The More Options button in the Rasterize Separations activity in Nexus RIP shows a sheet with various options. One of the options is Image Sampling. What is Image Sampling, and what do the other options mean?
The resolution of images used in a job will rarely match the output resolution of the ripped job (for instance, the job may be ripped at 2540 ppi, but the images in the input PDF may be supplied at 300 ppi). So the images need to be re-rendered at the output resolution of the job, and the rip need to use a method to convert each pixel of the input image to the corresponding pixels of the output image. The methods used by TotalRIP, work by considering sets of pixels in the input image and then changing the resolution of the input image. This process is called resampling.
TotalRIP offers three methods:
Nearest Neighbour This is the fastest of the resampling methods, and is the default method. It is the least accurate of the methods, but is perfectly adequate for 99% of jobs. We recommend that this method is used unless the user has a specific reason to use a different one.
Bilinear Interpolation This is a more accurate resampling method than Nearest Neighbour, but is considerably slower. If you are finding that an image contains fine detail that is getting lost when using the Nearest Neighbour method, using Bilinear Interpolation should provide a better result. But at the cost of increased ripping times.
Bicubic Interpolation This is the most accurate resampling method offered by Nexus TotalRIP, but it is considerably slower than the other methods. We recommend that this method is used only when the other two methods have failed to yield good results. If your input images contain very fine lines, especially in patterns, and the result when using Bilinear Interpolation is not good enough, it may be necessary to switch to Bicubic Interpolation. While it gives the most accurate results, for most jobs there will be no discernible difference between using this method and the Nearest Neighbour method. So, while it is tempting to switch to this method for all jobs, we strongly recommend that this is not done. This is because the improvements for most jobs will be negligible, and ripping times will potentially be increased considerably.
In short, we strongly recommend that the Image Sampling method be used, as per the default settings. If the images contained in a job begin to lose fine details when using Nearest Neighbour method, then the other methods should be considered. But remember that ripping will be slower if other methods are selected.