange colors. The default is to treat no brightness range in this way.|
Note that the more image manipulation you do, and the more complex the image, the longer the conversion will take. However in most cases the image will be processed in only a couple of seconds per page.
If you have a significant application which you think could benefit from a 'tweaked' conversion method, please contact Copia support.