There can be several reasons why a converted image may appear blurred, and these depend on the way the unit is being used:
DOWNSCALING AN IMAGE
If you're changing from a high resolution to a low resolution, you'll never get quite the same image quality on the output. This is because you're 'squeezing' multiple pixels into fewer pixels and thus losing information.
This is especially true when converting from a PC image to a composite video or S-video connection. These signals cannot properly represent a PC image. The graphics seen on TV that may appear clear, on closer inspection, will appear soft with quite fuzzy colour (TV broadcast text overlay colours are chosen very carefully to try and avoid such problems).
PICTURE IN PICTURE (PIP)
This is the same as down-scaling - you'll lose information that was present in the original because you're squeezing more pixels into fewer pixels.
ANALOG RGB TO RGB CONVERSION
Assuming you're not down-scaling (see above), then run an Autoset (Autoset feature) from the Adjust sources menu. Alternatively, manually adjust the 'Phase' option in that same menu. This will help the analog to digital convertor sample in the middle of each pixel rather than at the edges.
DIGITAL RGB TO RGB CONVERSION
Digital data should not suffer any quality loss, unless you're converting the image size (e.g. PIP or DOWNSCALING). However, even upscaling can introduce some artefacts that are unavoidable.
COMPOSITE VIDEO INPUT
These signals can lack image clarity, and thus upscaling them will just make the image bigger and not necessarily clearer. Try adjusting the 'Sharpness' control in the 'Adjust sources' menu.