Realtime Sweepable Mid-EQ - SGX-2 Piano Sound Engine | Part 3

Media Composers

By John Walden

Whether you're adding some music to your best mate's wedding video or you've just received a call from Spielberg's people about his next mega‑film project, writing music for picture can be a tremendously satisfying creative process. Cubase has an excellent suite of tools for this sort of work so, for those about to score, we will guide you….

In The Box

The Import Video dialogue box provides information on the format of the video file selected and a self‑explanatory option to 'Extract Audio From Video'.

Mark Wherry explained, back in SOS November and December 2005, how composers can work with video played back on an external device (locked to Cubase via MIDI timecode) or using a digital video file in the Cubase Video pane. Five years is a long time in music technology, and with film and music collaborators increasingly delivering video and audio files via the Internet, the 'internal' approach is becoming much more common. It also has the practical advantage that we media composers require no additional kit!

Whether on Mac or PC, Cubase supports a range of video formats. Nonetheless, access to some dedicated video format‑conversion software can save a lot of hassle. There are plenty of options for such software, but as v5.5 of Cubase adopted the QuickTime engine for video playback, the full version of QuickTime Pro (at about £20) seems, to me, to be a sensible starting point.

Going In

If your video file plays back successfully in the QuickTime Player, importing it into Cubase should be straightforward. The most direct route is via the File/Import/Video File menu option, which imports video into the Pool and places it on a Video track in the Project window. If you tick 'Extract Audio From Video' in the 'Import Video' dialogue box (see screen above), you will also get the audio from the video file — perhaps dialogue and sound effects — placed onto a separate audio track.

Cubase will automatically set the Project frame rate to that of the imported video file (this can be checked and set manually via the Project Setup dialogue box). From v5.5 onwards, an error message is displayed if the frame rate is unsupported (which is one reason why format-conversion software is useful). The Device/Video Player menu option opens the Video window. This is resizable, so you can customise both size and position to best suit your screen real‑estate. The example shown in the screenshot left includes 'burnt‑in' SMPTE time code, with playback here at 50 seconds and six frames. If your film‑making collaborator can include this in the digital video file, life becomes much easier for you, as it provides an exact frame of reference for placing music cues and sound effects.

Once the film is imported, you should also use the Set Timecode At Cursor dialogue (from the Project menu) to match the Project timecode to that of the burnt‑in timecode. If you do this once anywhere within the project, the two should be in sync throughout, and you can then lock the position of the video track and its associated audio track, so that they don't get moved accidentally.

One Project, Many Projects?

The thumbnail view stills give you a visual cue of scenes in the overall project. (The Preferences dialogue box allows you to allocate more space to the thumbnail cache, and can be helpful if screen redraws become sluggish.) Note also that both the video and its associated audio track are locked in position to avoid accidentally moving them. In addition, a Ruler track is displaying timecode and a Marker track has been used to identify key positions along the timeline, such as scene changes or visual hit points.

If the video project is longer than 15 minutes or so, and it is likely to feature multiple musical cues, there is a practical question to consider before starting work: do you create a separate Cubase project for each cue, or do you have a single project and create all the cues within it? While the latter might seem the obvious choice, there are pros and cons to both approaches. That said, given the gradual improvements that Steinberg have made in how Cubase can handle video playback (for example, the ability to use multiple video clips on a Video track), my instinct now is always to start with a single project containing all cues. I would, however, qualify that by encouraging a scrupulous attitude to backing up project versions as the work develops.

Hit Points & Hitpoints

The Marker track can be used to identify all the key visual points within the film. In composer speak, these are often referred to as 'hit points' — that is, key visual events that the music has to 'hit' in some fashion — scene changes or particularly dramatic moments. These are not, of course, the same as Steinberg's use of the term 'hitpoints', which refers to transients within an audio file that can be used to create audio slices in, for example, drum loops.

Given that these hit points (composer speak) represent fixed time points in the film (well, fixed until you are supplied with yet another edited cut of the film where scenes have been made shorter, longer or re‑ordered just to make the composer's life more fun!), the Marker track should be changed to Linear mode (see the screenshot — a small clock icon is displayed within the Track List). This locks the markers to their SMPTE timecode position rather than the bar/beat position, so if you make lots of tempo adjustments, as described below, the markers stay locked at the correct timecode location.

Slow, Slow, Fast, Slow

The Video Player window showing a video containing burnt‑in timecode.

Tempo plays a very significant role when producing music to picture, not just in setting a tempo suitable for the musical mood, but also because subtle shifts in tempo are often used to get a musical 'hit' to synchronise with a visual hit point. For example, in the main screen shot, scene three starts just before bar 57, but for musical purposes, I might require the scene start to coincide exactly with a bar start. A small tempo increase prior to bar 57 could bring the two events into sync. Therefore, before doing any significant musical creation, composers will generally spend some time creating a tempo map for the project.

Using a combination of the Tempo track (Project/Add Track/Tempo Track) in the Project window and the Time Warp tool allows you to add tempo-curve points to develop tempo changes suitable for each of your cues. For example, with the Time Warp tool selected, I can easily grab the bar 57 gridline and drag it so that it's aligned with the marker placed on the first frame of scene three. This process works by automatically adjusting the tempo value at the previous tempo-curve point.

This is fine — unless, of course, that previous tempo was essential for the timing of an earlier cue! In practice, what this means is that you need to add additional tempo points just after the end of each musical cue, so that any tempos set within the cue itself are not altered when you use the Time Warp tool further along the timeline. It's also not a bad idea to add an additional tempo-curve point just before the start of a cue. When in Time Warp mode, these extra tempo-curve points can be added by simply clicking at the required grid point in the Project window.

Creating tempo maps in this way is simple in principle, but it usually seems rather more difficult in practice. It's quite easy to get yourself in a tempo tangle and add all sorts of unmusical tempo changes in an effort to 'make the cue fit' the musical bar/beat grid. Practice makes perfect, but it's worth spending time on this before you start any audio recording. MIDI data can be made to follow subsequent changes to the tempo map, but you don't want to splash out on a real string section, only to find out later that you need to adjust the tempo of a cue to make it 'fit' the visuals better. Yes, I know that it's possible to manipulate recorded audio with time‑stretching or audio warping, but it's much easier — and, more importantly, you'll get much better results — if you can avoid this.

Safe Delivery

The Export Audio Mixdown menu allows stems of particular instrument groups to be created. In this orchestra‑based project, the Channel Batch Export has been selected and a stereo mix will be generated for each of the four Group channels selected. Providing all the individual instruments are routed to the appropriate Group channel, this will create an audio stem for each of the main orchestral sections.

Of course, with all the above technical details resolved, writing the perfect score is a breeze. OK, even if that's not really true, let's assume that you've written a score you're happy with; the final technical issue is getting the music out of Cubase and back to the client (or your mate who just got back from his honeymoon...). You might share previews with them by embedding the audio into a QuickTime movie and sending that, but for the final delivery of the sound files, you'll need to fully understand what format they require for the audio.

Ignoring the whole issue of surround mixes, this might be as simple as a single stereo mix starting from frame one and, if so, export your final mix as you would any song‑based musical project. However, make sure you check the required sample rate: 48kHz (not 44.1kHz, as on music CDs) is the standard in the world of film, as it ensures an integer number of samples per frame.

In more professional contexts, you might also be asked to provide stems (separate stereo mixes for each instrument group) so that the sound mixer and/or director can adjust the music to their own taste and balance it better with the dialogue and sound effects mixes. Fortunately, the Export Audio Mixdown dialogue from the File/Export menu (see screen above) provides all the options you might need, allowing you to generate a series of stereo mixes, each based on particular sets of tracks.    

Published May 2011

What is Damper Resonance, Mechanical Key Sound & Damper Noise? NAUTILUS | Part 2

Keyboards & Other Controllers

By John Walden

The low-velocity MIDI notes generated in this MIDI guitar performance (top) were removed using the Input Transformer. In the 'after' example (above), these changes have been made permanent simply to illustrate which notes the Input Transformer has filtered out.

Not everyone who uses Cubase is a competent keyboard player. Guitar, drum and wind MIDI controllers offer alternative means of getting MIDI information into a project — but all represent something of a compromise. For example, pickup‑based MIDI guitar systems, such as the Roland GK3, rely on a conversion process to translate string vibrations into MIDI pitch and velocity data. Brilliant though they can be, the conversion process sometimes gets confused by all the fret squeaks, damped notes and other incidental noises in a guitar performance. When using such controllers, therefore, you'll probably need to do a little more MIDI editing to clean up the performance than you would with a well‑played keyboard part. Thankfully, Cubase offers a collection of tools that make this 'tidy up' process less of a chore.

On The Way In

One way to reduce the requirement for post‑performance editing is to clean up the MIDI data as it is being recorded. The Input Transformer (which can be switched on via the Inspector Panel) can help here. This provides a subset of the MIDI-editing functions available in the Logical Editor. However, before looking at the Input Transformer itself, make sure you insert an instance of the MIDI Monitor plug‑in, which displays the MIDI data for a track in real‑time, and thus allows you to see 'live' what the Input Transformer settings are doing to the MIDI data input as well as events occurring on playback.

The Input Transformer window has two panels, the upper of which is used to specify what MIDI events are to be 'caught' by the Input Transformer. Events that don't meet these conditions are allowed to pass. Those that are caught can be either filtered (removed) or transformed (altered in some way). This choice is set in the Function box, which is found bottom‑centre of the Input Transformer window. If 'Transform' is selected, then the lower panel is used to specify how the events that meet the condition in the upper panel are to be changed. Up to four such configurations can be set, one in each of the Input Transformer's four modules.

Here, the Input Transformer is being used to adjust the velocity of just the snare (note D1) from a MIDI drum controller performance.Let's look at some useful applications of the Input Transformer. First, if your MIDI controller generates a supply of unwanted extra notes, you can attempt to filter these out. For example, MIDI guitar devices often produce low‑velocity notes as a result of fingering noise and other playing artifacts. These can be automatically removed as you play, by setting one of the four Input Transformer modules to filter out all notes below a certain velocity value. To achieve the results reflected in the screenshots above, all notes with a velocity equal to or less than 25 were filtered out — you can, obviously, adjust the velocity threshold to your own needs and from performance to performance.

A second application is to filter out unwanted pitch-bend data if, for example, you're playing a virtual piano that shouldn't respond to pitch-bend. Doing so with the Input Transformer is often easier than diving into the menu system of your hardware controller to switch pitch-bend data off.

As a final Input Transformer 'quick fix' example, I find that my snare-drum trigger (purchased separately from the rest of my electronic drum kit) consistently generates rather low velocity values. To avoid having to hit the drum harder, or delve into the drum-brain's settings, the response can be adjusted by increasing the velocity of all the snare-drum notes by a fixed amount. The Input Transformer setup I used to achieve this is shown in the screenshot above.

After The Fact

The original drum performance (before) and the processed performance (after). This example is shown for illustration; in reality, you might not wish to clean things up quite so clinically.

There are some things that the Input Transformer cannot do, such as ignoring the very short notes which are sometimes generated by MIDI guitar systems, reducing long notes to a fixed, shorter length (my own drum controller sometimes generates such long notes), or removing 'doubles' where two notes are generated by the controller when only one note was intended. What these conditions have in common is that none of them can be recognised until after a Note Off is received.

Thankfully, the Transformer MIDI plug‑in — which looks very similar to the Input Transformer — acts after a recording has been made and can offer you more control over the result. As it operates in real time during MIDI playback, there's no risk of making unwanted permanent changes to your MIDI performance. The Transformer plug‑in includes presets for ignoring notes below a specific velocity, and for filtering out controller data such as aftertouch (this preset can easily be edited to target pitch-bend). If you don't wish to permanently eliminate such data while recording, in case some of it proves useful later on, the Transformer is a better choice than the Input Transformer for filtering.

One further useful Transformer preset is 'delete short notes'. This defaults to a setting in pulses per quarter note, but you can change the time-base to seconds, samples or frames if you prefer, and then adjust the maximum length of note that you wish to filter out. As mentioned earlier, this can often help clean up short notes that have been falsely generated by a guitar or drum MIDI controller.

Permanent Commitment

The Quantize Setup dialogue.

Some simple clean‑up tasks require permanent changes to the MIDI data (though you can use the Undo command if things go wrong!), and these can all be achieved via the MIDI menu, using a combination of the Quantize Setup, Functions and Logical Presets options.

Let's consider an example. My MIDI drum controller has a habit of generating multiple short notes when I intended to hit a drum only once, as well as the occasional spurious long note. The first issue is more significant, as on playback it produces several sample triggers close together in time. The second simply looks a little odd in the standard Note Editor.

To get from the raw MIDI data to a suitably cleaned version there are a number of steps, but as this is going to be a destructive process it's probably best if you first make a copy of the original performance. Applying the settings shown in the Quantize Setup screenshot (below) pulls all the separate notes to the same start position. You'll need to set the grid size to suit the basic divisions of the pattern (in this instance it was eighth notes, for the hi-hat part), and the 30 percent Magnetic Area catches the most likely 'double notes'. Incidentally, once you've configured the Quantize Setup dialogue, the first step described above could also be achieved by turning on the auto-quantise option while recording.

Setting the Iterative Strength to 100 percent forces all of these notes to exactly the same start position. This is essential for the next step, where the MIDI/Functions/Delete Doubles option deletes any notes with the same pitch and start position — but it only works if the notes have exactly the same start position. Having got rid of the double notes, the varying note lengths can be dealt with. Although this is mostly cosmetic, when it comes to triggering drum samples it makes things look a lot neater, and I find that this in turn makes things much easier to edit.

Returning to the Quantize Setup dialogue, set the grid to something smaller than the smallest division in the performance. For example, for our eighth‑note pattern, a grid of 1/16 or 1/32 could be used. If you execute the MIDI/Functions/Fixed Length command, all notes will be set to the length of the chosen grid division. This process is easiest to follow in the Key Editor. It won't make any difference in the Drum Editor where note lengths are not displayed.

There is an obvious down side to doing all of this quantising, though: although you may retain the dynamics of the playing, it can easily make the performance feel rather mechanical. A quick return visit to the Quantize Setup dialogue to add a bit of swing or some random quantise (to create some subtle 'human error') can breathe a little life back into the performance, as can some groove quantising.  

Published June 2011

What is Velocity Bias? SGX-2 Piano Sound Engine on NAUTILUS | Part 1

Send Effects: More Than Reverb & Delay

By Matt Houghton

Here, a tape-emulation effect has been set up as a send effect from the drum bus. With the high end rolled off, this can add solidity to the bottom end and mid-range of drums, leaving a more natural, airy sound from the cymbals.

We've discussed the difference between inserts and send effects many times in these pages, but for the newbies, let's spell it out: an insert slot is typically used for processors that change the whole signal — things like compression, gating, limiting and EQ, which can alter the character of the audio on the track — whereas send effects are typically used to add effects to a signal, such as by introducing delay lines or reverberation, while leaving the sound of the original track unaffected: you decide how much of the effect you want, using either the source track's send-level controls, or the faders of the effect 'return' channels ('FX' channels, in Cubase). I say 'typically', because you can, of course, break these 'rules' — and that's the subject of this month's column.

Before I get into any techniques, I should say that I'm not sure why Cubase has both FX and Group channels. You can send to a Group just as you can an FX channel, but Groups also seem slightly more flexible in terms of routing. For that reason, I often use Group channels in place of FX channels.

Reverb & Delay

I tend not to use reverbs and delays as inserts, because (a) they're generally relatively CPU‑intensive effects — so it makes sense for multiple channels to share the same instance of a plug‑in — and (b) I often want multiple source channels to share the same virtual 'space'. I occasionally use reverb as an insert to add a bit of body to a thin sound, using an ambience-only patch, before running the whole signal through a processor — but that's about it. I do, on the other hand, often find myself using and abusing Cubase's FX and Group channels, sending individual source tracks to processors for a variety of reasons...

Parallel Dynamics

Parallel compression is a common 'send trick', and we touched on this in SOS April 2009's Cubase column (/sos/apr09/articles/cubasetech_0409.htm). The idea is that you put a compressor on a Group channel and send a signal to it. You route both the source and compressed channels to another Group, and balance the processed and unprocessed signals to taste. The aim is the sort of increased tightness or 'focus' that you'd expect from standard compression, while still letting some of the dynamics of the original through. Some plug-ins (not the native Cubase ones, though) provide a wet/dry control, which in essence achieves the same thing, but these offer you less flexibility than genuine parallel compression, as when using sends you can process the original and compressed signals differently — for example, by applying EQ. Adding a subtle 'smile EQ' before mixing the two signals back together can yield a pleasing sound, and high and low-pass filters are particularly useful. For example, I sometimes use a similar technique on a drum bus, but substituting the compressor with a tape‑emulation plug-in. I'm able to decide which frequencies go to tape: I can leave the cymbals nice and open, but get the desired solidity by low-pass filtering before the tape effect somewhere around or below the 4-5kHz region.

Gates can also be used in combination with more conventional send treatments to create interesting spot effects. For example, a lead vocal or guitar part might be sent to a gate so that only the loudest notes are allowed through, and this signal can then feed a delay line.

De-ess Success

Sending to a Group channel to create a vocal de-esser. Note the automation singling out the offending 'esses', and the orange polarity-invert button on the dedicated 'ess' track, which causes the desired ess-cancellation.

A reasonably good-sounding de-esser plug-in comes with Cubase, but it offers little controllability — so if you need greater precision, try this trick.

• Create a Group channel and send the full source signal to it (set the send to value 0.0 and make sure it's a post-fade send).
• Reverse the polarity of the Group, using the 'phase' button at the top of its mixer channel. The two signals should cancel when summed, despite the meters on both channels showing activity.
• Set the original channel output to 'No Buss' to silence it (the Mute button would stop it feeding your Group), so you hear only the sent signal.
• Now automate the Group's fader so that only the vocal 'esses' are playing, and the rest is muted.
• Place an EQ in the first insert slot and use a high-pass filter, rolling it up the frequency spectrum until you filter out as much of the vocal, other than the esses themselves, as you can.
• Finally, set the vocal-channel output to the stereo bus again, and the Group channel should now only cancel the vocal track when the esses occur.

Of course, this is cancelling the whole of the vocal ess, so you need to fine-tune the automation to reduce the amount of cancellation. While the fader's at unity gain, there's full cancellation, and rolling it back reduces the amount of cancellation.

You'll often read about a technique known as 'multing' in the pages of SOS, particularly in Mix Rescue. This is where you duplicate a track so you can apply different treatments to the two parts in different sections of the song, without having to resort to detailed automation. You can achieve much the same thing as discussed above in this way, editing the waveform to leave only the esses, rather than using automation. However, that leaves you having to revisit your esses every time you make further edits or processing tweaks to the source channel — whereas using a send dynamically links the two parts.

Cutting Through The Mix

Automating the send level of a snare to a distortion processor can help the snare to cut through at critical points in the mix.

I often use distortion or a harmonic enhancer as a send effect. Most enhancers, such as the Aphex Aural Exciter, for example, work at least in part by adding harmonics that are related to the source signal, and blending them back in with the source, making it appear brighter or fuller. By placing the enhancer on an FX channel, you can route the full signal to it, and then filter out the unwanted frequencies — just as I described with the tape effect.

Overdrives, distortion processors and guitar amp and cab emulations also introduce harmonics, but they can be seriously unsubtle and produce some nasty side‑effects if applied to whole sources, such as vocals or snare drums. Even on sources more usually associated with distortion, like electric bass and guitar, they'll fundamentally change the sound character, and that's not always what you want. Using these as sends, though, you can be more subtle.

In a recent mix, for example, the snare drum was just as I wanted it for most of the track. The later choruses and outro, though, were very busy, with more elements piled into the mix, and the snare was struggling to be heard. I set up a Group channel, routed the snare to it, and dialled in some distortion using Cubase's bundled 'Distortion' plug-in, before rolling off the bottom end of the distorted signal. That left me with a snare 'presence' fader that I could ride up and down during those sections to get the snare to cut through. Of course, I also used automation to mute the distortion channel in other parts of the song. The same technique — without the automation — can also be used to fake an under-snare mic (maybe you had insufficient inputs on the session, for example), to give you a lot more control over the snare tone.

Transient Design

You might also use this 'send-multing' approach to layer and sculpt sounds. Let's continue with the snare example again, but this time you want more body and less 'snap', and EQ alone isn't cutting the mustard. One possibility would be to send the snare to a Group channel and then insert Cubase's Enveloper (or, better still, a third-party transient‑design plug‑in) to roll off the attack portion of the snare. You can then add a distortion effect, a compressor or limiter and maybe some EQ to sculpt this sustain section of the snare. Finally, bring up the Group's fader to mix the original and the duplicate. You could even set the output of the original to 'No Buss' but send it to two different Groups, one to control the attack and the other the sustain. You can achieve much more precise control over the results of Cubase's Enveloper in this way.

Conclusion

If you're new to either Cubase or the mixing game more generally, I'd strongly recommend that you start out by learning to use inserts and send effects in a 'conventional' way, as described at the beginning of this article. It remains a good rule of thumb to follow. But as you become more adept at using your processors, you'll find there are definite benefits to experimentation with those send effects, whether for tonal shaping or for problem solving.   

Published July 2011

modwave native - wavetable synthesis powerhouse - now available for Mac and Windows

Vocal Comping

By John Walden

Keep History provides the best Audio Recording Mode option when capturing multiple takes.

Compiling — or 'comping' — the perfect vocal performance from multiple takes is one of the most common editing tasks that DAW users undertake. We've already considered this process (back in SOS July 2007), but as the upgrade to Cubase v6 brought some welcome improvements in this area, it's time to revisit the subject.

What's New?

Cubase has allowed multiple takes to be recorded onto a single audio track for quite some time, and in previous versions these takes would be displayed as a series of 'lanes' inside the main audio track. Prior to v6, to get the best results you had to record in cycle mode, typically recording one section (for example, a verse or chorus) at a time. This worked well enough, but if you decided to return to a song section and try a few more takes, things could get untidy very quickly.

In v6, the lanes-based approach is much more flexible. Any recording made on an audio track that overlaps with a previous recording on the same track simply becomes another lane, whether the recording is made with cycle mode on or off, and regardless of when the recordings were made. You can, therefore, adopt a more flexible approach to recording multiple takes. The most recently recorded take assumes playback priority, although to get the most from the new comping features with vocal tracks, it's best to record all the takes for a specific song section on a single audio track.

It's also worth noting that Steinberg have revamped the recording mode options for handling tracking of multiple takes. For audio, three options are now available in the transport panel. If hard-drive space is not an issue, the default 'Keep History' setting is safest, as it retains all audio takes, whereas the two other modes will, to different degrees, remove earlier takes on a track when you record over the top of them.

Down The Lane

In this example, four vocal takes have been recorded, each displayed in its own lane under the master audio track.

Having captured your takes, the Show Lanes option needs to be engaged (the button goes blue in the Track List) to display them. Each take is then shown indented below the audio track, rather like the display of audio tracks that are nested inside a Folder track.

By default, the bottom-most (most recent) take has priority for playback. In the example shown below, this is in Lane 4, although note that because the take in Lane 4 is shorter than that in Lane 3, Lane 3 assumes priority for a few bars of this song section. It's also worth noting that the waveform display in the audio track itself shows the 'comped' vocal — that is, it shows the waveform from each take that has priority — in this case, a combination of take 3 (bars 3-9) and take 4 (bars 10-13).

Engaging solo on one of the individual lanes allows that lane to be heard on playback. As will become clear, this is useful if you want to audition a particular take without disturbing your existing comping. If you want to hear an individual take in isolation (that is, properly soloed without the backing track), you need to engage the audio track's solo button as well.

Cut! It's A Take

My own experience of using the new comping tools has resulted in a two-stage workflow. The first step is to split the takes into individual phrases, something the new tool set makes very easy. With the standard Object Selection tool (the arrow) selected, simply hold down the Alt (Windows) or Option (Mac) keys and then click at the position where you want to cut the take in two. This cuts all of the overlapping takes at this point, not just the currently active one. It is very neat for comping vocals, where the phrasing for a particular line is likely to be similar in each take. In the example opposite (top), note that the Snap setting is active; if you want finer control over the position of the cuts, turn this off.

If you wish to subsequently adjust a cut position for all takes simultaneously, simply hold the mouse over the cut in the upper half of one of the takes. The cursor changes into two vertical lines with arrows pointing in opposite directions and you can drag the cut point to a new position. Alternatively, if you want to change a cut in a single lane only, hold the mouse over the lower portion of a take and the usual adjust length tool icon will appear.

Having split the audio into distinct phrases, the comping process itself is easy. With the Object Selection tool selected, click on whichever take you wish to be given priority on playback; this automatically gives this take priority and mutes any overlapping sections of takes on other lanes. With playback cycling through the song section, auditioning the best combination of takes becomes an absolute breeze.

Range Finder

Having split the takes into suitable phrase-based sections, the various takes are then easy to select and audition using the Object Selection tool.

The approach described above generally gets the bulk of the work done very quickly, but if you need greater control over your selection (for example, to pick specific words from different takes), a second stage of comping is required. This is best based around the Range Selection tool, and can be used whether you have already split your track into phrase sections or not.

Having selected the Range Selection tool from the main project window toolbar (the square icon positioned next to the Object Selection arrow tool), the mouse can be clicked and dragged across the section of a take that you require.

As shown in the screen shot (right), this selection will be highlighted in pale blue. Double-clicking on this selection will set it as the active take, automatically creating cuts/mutes in all of the other lanes, as required. This is a great, hassle-free way of making precise selections, and is a big improvement on the functionality of previous versions of Cubase.

Curve & Bounce

The Range Selection tool has been used to select a section of the take in Lane 2 and this is highlighted in blue (above). Double-clicking on the selection sets that portion of the take as active, creating cuts at the same positions on all lanes (beneath).

Once you've compiled a 'master' take, a couple of finishing touches can really neaten things up, but before doing these, remember to save a version of your project under a different name, so that you can return later to fine-tune your selections, if you need to.

The first tidy-up task is to check for any audio glitches at transitions between the different takes. Cubase can automatically apply fades to audio clips and crossfades between clips (using the Project/Auto Fade Settings dialogue), and the default settings may be all that are required. If you still have some lumpy-sounding edits, you can also set auto fades at the track level by right‑clicking on the track in the Track List and selecting the Auto Fade Settings option. This dialogue has the option to turn off the global project auto-fade settings and create custom settings for that track. Manual editing of specific fades might also be used, but always solo the track to ensure you've dealt with any pops and clicks as cleanly as possible.

The final step is simply to bounce your selected takes from their various lanes into a single audio track. This isn't essential, as you can simply switch off the Show Lanes option to hide the lanes — but I find that this sort of tidiness helps to manage a project. Select all the events in the various lanes (both used and unused within the comp) and execute the Audio / Advanced / Delete Overlaps menu option. This removes any unused events and places all your selected sections on the top lane (a final solo audition is worth doing at this stage). Once you're happy that everything sounds hunky dory, simply turn off the Show Lanes option, select all the events you wish to combine and execute the Audio / Bounce Selection command. Click 'Replace' when Cubase asks you if you wish to replace the events and, hey presto, your ultimate vocal take exists as a single audio event.  

Advanced Tips

A couple of additional features are worth a quick mention. First, if you find yourself needing to adjust the timing of part of the performance in one of the lanes, holding down the Ctrl and Alt (Windows) or Command and Option (Mac) keys allows you to 'slip edit' the audio within an individual event — which is very useful when the performance is good but the timing of certain words needs tightening.

Second, the full version of Cubase 6 also allows you to perform multitrack comping processes using the Group Editing options in the project window. This is potentially very useful for editing multiple takes from multitrack recordings of backing vocals or drums — but that's a topic to save for another day! 

Published August 2011

SQ-64 Sequencing Strategies Masterclass with Kabuki - SQ64 xARP 2600 Patch Breakdown

Liven Up Your Drum Loops, Part 3

LoopMash 2 includes a number of new tools for creative loop mangling.

LoopMash caused quite a stir when released with Cubase 5, but it was significantly improved with LoopMash 2 in Cubase 6. While LoopMash's primary function is — well, the mashing of loops, obviously — it's also useful for other loop-based tasks, and that makes it ideal for refreshing your tired old loops. So let's get mashed...

LoopMash Basics

We took a brief look at LoopMash back in SOS October 2009, and the basic principles of operation remain the same, so I won't re-tread that ground in detail here. Instead, following a basic refresher of Loopmash's main function, we'll focus on some of the more creative possibilities the updated version presents.

Essentially, LoopMash is designed to 'mash' loops together by combining beats from one or more loops into an overall rhythmic pattern defined by a master loop. You slice sounds from up to seven other loops, and layer these over the master loop. Loopmash finds elements in these loops that are in some way sonically similar to slices in the master loop, and thjese similar slices are then triggered as the master loop plays back. How much emphasis LoopMash places on any of the other seven loops (that is, their relative 'similarity' to sounds in the master loop), how many slices can be triggered simultaneously, and the relative volumes of slices from each loop, are all controllable by the user.

Mashed

Using LoopMash 2's Always Solo and Exclude slice-selection modifiers for drum replacement.

If you've catalogued all your audio loops using MediaBay, as suggested in the first part of this series (in SOS September 2011), you should be able to drag and drop from there, or from the Project window, onto an empty track in LoopMash. LoopMash will usually do a decent job of recognising the loop length and tempo before slicing the loop, although I still experience the occasional glitch in this process, and find that temporarily changing the project tempo to match that of the loop before doing the drag and drop can provide a workaround in the event of problems.

The master track is identified by the lit button immediately to the left of the waveform display. Overall control of which of the additional loops might supply slices to complement the master track is set via the Similarity gain-sliders on the left of the display. A number of other performance controls remain the same as in the original LoopMash — and the Cubase 6 Plug-in Reference PDF does a much improved job of explaining how these various controls interact to create a particular 'mash'.

Sliced & Diced

The slice-selection modifiers provide some excellent ear-candy effects.

Even the shortest exploration of the supplied presets demonstrates that LoopMash is capable of performing fabulous loop mangling. However, as discussed earlier in this series, sometimes you need to be able to take sounds from a key loop or two to build your own variations, or add extra beats over an existing loop. Usefully, once loops are sliced inside LoopMash, individual slices can be auditioned (simply click on the slice) and then simply dragged and dropped on to Groove Agent One pad, so that additional hits can be triggered via MIDI notes. This process works very well and the only down side is that the rather wonderful slice-selection modifiers (described below) are not carried across. Maybe this is something that could be added in a future update (please!).

In essence, I suppose that this is not that far removed from the sort of drum replacement that can be achieved by combining beat slicing in the Audio Editor and dragging-and-dropping slices into GAO, as discussed earlier in this series. By default, LoopMash does this drum replacement in a 'with bells on' fashion, but with a few setting tweaks a more straightforward sound replacement job is possible.

Loopmash 2's Audio Parameters tab offers some new and useful tools...

Once you've identified your master loop and a loop (or loops) containing the sounds you want to use, you'll need to get busy with some of the slice-selection modifiers. Right-clicking on a slice opens a menu containing a large number of these modifiers, but for this task the most useful options are Always, Always Solo (both of which are only available on the master track), Exclude and Boost. Selecting one of these applies the modifier to the current slice, and a small icon is placed on the slice to indicate which modifier is being used.

For simple drum-replacement, the trick is to first use Exclude on both the master and other track(s), so that LoopMash ignores any slices with sounds you don't want to use, and then use Always or Always Solo on your master track slices for any sounds you want to retain. This allows LoopMash to focus in on just the sounds you want to use. If you then decide you want to hear more of a particular sound, you can also apply the Boost modifier to that slice.

Finally, you simply adjust the standard LoopMash controls for wet/dry mix, track similarity, track volume, and so on. Hopefully, the result will be a loop that is rhythmically the same as your master track but which features a replacement sound set. Given the way LoopMash does its similarity processing, the results are not as predictable as with a more conventional approach to drum replacement — but, in creative terms, I find that this is actually half the fun, as there's potential for happy accidents!

Get Creative

...and its Performance Controls tab reveals plenty of fun options for live tweaking of your loops!

Simply mixing and matching loops with the similarity sliders, and varying the global controls under the Slice Selection and Audio Performance tabs, provides a wealth of 'mashing' possibilities. There are some options here that are new in Cubase 6, but for some genuine 'ear candy', the other slice modifiers and the Performance Controls are the best of the new LoopMash features.

As with the Always and Exclude slice-modifiers mentioned above, the remaining slice-selection modifiers operate at the level of an individual slice. These include mute, reverse, staccato, scratch, backspin and various types of stutter, slur and tape effects. While only one such modifier can be applied to a slice at any one time, they're stored as part of one of LoopMash's 24 scenes for easy recall. You might, example, create several identical scenes based on the same loops and then apply different slice modifiers to each scene, switching between the scenes during playback (either by clicking on the scene pads or via the MIDI key range C0 to B1), to create variations on your loop.

A similar range of effects is available via the excellent Performance Controls, but these are applied across the whole of LoopMash's output, rather than to individual slices. As their name suggests, these are intended to be 'played' (via the on-screen buttons or MIDI key range C3 to G#4). Any fan of electronic processing of drum patterns will have a lot of fun with this.

More Than Beats

I've deliberately focused on using drum loops through this short series, but it's well worth me saying that all of these tools can easily be used to breathe new life into loops of other sounds as well. Bass loops make a particularly good target: you can achieve great results from something as simple as grabbing an individual note from a loop as a slice, and then re-pitching it in a sampler plug-in so that you can play in a variation on the original bass line. Indeed, given that Groove Agent One allows you to adjust the pitch of samples held on a pad, this is perfectly possible using just the tools that we've discussed in this mini-series. LoopMash also has plenty to offer in this instance: just load a couple of different bass loops and see what happens as you adjust the similarity sliders. You might be pleasantly surprised!

Beat This!

It's not my intention to stop you buying commercial loop libraries, but I hope you won't feel the need to buy them out of boredom with your existing loop collection — because this short series should have shown you that Cubase includes an excellent set of tools for making more from whatever loops you have at your disposal.  

Coming To The Third Party

The Sample Editor/Groove Agent One (GAO) combination for generating sliced drum loops and MIDI files described earlier in this series has only been available since GAO was introduced in Cubase 5. If you're using an earlier version of Cubase, and don't have access to something like Kontakt, Stylus RMX or Phatmatik Pro (all of which provide their own takes on this process), there are some freebie options. UVI Workstation (www.uvisoundsource.com), for example, is available for both Mac and PC. It offers the ability to slice WAV and REX audio loops and then drag and drop a MIDI file back into your host to trigger the slices. Unfortunately, there's no way to drag and drop individual audio slices from the loops to use in a drum sampler, but this could easily be worked around by recording individual drum hits to an audio track and working with these. 

Published November 2011