Professional noise reduction is possible with the denoise effect.

Core workflow

After the effect is applied to a noisy layer a highlight box will be displayed in the Viewer. This can be adjusted using the two control points.

This box should be positioned over a noisy area of the frame. This will provide the effect with input to analyze the video. Ideally select a flat region with mid-level brightness.

Click the Analyze button to analyze the selected area.

Some noise will be removed immediately.

To further refine, change the View setting to Analysis Box.

Move the control points so that the analysis box is over darker noisy area.

Click the Analyze Brightness button in the Analyze control group.

This will analyze the difference in the noise levels, between the brightness of the original analysis and the brightness of the new analysis region.

This will refine the noise removal.

The process can be further improved by increasing the number of frames used during the analysis. The more frames it uses, the more time it will take to render, but usually with superior results.

Controls

  • Analyze button – Only visible when image has not yet been analysed. Uses the selected analysis area, to analyse the noise of the image.
  • Frames – The number of frames used to remove the noise. It should be kept at ‘1’ if applied to an image.

View modes

  • Result – shows the result of the noise reduction
  • Analysis Box – shows the region that will be used for analysis if one of the analysis button is pressed
  • Frequency Y – shows how much the brightness changes at the selected frequency level
  • Frequency U – shows how much the u color changes at the selected frequency level
  • Frequency V – shows how much the v color changes at the selected frequency level

Frequency View Options Group

Only visible if the selected View is one of the frequency options.

  • Level – The frequency level to be viewed.
  • Contrast – Allows you to change the contrast of the displayed frequency. Sometimes the changes in frequency are too subtle to see. This control allows you to increase the contrast, to help you to see it.

Analyse Brightness

Used when noise amount varies depending on the brightness.

Updates the threshold values in the ‘Threshold – Brightness’ control group.

Re-Analyse

Discards the original analysis that was performed and re-analyzes with the currently selected analysis area.

Threshold

The controls under this group help the plugin to identify what is and isn’t noise, by manually adjusting the threshold values.

  • Brightness – The controls under this group, allow you to change the threshold values dependent on the brightness of the image. If, for example, not enough noise is being removed from dark areas, just increase the threshold of the ‘0’ slider. If the problem is in the white areas, adjust the ‘100’ slider. The sliders act like a graph, adjusting one will cause the nearby sliders to also be adjusted.
  • Channel – These controls allow you to change the noise threshold depending on the YUV channel.
  • Frequency – This allows you to change the noise threshold depending on the frequency of the data. It is easier to edit these slider if viewing the correct frequency: simply change the view to one of the frequencies; then change the frequency options level to correspond with the frequency slider.
  • Keep Edges – This slider sometimes helps to bring more of the edges back. But as the value is increased, more of the noise might be brought back as well.

Removal Amount

These controls allow you to adjust how much of the noise is removed.

  • Y controls how much noise is removed from variations in brightness, of the image. By default it is set at 80%, as removing all of the noise from the brightness tends to not look as good.
  • U/V controls how much noise is removed from variations in colour.
  • Frequency controls how much noise is removed dependent on the frequency of the data.

Was this helpful?

Yes No
You indicated this topic was not helpful to you ...
Could you please leave a comment telling us why? Thank you!
Thanks for your feedback.