Limiting noise reduction – Blackmagic Design DaVinci Resolve Advanced Panel User Manual

Page 239

Advertising
background image

CHAPTER

COLOR

COLOR

239

9

NR Radius, How Low Should You Go?

Lower NR Radius settings can dramatically improve the quality of high detail regions in shots
where you’re using aggressive noise reduction, but it’s not necessary to always jump to the
lowest value that provides the highest precision. In many cases, when evaluating an image
that you’re applying noise reduction to, you may not actually be able to perceive the additional
quality, and you’ll waste processing time on an unnecessary level of correction.

It’s a good idea to make sure that you’re evaluating the full-frame image on a large enough
display to see the noise you’re working on within the viewing context of the intended audience.
Zooming really far into a clip while applying noise reduction may encourage you to use higher
quality settings then are necessary, because an excessively enlarged detail of an image lets
you see subtle changes that you wouldn’t notice at actual size.

Limiting Noise Reduction

As with any other correction in the Color page, noise reduction can be limited using HSL Qualification
or Power Windows. This means you can focus your efforts on reducing noise in the most problematic
areas of an image (for example, in shadows and background regions), while sparing elements that you
don’t want to affect (such as faces or better-lit areas of the image).

Furthermore, you can use noise reduction in lieu of the Blur settings to perform a subtler form of
complexion smoothing, using the HSL Qualifier to key an actor’s skin tone for targeted noise reduction.

Controlling the Order of Operations for Noise Reduction

You can apply noise reduction at any point in your image processing tree using a dedicated node. If you
have an image with noise that you think might be enhanced by whatever corrections you need to make,
for example increasing the contrast of underexposed clips often increases whatever noise is within an
image, there are two approaches to noise reduction:

Apply noise reduction at the beginning of a node tree:

This lets you preemptively eliminate any noise before it becomes a problem as
a result of whatever adjustments you’re planning on making. The end result can
be smoother, but you may also notice that the edge detail within the image is
a bit softer as a result.

Apply noise reduction at the end of a node tree:

The alternative is to make your adjustments first, and then apply noise reduction
in a separate node afterwards. In this case, you may find that the noise reduced
regions of the image aren’t quite as smooth, however the edge detail within the
image may be visibly sharper as a result.

Advertising