Let’s Learn About File Formats

VersaWorks will allow loading of PDF, EPS, TIFF, JPEG, & PS file formats. Add to that the VersaWorks job save files – RVW.  Note that SVG, PNG, & GIF files are not on this list.

Computer graphic files can be separated into two broad categories – files that have vector elements and those that strictly have raster elements.

PDF & EPS Files: Vector elements are directed line segments – mathematically described lines that have nodes and direction and when scaled, mathematically enlarge with no degradation in quality.

TIFF & JPEG Files: Raster elements are much like the squares on a piece of graph paper. They can be made colored or grayscale. When raster files get enlarged – the squares are enlarged and eventually will be visible as “jaggies” or stair steps that reveal how they are created. If there are enough raster elements – the jaggies are not seen and the file will look fine.

Another way to separate the two categories – vector and raster – is by describing them as output independent and output dependent. Once a vector file is created it can be printed at any output size with no quality loss. The quality is independent of the output size. Scale the image up and the vectors can be scaled and the image looks the same. Raster files must be predetermined in size and resolution in order that quality is assured. If we enlarge a raster file – the squares get enlarged until our eyes see the squares. We call that type of file output dependent – the quality is dependent on how large the file is printed. Early sign cutters used vectors (fonts are made of vectors) as the command to tell the cutter what size and direction to cut.

But it gets complicated: In the early days, EPS files were strictly vector files – our cutters cut. Now a raster element can be placed into an EPS files and printed. Once a raster element is placed in an EPS file it is no longer output dependent and care must be taken to assure the raster portion meets the quality demand of the job.

EPS vs. PDF: EPS files are Encapsulated Postscript. Postscript was the programming language that told the printer how to image the file. Early on it was used for typesetting. Fonts were loaded as a set of vector characters and Postscript knew how to image the fonts. Graphic software allowed the other elements to be drawn as vectors as well. Way back in the day we had crappy printers that printed at low resolutions – at best 300 dots per inch – dpi. Output Bureaus had very expensive imagesetters that could print at 2400 dpi. Postscript allowed up to design with our crappy Postscript printer and then pay the Output Bureau to image it at high resolution on their expensive Postscript printer – often to film used to make printing plates. Postscript was the glue that made this work. Both printers printed with the same imaging software. EPS files were little snippets of code representing images that we could embed into our layout software – usually they were graphic elements like logos among the text.

PDF Workflow: PDF files are Page Description Format files. While EPS represented graphic elements – PDF represents multipage documents. This format was initially used to package documents for the web – but grew more robust to allow for whole magazines. PDF has largely replaced EPS in the graphic industry – although they share a lineage. While an EPS file is all the graphic elements in a file – the PDF allows for multipage documents and the page size – or artboard – is used as the edge of the graphic page. Incremental improvements to programs like Illustrator have added multiple pages and have slowly eliminated the need for EPS files – although they are still useful when making large documents with small graphic elements.

As you will see below – vector files are very small and light duty files and have been the backbone of the large format industry.

JPEG & TIFF Files: JPEG files are Joint Photographic Engineering Group files. They are a format used by photographers that allow for large files to be compressed and made smaller. The compression is “lossy” meaning that every time they are compressed the quality of the photo gets worse. We can change the amount of compression from a little to a lot, but if we keep recompressing we will start to see remnants of the compression algorithm. TIFF files are Tagged Image File Format. TIFF files are raster files for saving photographs as well. TIFF files tend to larger for the same data as JPEGs because they either use no compression or “non-lossy” compression. They essentially compress all rasters (squares) of the same color to a (color)X expression. If 100 blue squares are in a row – LZW compression will make that area blueX100 – saving 99 data points.

Be aware that the way files are saved can wreak havoc on a RIP like VersaWorks. VersaWorks accepts TIFF LZW compression, but not others. VersaWorks does not accept JPEG progressive scan options in the save dialog box choices.

DPI & PPI: Dots per Inch [DPI] – is the number of dots a printer can resolve per inch. It represents the quality of the output device. These days we are at 600 dpi to 2400 dpi and more. Most how laser printers are between 300 dpi and 600 dpi. Roland printers are 720 dpi to 900 dpi. This is the dot size of each of the printers colors. With an 8 color printer that is 900x900x8 colors per inch. Lots of Dots. Pixels per Inch [PPI] – is the proper way of desribing a raster file. It describes how many pixel (picture Elements) are in each inch of the file. The number of PPI for a file is directly related to the number of inches in width and height the file has.

A 72 PPI file 3″x3″ has very few pixels compared to a 72 PPI file at 96″x120″. This difference will also be very identifiable by the file size. The 3″x3″ will be very small. Also, the color mode or color depth of each element has an effect on the file size as well. A pixel can be black or white – 1 bit depth, or grayscale – 8 bit depth, or full color – 24 bit depth. The 1 bit file is black and white and will need to be the PPI of the printers DPI to not see squares. The grayscale image gets a bit more complicated and would involve a search about linescreens. The 24 bit file is basically a RGB file with one grayscale image for each color channel. A 32 bit file is essentially the same idea, but in a CMYK color space. A 24 bit file is three times the size of that file made into an 8 bit grayscale files. That grayscale file will be eight times the size of that file made 1 bit black and white.

Making Raster Files Good Enough Quality: Because raster files are output dependent we need to reverse engineer the file before we begin to design. A four foot by eight foot sign made as a raster file will need to be at least 150 PPI at that size to assure the file will print reasonable at its full size. An RGB file with this setting is almost 300 MB. When computers were not as fast as today – this way of designing was out of the question. Today we might be able to get away with this if we had a very good processor and a lot of RAM to avoid writing to disk all the time. When filters are used the time to process could be very slow. Using 150 PPI is actually a minimum for most closer viewing distances. While it might work on a box truck, it would start to show jaggies on a tradeshow sign. This is the reason we in the sign industry predominantly focus on vector graphics.

I hope this little trip down memory lane helps you a bit in understanding the lineage and use of the files in our industry. Email me if you have questions or corrections: