To remove grain or visual noise, use the
Remove Grain effect. This effect uses sophisticated signal processing
and statistical estimation techniques in an attempt to restore the
image to how it would look without the grain or noise. While many
techniques, such as applying a mild Gaussian Blur effect or the Median
effect, reduce the visibility of noise in an image, the tradeoff
is an unavoidable loss of sharpness and highlights. The Remove Grain
effect, in contrast, differentiates fine image detail from grain
and noise and preserves the image detail as much as possible.
The Remove
Grain effect provides several options to precisely balance the reduction
in noise and the amount of sharpness retained in the image. Additionally,
the Remove Grain effect can analyze the differences between frames
to further improve noise reduction and sharpness; since this process operates
over time, it is called temporal filtering.
Note: Good degraining depends on good noise sampling. The results
of the automatic sampling depend on the image content and noise
type. You can also change the number, size, and position of the
samples to get the best results for a particular image.
The
Temporal Filtering controls of the Remove Grain effect use a statistical algorithm
to blend the current frame with previous and next frames. These controls
are especially effective in removing compression artifacts from
DV or video footage. Because temporal filtering works on the basis
of differences between frames, it is useful only for sequences.
To properly evaluate the results of this filter, the sequence
must be viewed in real time, either with a RAM preview or by viewing
a movie rendered to a file.

To increase the speed of the Remove Grain effect
preview, adjust the Remove Grain controls in order in the Effect
Controls panel. Specifically, the most efficient workflow is to
find effective degraining settings first and to adjust the last
three controls last.
This effect works with 8-bpc and 16-bpc color.

Original (left), and with effect applied (right)