Which Format is Best?
We are often asked what the best format is to use. Assuming there are no pre-determined requirements, then the answer to this question can be one of the most important decisions made in a project.
Obviously, if there are predetermined requirements, then the best format to use is the one called for in those requirements, unless you can get them changed.
Raster vs. Vector
AutoXChange 2011 will convert to both raster and vector formats. In a raster (or bitmap) format the image is comprised of a large grid of colored dots (called pixels, short for "picture elements"). Since everything has been devolved into individual dots, all semblance of intelligence is lost when converting to a bitmap. Another drawback of bitmaps is that they either have to be very large or they lose detail very quickly. When you zoom in on a portion of a bitmap image the dots get bigger, but the detail is not there. So instead of a nice smooth circle, you will see a very jagged representation of a circle. As you continue to zoom in more and more detail is lost.
In a vector format the image is comprised of geometrical elements like lines, circles, text and line strings (called polylines). While, depending on the format, some intelligence may be lost, the majority of intelligence is still present in the resulting image. A line is still a line, not a series of dots. Tailor Made Software has always tried to maintain as much intelligence as possible in the resulting drawing, so, unlike some of our competitors, text remains text when we can, circles remain circles and don't become polylines, etc.
The main advantages of vector formats are size and resolution. As mentioned above, when you zoom in on a raster you see grainier and grainier set of dots. When you zoom in on a vector image you continue to see a smooth circle or line or whatever. You no longer have to produce a huge drawing size just to maintain a reasonable degree of resolution like you do in a bitmap.
Types of Vector Formats
The various vector formats have far fewer inherent advantages or disadvantages than the various raster formats do. For raster formats most viewers will display a wide variety of raster formats. For vector formats most systems will only display a few formats and may really prefer one. Therefore the choice of which vector format to use is really determined by the system that will use the resulting image, and not by any inherent advantage of one format over another.
Are there differences? Sure, but they are not usually overwhelming. Uncompressed PDFs or AI files are huge compared to other formats, but compressed PDFs are very reasonable in size. The same can be said for DWFs, although binary DWFs will be smaller than the equivalent uncompressed PDF.
PDFs, and its PostScript-based cousins like Adobe Illustrator, do have one unusual quirk: they do not have a circular arc or circle element. Instead all curves are made from Bezier curves. Mathematically this makes things simple, but it does mean larger files with elements that can be less precise than normal circular arcs (Beziers can approximate circular arcs, but can also "wobble" a little in tracing the circular arc path).
Types of Bitmap Formats
AutoCAD drawings only have 256 colors defined in them. Any format that defines more than 256 colors is just wasting space. As a result JPEG, which uses either 24 or 32-bits and allows the definition of millions of colors, is a very inefficient format for rendering CAD drawings. GIF or PNG, both of which are 8 bit formats allowing 256 colors are great for color images of CAD drawings. Now that the UNISYS patent has expired there is no real reason to use PNG instead of the more popular GIF.
However, if you do not need to retain colors, then the best raster format to use (if you can) is TIFF Group IV. This format is a highly compressed format, but is only black and white. Not all viewers will display TIFF files so you may be forced to use GIF instead.
A second reason not to use JPEG is due to how JPEG files are compressed. JPEG uses what is known as "lossy compression". Basically it gets large file size reductions by tossing out some colors and blurring areas together. This works great in multi-million color pictures where the human eye can't really distinguish that many colors in such a small space. It works very poorly for line drawings where black lines are "smudged" with large white areas resulting in fuzzy gray lines. The only way to get around producing fuzzy JPEG drawings is to create non-compressed JPEGs. However, the resulting file sizes are huge. There is no good reason to use JPEG for CAD drawings (unless the customer requires them and you can't change their mind) so stick to GIF or TIFF Group IV if you must use rasters.