There are two things you need to master when working with video in Cubase: first, the mechanics of actually getting video to play in time with your Cubase Project and working with timecode, and second, building tempo maps that allow you to accurately place musical moments at specific time locations in the video. In this month's Cubase workshop, we're going to look at the first of these areas — but stay tuned, as we'll be turning our attention to learning how to build tempo maps specifically for writing to picture in a future issue.
There are two ways you can run video alongside a Cubase Project. The first is to use Cubase's built-in video player and make your video file (or files) part of your Cubase Project. The second is to synchronise an external video device via MIDI Time Code (MTC) so that when Cubase's transport is running, the external device chases and runs in sync. There are many advantages to running video inside Cubase: not requiring any extra equipment is an obvious one, but another is that the synchronisation between Cubase's transport and the video is instant, so the picture starts playing immediately with the Project, and the appropriate frame at the current position of the Project Cursor is always shown.
Despite these advantages, though, running picture in Cubase isn't always the best solution. As anyone involved in film or TV will know, it's rare to only ever receive one cut of video. A far more common situation is that you receive multiple cuts, on a weekly (or sometimes daily) basis, that have to be re-integrated into your Project. Let's say you're working on a 30-minute TV show that comprises 20 different music cues, and each of these music cues is a separate Cubase Project, each containing video. Every time you get a new cut of video, you'll have to update the video in 20 different Cubase Projects: if you're working on a feature film that might have 50 to 100 cues across multiple reels of video, this situation becomes really frustrating, really quickly.
For this reason, a large majority of working media composers tend to run video outside of their sequencers, and even if your main sequencer is Cubase it's pretty common to see a separate system, locked via MTC, being used to run video. The big advantage of an external video machine is therefore that you can reload the video once and every Cubase Project will lock to whatever the external video machine is playing. The disadvantage is that it usually takes a second for the video machine to lock in with Cubase, and getting the video to display the exact frame at the current position of the Project Cursor isn't always as easy as you might think.
So if you're working with one or two video files that won't be updated on a regular basis, the best approach is probably to run video within Cubase. However, if you're working with multiple video files that will have frequent updates, the extra expense of a separate computer to run video is probably justified by the time you'll save. While this article covers using video in Cubase, in the near future we'll be looking at tips for running a separate video system alongside Cubase.
Understanding Cubase's Video Player Device
Cubase's built-in Video Player Device is fairly straightforward, but there are some differences to be aware of, depending on what type of movie file you're working with, what codec the video is compressed with, and whether you're running a Mac or Windows computer with or without external video peripherals. To configure the built-in Video Player Device, select Devices / Device Setup to open the Device Setup window, then choose Video Player within the Video folder, to see the available options in the panel to the right of the window. At the top is a pop-up menu labelled Playback Method, where you select which video engine is used to play back the video. By default, Mac users only have one option, 'Quicktime Video', while Windows users have this option along with 'Direct Show Video' and 'Video for Windows'. The Video Properties group offers various parameters for configuring the chosen Playback Method.
If you do end up using an additional video output device in conjunction with Cubase's Video Player Device, there's a useful Frame Offset parameter in the Quicktime Video Playback Method's Video Properties group that lets you specify by how many frames to play the video ahead of the actual Project time during playback. For example, setting this value to three means that the video will always play three frames ahead of the Project time during playback, which allows you to compensate for any latency introduced by the video hardware. So if you were using the ADVC110 you'd want Frame Offset to be set between five and six frames to allow for the digital-to-analogue video conversion.
The 'Direct Show Video' Playback Method for Windows supports the same codecs and file formats as the Quicktime player, with the exception of support for Quicktime files themselves, but with the addition of support for Windows Media Video (WMV) files. So, for example, if you need to play back Quicktime files, you'll need to have the 'Quicktime Video' Playback Method selected; but if you have a WMV file, you'll need the 'Direct Show Video' Playback Method instead. The 'Video for Windows' Playback Method only supports AVI files and the Cinepak and Indeo codecs, although Motion JPEG can also be supported, depending on what video hardware you're using.
The only Video Properties available for the 'Direct Show Video' and 'Video for Windows' Playback Methods are the choice of three possible sizes for the Video Player window. While the 'Quicktime Video' Playback Method doesn't offer these choices, it is instead possible to resize the Player window to any size you like by dragging the bottom-right corner as you would any other window.
When you're running video in Cubase, you can be grateful that it has possibly the best internal video handling of any sequencer or digital audio workstation available, thanks to functionality that was originally introduced in Nuendo 1.0 way back in 2000. Basically, an audio file is referenced in a Cubase Project as an Audio Clip, and when you use an Audio Clip in the Project window you create an Audio Event, which can be dragged around, resized, and so on. A video file is therefore referenced as a Video Clip, and a Video Clip appears in a special Video Track in Cubase as a Video Event, which can be dragged around, resized and handled just as if it was an Audio Event.
To add a video file to a Cubase Project, select Pool / Open Pool Window, click the Import button at the top-centre of the Pool window (or select Pool / Import Medium) and choose the video file to be added. A Video Clip will be automatically added to the Pool to represent the video file. If your video file contains audio content, such as dialogue or temp music, you'll need to extract this data as a separate audio file, as Cubase has no way of playing audio contained within a movie file. To do this, select the Video Clip in the Pool and choose Pool / Extract Audio from Video File. A new Audio Clip is added to the Audio Folder in the Pool.
At this point it's worth noticing the Info column in the Pool, which for a Video Clip tells you the frame rate of the video (in addition to its length and resolution), and for an Audio Clip shows what sample rate, bit depth and channel format are used, along with the length of the file. It's important that the frame rate of the Cubase Project matches that of the video, so that the timeline of the Project exactly matches the timeline of the video. To set the Project's frame rate, open the Project Setup window by selecting Project / Project Setup (or pressing Shift+S), choose the appropriate frame rate from the Frame Rate pop-up menu, and click OK.
The next step is to create a Video track in the Project window (Project / Add Track / Video) and an Audio track (Project / Add Track / Audio) of the appropriate format (such as stereo or mono, as indicated by the information in the Pool). Now, drag the Video Clip from the Pool onto the Video track so that it's positioned at the start of the Project; and do exactly the same with the Audio Clip, so that it lines up precisely underneath the Video Event. In order to keep the video and its audio together, it's a good idea to group these two Events, by selecting them both and choosing Edit / Group (or pressing Ctrl/Command+G). Now, any edit operations you apply to either Event will be applied to the other Event. You can later ungroup these, if you want to, by selecting both Events and choosing Edit / Ungroup (or pressing Ctrl/Command+U).
Once you have Video Events in a Video track, you can open Cubase's video window where the playback of the video will be displayed by selecting Devices / Video or pressing F8. See the 'Understanding Cubase's Video Player Device' box (below) for more about the various playback options.
At this point, it's important to make sure that the video is placed (or 'spotted') to the correct timecode location in your Cubase project. Hopefully, your video will contain a timecode burn-in (literally, timecode that's visible, or 'burnt', into the image you see) from the editorial department that gave you the video; if it doesn't, you may need to ask them what the first frame of picture should be. If there is indeed a burn-in on your video, you need to make sure that the timecode in Cubase precisely matches the timecode on the video.
If the timecode starts running at the start of the video, all you need to do is take a note of what the SMPTE time is on the video. This is easy if the Video Event is at the start of the Project, as you can just set the Project Cursor to zero and look at the Video Player window. Next, open the Project Setup window again and set the Start value to the timecode value at the start of your video, before clicking OK. Now, wherever you place the Project Cursor, the timecode visible on the video should be exactly the same as the timecode readout in Cubase. If you can't see a timecode readout in Cubase, set the Secondary Display Format to Timecode by clicking the Select Secondary Time Format button (which is located just above the Record button on the Transport Panel) and selecting Timecode from the pop-up menu. Usually, it's best to leave the Primary Display Format set to Bars+Beats, since that's the time format you'll need to work in once you start writing some music.
The only slight problem you might encounter is if the timecode on the video doesn't start running at the very start of the video — in this case you'll need to crop the Video Event so that the start of the Event is the first frame of video you can reference in terms of timecode. This is pretty easy, as you can change the start point of the Video Event just as you would that of an Audio Event, by dragging the Event's start handle, or numerically by selecting the Event and changing the Start time in the Event Infoline. If you're going to adjust the start of the Event in the latter way, you might want to swap the Primary and Secondary Display Formats temporarily, so that you can adjust the start of the video in terms of frames rather than bars and beats. Do this by clicking the Exchange Time Formats button (which is an icon just above the Stop button on the Transport Panel and looks like a thin line with two arrows).
By default you'll notice that the Video Event displays a series of frames from the video, to give an approximate idea of what's going on in the picture at any given position along the timeline. The option to see (or not see) these so-called 'thumbnails' is located in the Event Display-Video page of the Preferences window, and you'll notice that this page also contains an option for setting the size of the Video Cache. The Video Cache is used to store the thumbnails generated from the video file in memory, so that Cubase doesn't have to read from the disk every time the thumbnails need to be redrawn. If you find yourself zooming in and out of the Project window horizontally quite frequently, this causes the thumbnails to be constantly redrawn, so it's worth increasing the cache size if you notice either your computer slowing down or the process of redrawing the thumbnails on the Video Event being sluggish.
Cubase SX/SL 3.1 adds a further new trick for improving performance when working with thumbnails, by offering the option to create a thumbnail cache file on disk as well. The way this works is that if Cubase needs to redraw a thumbnail that doesn't exist in the memory cache and would take too much processing to generate from the video file, it uses a lower-resolution thumbnail stored in the cache file instead. Once the processor is free to redraw the thumbnail from the video file, Cubase will do this, and the thumbnails will be redrawn at a higher resolution. To create a thumbnail cache file, select the Video Clip in the Pool and choose Pool / Generate Thumbnail Cache.
At this point I have a small confession to make: there is actually a quicker way of performing the tasks I've just described. Selecting 'File / Import / Video' displays a file selector with two built-in options that are enabled by default: 'Extract Audio' and 'Generate Thumbnail Cache'. If these options are enabled, Cubase will automatically extract the audio, generate the thumbnail cache, add the appropriate Audio and Video tracks and the Audio and Video Events.
The reason for explaining the whole of the process is more for those situations where you might need to do something out of order, or work with multiple video files in the same Project. This latter point is actually really important, because Cubase SX is one of the only sequencers on the market that can deal with multiple video files in the same Project, since the Video track, like any Audio track, can contain multiple Events from different files. This is incredibly handy, because TV shows and films are often broken down into acts and reels, and Cubase allows you to load these in a single Project if you wish, without having to use a separate application to merge them together.