In Part Two we looked at the interface and import workflow of The Foundry Nuke and NukeX. In this final part we’ll look at the playback, rendering and export options in Nuke.
Nuke’s rendering system
Nuke renders mainly using the CPU, and multiple cores and/or threads are supported. Whenever you make a change to the node graph, Nuke computes the image once again (except for the precomped regions) and displays it line by line.
This is called Scanline rendering. The cool thing about this is you don’t have to wait until the entire frame is rendered before knowing if your effect has worked or not. If you don’t like what you see, you can cancel the render and continue working.
You can go even further. You can zoom into a region of the viewer to render only that portion. You can choose this region accurately using the Region of Interest (ROI) button above the Viewer.
Once you are satisfied with a still render of a frame, you will want to see how your composite works in motion. To do this you must first render the frames in the entire shot and then play them back for inspection.
This process is called flipbooking. There are three ways to flipbook in Nuke:
- You can use the automatic disk caching feature which will let you play back your shot in the Viewer itself. However, The Foundry advises that this will not allow you to set the frame rate correctly which means you can’t judge motion characteristics perfectly.
- You can use FrameCycler (from Iridas, installed along with your copy of Nuke). This has the advantage of playing back the rendered shot in the required frame rate.
- You can set up an external flipbooking application via Python.
Out of these three, the most accurate playback you can get out of the box is via FrameCycler. Real-time playback is only practical if you have a really fast hard drive or RAID array. If you’re working in 32-bit or 16-bit mode, the data rate is easily quadrupled or doubled from 8-bit mode.
To access FrameCycler, go to Render > Flipbook Selected (you must select a node first):
You can choose to render the original full size frame or a proxy, with a unique LUT. You can choose to display the final image (rgba channels), or separate channels individually. You an also export Audio if it is available via the AudioRead node.
Click OK to render.
Regarding ‘Render in background’ Nuke advises:
If you check this, you can also set #CPU limit and Memory limit controls. The former limits the number of threads that Nuke uses in the background and the latter limits the amount of cache memory that Nuke uses.
If you’re rendering multiple sequences in the background, this can take up more than the total RAM on your machine. When running background renders of any type, you need to make sure they don’t require more RAM all together than what’s available on the machine, otherwise you may experience problems such as hanging.
The output of this operation is defined according to your project settings and the resolution and frame rate you have applied there.
You can also preview your sequences on a broadcast monitor. It requires additional hardware:
Currently, development and testing is done using BlackMagic Decklink HD Extreme 2 and AJA Kona/Xena LHe.
are limited to 10-bit. To read more about how to playback on an external monitor, read page 506 of the User Guide.
Supported codecs for export in Nuke
Here’s an incomplete list of supported codecs and file formats for export:
- AVI (Codec must be installed in your computer in Windows)
- Quicktime (Codec must be installed in your computer, both Windows and OS X, bit depth or gamma metadata does not show up, so the image might look ‘off’)
- YUV (resolution independent, Nuke assumes 720p) 8-bit
Still images and image sequences
- Cineon 10-bit log
- DPX up to 16-bit
- OpenEXR 1.0 and 2.0 16 and 32-bit
- TIFF up to 32-bit
- Targa up to 8-bit
- PNG up to 16-bit
Export workflow in Nuke
Rendering in Nuke is also ‘connected’ to the node graph. Make sure your project settings match your delivery or output format. Pay specific attention to the color space or any LUTs you might have used, and also the output gamma of the video (Nuke works internally in linear space).
Create a Write node (Hit W) and link it to the node graph. In the Properties pane, you’ll get the following options:
If you type in the file extension in the file name, Nuke will automatically assign the file type. Choose the colorspace and ensure the write views is selected.
The write node can also be connected to the viewer node so you can recheck you’re rendering the right node tree.
Most of the time you’ll be selecting the ‘input’ Frame range, which is the frame range of your shot. You can also choose ‘global’ (defined in project settings), ‘custom’ or ‘viewer#’.
If you’ve allocated resources for background rendering in the Preferences, you can check Render in background. Click OK.
If you’ve chosen Quicktime or AVI, you’ll get a list of installed codecs to choose from. (Note: Even on a Windows machine, Nuke didn’t display the AVI export option, though it was available for import. Strange.) Click on Advanced to tweak your codec. You can also select an audio file to be multiplexed with the video.
For instructions on rendering to a render farm, read page 517 of the User Guide.
A few more things to know
Nuke is not a complicated program as far as workflows are concerned. However, it offers so many tools and features that it will take years to become proficient at it. You can specialize in core activities within Nuke, like keying and color, 3D compositing, rotoscoping, tracking, etc. The sky’s the limit!
Before I close this crash course, let’s look at three important features of The Foundry Nuke that you need to understand.
The Dope Sheet
The Dope Sheet lists all the nodes in your node graph that can have keyframes assigned to them:
This gives you an After Effects-like feel that lets you trim clips and manipulate keyframes easily. With effects nodes you might be able to add keyframes to individual parameters as well. You can also add an audio track with the AudioRead node.
To know more about using the dope sheet, head over to page 86 of the Getting Started Guide.
The Curve Editor
The Curve Editor also shows keyframes according to time, but as a value (y-axis) so you can manipulate the value intuitively and quickly. Instead of the straight lines you see in the Dope Sheet, you get the same thing but it’s wavy here.
In addition to providing an easy access to the values of the parameters, you can set curves to follow mathematical properties with expressions (formulas, just like in After Effects), and also perform interpolations to get smooth animations. To read more about the curve editor, head over to page 90 of the Getting Started Guide.
Sometimes, if you’re using both the curve editor and the dope sheet, you might want to get them to match time-wise (actually frame-wise*). You can sync the two by right-clicking on the dope sheet and choose View > Synchronize frame range.
*By now you might have realized that everything in Nuke is frame-based, which is why Nuke works best with image sequences. But that doesn’t mean you can’t have frame rates like 23.976 or 59.94, etc. Nuke supports frame rates from 0 to as many numbers as you can fill in the box, but to play them back you’ll need to conform to something under 60p for most display technologies.
The Terminal is not really a part of Nuke, but you can use it to get a lot of stuff done quickly. The Terminal is just the:
- The Linux Terminal
- Windows Command Prompt
- OS X Terminal
The most common usage for the terminal is to launch Nuke. Now you might want to know: Why not just click the icon to launch Nuke? One good reason is if you want to launch Nuke with customized options on the fly, you can use the command prompt instead. E.g., if you type nuke -p, Nuke launches in proxy mode – even if your script was in full size mode. There are tons of things you can do with the terminal. See the potential? To take a detailed look, head over to page 559 of the User Guide.
There you have it. We’ve covered a lot of ground in this three-part crash course, and I hope I have given you enough information so you can decide if Nuke is the right tool for you or not. Your next stop, if you choose to use The Foundry Nuke, is to read (or at least skim) the Getting Started Guide and User Guide from end to end. You might have to bookmark the sections you’re interested in, because it really does cover a lot of ground.
Ultimately, don’t forget that you can’t fool people into believing you’re good at Nuke. It’s not what you know, but what you produce with it that counts.