The Viewer is a suprisingly complex aspect of Nuke's interface. Using it, many channels and layer sets can be previewed, on-screen rendering processes can be moderated, the results given onscreen stress-testing, low-rez proxy mode enabled and the time-line skimmed. In short, it is the 'hub' in which the results of the script are presented and overseen.
- 1 The Viewer interface
- 2 Render Control
- 3 The Viewer node
- 4 Viewer Shortcuts
- 5 Links
The Viewer interface
This is where channels are managed. In day to day application, this is mostly used for viewing the alpha channel in the viewer window using the RGB pull-down menu (see #Viewer Shortcuts and Script Evaluation#Channels in the Viewer window). render passes from a 3D app can also viewed here.
This is a handy set of controls for enabling one part of the script to be previewed against another within the same viewer. To use, sleet the first node that you wish to preview, and press '1'. Then select the other point you wish to preview from and press '2'. If you select 'over' from the Centre value, then both feeds will be visible in the drop down lists on the light and left. Thereafter, they may be compared to each other using the on-screen widget. Comparisons may be done 'wipe-form' or faded into each other.
The 3D menu set. This only becomes relavent when you are in 3D mode. 3D mode is enabled by pressing 'Tab' when you are in the viewer window. Suprised by what you see there? Scared by the unfamiliar enviroment? Then get the hell out by pressing 'Tab' again.
D: f-stop and gamma sliders
These sliders change the gamma value and the exposure settings of the preview image. The main use of these sliders is to 'visually diagnose' the image, e.g. To test if an alpha channel has any holes in it. Their use is discussed here. If they are set to anything other than their default values, they they will turn red. This is to remind you to set them back after use.
E: render management
A number of interesting things here all useful to manage the render so that it is quicker. Details in the following section.
F: value information bar
Here numerical color values from the preview are displayed that correspond to the region directly under the cursor, or from an average derived from a marquee-selcted area (F1). See: here). More advanced forms of the same thing are a aisle in the Pixel Analyzer.
A simple time bar. Final cut pro it ain't, but is easy enough to learn (so easy I am not going to say much about it). Every time a key point is set on an active node a red marker appears in the timeline. Setting a 'frame range' is done by setting the parameters on the far right of the bar (see keystrokes below for a more convenient way). This locks the playback to the prescribed range. Press the 'frame range lock' to reset.
The following controls are available in the render control set (region 'E').
H: proxy toggle
Processing and rendering a Nuke script can be very processor intensive. One way of addressing this issue is by using a proxy. A proxy is A smaller version of a source file. Any edits upon the file are scaled down to the same degree. When it comes time to review the final version, the source file and its edits are scaled back up to their original size. Nuke recognizes two flavors of proxy. 'Pre-rendered' proxies call upon proxies of the source files that are rendered in advance. Pre-rendered proxies are simple to render: simply point the proxy file path of a Write node at a conveniant location, set the render to proxy and hey presto. These can be read in from the same location. More common are 'on-the-fly' proxies. These are rendered by Nuke as and when it needs them. The percentage by which a proxy is scaled down can be defined in the '[Project Settings]]. The default is 50%.
Reduces the resolution of the preview in the viewer. Speeds up the render within the viewer but, unlike the proxy value, does not affect the speed at which the script processes.
J: region of interest
If this is activated, an editable rectangle widget appears within the viewer render window. This restricts the limit of the rendering of the script within the viewer window, hence speeding up the viewing process.
K and L
Stops and starts the render. This is good for when you need to update the parameters of a node without the viewer updating as well (and slowing down the preview in the process).
The Viewer node
As well as being a window, the Viewer is also a node. The Viewer window previews the script from any point in the node tree. The Viewer node is the viewer's presence in the Node Graph. It will preview from any point in the node tree that it is connected to. The shortcut voodoo for this is to select the Node from which a preview is required and to press the '1' key. If another point in the node graph is selected, and the number '2' is pressed, then that that point in the tree is also previewed. Further previews are available to sequential numbers.
New viewers windows can be added by adding new viewer nodes: command 'i'.
|R, G, B, A or Y||Isolate the Red, Green Blue, Alpha or Luminance channel. Press again to return to composite view.|
|H||Home in viewer. Different to F, this fills the Viewer window according to the format,|
|Q||Hides any on screen controls (e.g. those of the Transform and Roto nodes). Press again to reveal.|
|Tab||Move into 3D view. Press again to return to 2D.|
|+ and -||Zoom in and out (also scroll wheel).|
|F||Fits and centres selected contents to pane (in the Viewer fits format to window, in the Node graph fits selected nodes to node graph etc).|
|Spacebar||When panes are docked (i.e. not floating) this fills the workspace with the current pane).|
|option click + mouse drag||Move the work space around within the pane.|