Audiomind's Turbulent and Kismetical Rambling Mind (audiomind) wrote in electrixa,
Audiomind's Turbulent and Kismetical Rambling Mind

MIDI & Digital Audio Music Sequencer Techniques

Music technology has finally gotten to the point where almost anything sonically imaginable can be achieved. Audio can be recorded, chopped into pieces, and then rearranged into something completely different - all with a few clicks of a computer mouse. Whole environments can be simulated so that while it sounds like your singing on stage in a huge concert hall, you're actually standing in the middle of your bedroom studio. Complete professional-quality songs can be recorded, produced, mixed, and mastered using a single standard computer system with MIDI/ digital audio sequencing software.

Of course, just because you can imagine something doesn't mean you know how to accomplish it. Today's digital audio sequencers pack so much power and contain feature upon feature, that it's easy to stay comfortable and just stick with the basics, especially when you're facing deadlines and don't have the time to experiment. 

We're going to take a look at some sequencing techniques that you can use in your day-to-day digital audio dealings. We'll cover a number of generic techniques that will work with any sequencing software you may have. I'm going to assume that you already know the ins-and-outs of MIDI and digital audio. Ready? Set? Let's go...

The real-time effects built into many sequencers aren't the only way to add excitement to your MIDI and digital audio tracks. With a few applications of the copy, paste, and pitch shift commands, you can easily simulate a number of different effects including delay, flange, and chorus. This can free up some of your computer's precious processing power for handling the more complex effects like compression and reverb.

Digital delay. This technique actually covers a number of different effects, and can be applied to both MIDI and audio tracks. Simply copy the MIDI or audio clip that you want to process, and paste it to a new track. Then slide the copied clip forward in time to create a delay. Depending on how much of a time offset you use, different delay- related effects can be achieved. Slide the clip between 5 and 10 milliseconds and you get a thickening effect. Slide the clip between 10 and 20 milliseconds and the thickening turns into doubling. Anything above that amount and you start creating discrete echoes.

If you want to create a delay that is synchronized in time with the music, just move the copied clip ahead by snapping it to a note value such as a sixteenth. To make things more realistic, make a few clip copies and slide each one a sixteenth note apart. Then lower the volume of each successive echo so that the effect fades over time. As an example, for three echoes, lower the volume of the first to 75%, the second to 50%, and the third to 25% of the original. You can even get a little crazy by sliding each copy to a different uneven time offset, and maybe give them different pan positions too. You can't get that kind of precise flexibility with a real-time effect.

Flange. This effect can be applied to both MIDI and digital audio tracks but they require slightly different procedures. As with delay, simply copy your clip to a new track. From here, you can easily produce a static flange effect by moving the copy forward in time from 1 to 5 milliseconds. While this creates an interesting sound, it doesn't have that authentic analog sweeping that we all love and cherish. To achieve that, we need to use a little pitch shifting.

For audio, instead of moving the clip copy, just shift its pitch anywhere from -12 to +12 cents. This will make the copy longer or shorter by a minute amount, causing it to drift out of sync with the original. Of course, depending on how long the clip is, eventually there will be so much of a delay that the flanging will turn into doubling. To get around that, just shift the pitch of small segments of the clip, alternating the direction of the shift with each segment. For example, shift the first two seconds of the clip by -9 cents, the next two seconds by +9 cents, and so on. This is also how you can control the rise and fall of the sweeping characteristic.

For MIDI, the procedure is similar except you need to apply pitch wheel controller changes to the clip copy. Just draw in some pitch controller changes over the course of the clip that vary by rising above and falling below the zero axis. I found it best to stay within the +600 to -600 bend range because any changes larger than that start sounding like detuning rather than flanging. To apply precise changes, many sequencers will allow you to insert a series of controllers that change evenly from a start value to an end value over a set amount of time. This will give you more precise sounding effects but you may prefer the more random fluctuations you can get from just drawing the controller changes in by hand. Either way, the effect is not quite as authentic as with audio. Many times it sounds more like chorusing than flanging.

Chorus. You create audio chorusing in almost the same way as flanging except you need to stretch the pitch, rather than shift it - meaning you need to change the pitch of the audio clip without changing its length. Most programs have this capability. To apply chorusing to an audio clip, simply copy the clip to a new track and stretch its pitch, plus or minus 9 to 12 cents. The more you stretch it, the deeper the effect. You can also use this technique to create a pseudo-stereo effect from mono parts by simply panning the original clip and clip copy to different locations. The more pan, the wider the stereo image.

If you really want to go wild, you can create something I like to call the Super-Combo Chorus Effect. This time, instead of creating just one clip copy, make 3 or 4 or more. Stretch the pitch of each copy using slightly different values from the rest. Now slide each copy in time using the delay techniques described earlier with slightly different values. For the final touch, evenly pan each copy throughout the stereo field. You can also give each copy a different volume level for more subtlety. When you're finished, hit play to experience one very cool chorus simulation.

When destructively applying plug-in effects to digital audio, many sequencers will let you either replace the original material or leave the original material intact and create a new track (or set of tracks in the case of stereo) for the new effected material. If you adjust the levels of the effect module so that the dry mix level is 0% and the wet mix level is 100%, your new track(s) will contain the effect output minus the original signal. Why would you want to do this? Because it gives you some very interesting possibilities.

For instance, you can now control your track effects in a variety of ways. Using volume controllers you can bring the effects track(s) in and out to add an artistic flair to the mix. You can also use panning to move the effected signal around the stereo field while the original signal either remains stationary or moves around as well but independently. Even more possibilities arise when you consider that you can add effects to the effects track(s). How about applying different EQ to the original and effected tracks? If your effected tracks contain reverb, how about applying another reverb effect on top of that for some very far out spatial displacement?

And if you happen to have multiple soundcards or multiple outputs, you can assign your effects track(s) to their own outputs. This way you can even use your outboard modules to add effects to the already effected track(s). Experimentation is the key here.

You've probably heard of the term "Groove Quantizing" or something similar, where a sequencer uses a pre-defined MIDI pattern that contains data in a specific rhythmic style and applies it to recorded MIDI data so that it conforms to the same rhythm. This feature is a great way to breathe new life into a tune or even give it a different sense of feel.

Well, in case you didn't know it, you can also use "Groove Quantizing" on audio tracks. You just need to do one small preparatory step. You need to break down the audio track (or audio clip) into small pieces or clips. Some sequencers will have an automatic feature for this. If yours doesn't, you can try exporting your track to an audio editing program like Sound Forge. Using Sound Forges' Auto Region and Extract Regions tools, you can easily break down a large audio file into smaller ones that contain a quantizable musical unit such as one beat of data.

When you import the small audio files back into your sequencer, each one will have its own start time, just like a MIDI event. This lets you easily quantize the playback of the audio segments, and you also don't get the weird anomalies that can arise from trying to stretch the audio. This technique works especially great with percussion clips but don't rule out instrumental or even vocal clips. Try turning a normal vocal track into a cool rhythmic scat. It's a lot of fun!

In addition to responding to the usual MIDI controller messages such as pan, volume, and pitch bend, most synthesizer modules have a number of "hidden" parameters that can be of great use in adding expression to your tunes. The way to get to these parameters is through the use of Non-Registered Parameter Numbers (NRPNs). As an example, let's take a look at the Roland Sound Canvas. This modest module responds to no less than 13 different NRPNs (see Table 1) that allow you to control things like the attack, decay, and release of a patch's envelope, as well as the cutoff frequency and resonance. As a matter of fact, by manipulating these parameters you can create brand new timbres based on the existing sound set, which is otherwise uneditable.

Table 1: NRPN's for the Roland Sound Canvas
Param. Control 98 Control 99 Control 6
Vibrato Rate 8 1 14-114
Vibrato Depth 9 1 14-114
Vibrato Delay 10 1 14-114
Cutoff Frequency 32 1 14-114
Resonance 33 1 14-114
Attack Time 99 1 14-114
Decay Time 100 1 14-114
Release Time 102 1 14-114
Drum Pitch Drum#0-127 24 0-127
Drum TVA Drum#0-127 26 0-127
Drum Pan Drum#0-127 28 0-127
Drum Reverb Drum#0-127 29 0-127
Drum Chorus Drum#0-127 30 0-127

For instance, let's say I wanted to change my basic piano track into something with more of a string feeling but keeping the piano timbre. I would open the event list for that track and insert nine new controller messages (see Table 2). The first three messages would set a new envelope attack, the next three would set the decay, and the last three would set the release. Controllers 98 (Non- Registered Parameter LSB) and 99 (Non-Registered Parameter MSB) are used to set the type of parameter to be changed (in this case envelope attack, decay, and release), and controller 6 (Data Entry MSB) is used to set the value of each selected parameter. All of the settings for each of these controllers depend on what MIDI device you're using them with. Have a look in your synthesizer manual and you might be surprised just how much control you have over that little black box.

Table 2: NRPN Example
Message 1 - Control: 98 Value: 99
Message 2 - Control: 99 Value: 1
Message 3 - Control: 6 Value: 70
Message 4 - Control: 98 Value: 100
Message 5 - Control: 99 Value: 1
Message 6 - Control: 6 Value: 100
Message 7 - Control: 98 Value: 101
Message 8 - Control: 99 Value: 1
Message 9 - Control: 6 Value: 70

A couple things to keep in mind when using NRPNs - send any patch changes down the wire beforehand otherwise you may end up with a different sound than what you were aiming for - although this could result in a pleasant surprise. The other thing to remember is that controller 6 is dynamic, meaning it changes the value of whatever was the last parameter that was set with controllers 98 and 99. In the above example, if later on in the track I sent another controller 6 message, this would change the envelope release since that was the last set parameter. In order to change either the attack or decay, I'd have to resend the appropriate values for controllers 98 and 99.

Another exciting aspect about NRPNs is that they can be sent in real-time, just like any other controller messages. What this means to you is that depending on your synthesizer module, you can now dynamically change the characteristics of patches or any other supported parameters and record those changes into your mix. In the case of the Roland Sound Canvas, you could easily add expressiveness to your sounds by varying their cutoff frequencies and resonance. In addition, you can dynamically vary the pan position, volume and pitch of the individual drums in a set. That could make for some very cool percussion parts.

Of course, typing in all these parameters would be a bit tedious. A better way would be to setup a special mixing template for just this sort of task. Most high-end sequencers provide the capability for you to create your own virtual mixing controller. In Steinberg Cubase the feature is called the MIDI Mixer. In Cakewalk Pro Audio and Sonar it's called StudioWare. Everyone has a different name for it but the feature is still the same. If you're lucky, your software will come with a pre-made template for your MIDI instrument. Cakewalk has a Roland GS template for use with a Sound Canvas but it's missing any drum parameter controls.

If I wanted to add individual drum panning to the existing Cakewalk Roland GS template, I would open a new copy of the template into my existing project. I'd then add a new Cluster widget to hold the new set of controls. Then I would add a slider object to represent the drum instrument number. From there, I'd set the necessary properties for the slider including Label (Drum Number), Alias (Drum_Number), Automate in Track (AutoTrack), Range (0-Min, 127-Max), Primary Action (Kind: Controller, Channel: Part, Number: 98, Value: Drum_Number), and Return Action (Kind: Controller, Channel: Part, Number: 99, Value: 28).

The Alias acts as sort of a program variable which is then used in the Value setting of the Primary Action property to tell Cakewalk to use the value of the on-screen slider (0-127 set in the Range property) as the MIDI controller value. In contrast, the Value setting of the Return Action is permanently at 28 because MIDI controller 99 is a constant value in this case. The Label is just an on-screen text name for the slider and the Automate in Track property just tells Cakewalk which track to record the automation data from this slider to during mixdown. In this case, AutoTrack automatically sets this property to whatever the currently selected track may be.

To complete the drum panning controls, I would also need to insert a Knob widget to represent the drum panning value. With this new control set I could now easily select any of the Sound Canvas drum instruments by moving the slider to the appropriate number and then pan the instrument using the knob - recording any changes to a designated track. Of course this whole procedure would be a bit different in another sequencing program but the concept is still the same. Once you learn how to create your own virtual mixer templates, the possibilities are endless.

When inspiration hits, you don't want to waste your time fiddling with sequencer setup parameters. You want to be able to start your software and get right to work. If you create a template file that contains everything set just the way you like it, you'll have a much better chance of getting that cool lick down before you forget it. Here's what you need to do to create the ultimate project template...

1. Use the File > New command to create a blank project. Then setup the track display with all of the instruments in your studio. Set the name, input, output, channel, bank, patch, volume, pan, etc. for each track.
2. Go through the rest of the program and set things up exactly the way you like them. That includes window positions, zoom, and even the last tool used. In Cakewalk Pro Audio and Sonar other parameters like timebase, file information and comments, tempo settings, meter and key settings, clock and synchronization information, and more can all be saved.
3. If your software has a built-in System Exclusive librarian, fill it up with all of your favorite Sysex macros, so that you have them readily available. This could be anything from a favorite bank of patches for one of your synthesizer modules or even basic messages like: Roland GS Reset (F0 41 10 42 12 40 00 7F 00 41 F7) and Yamaha XG Reset (F0 43 10 4C 00 00 7E 00 F7). If your software doesn't have a Sysex librarian feature, then just designate a special track for Sysex messages and save each macro as a separate sequence or clip. Be sure to keep that track muted so its data won't be transmitted during playback. When you need to use a certain Sysex macro, just drag-and-drop it to an open track.
4. Taking the Sysex track idea one step further, you can also create special tracks for your favorite MIDI controller macros. A MIDI volume track can hold different sets of volume changes such as fade in and fade out segments for a number of bars. A MIDI pan track can hold different sets of pan changes. Some interesting ones include a left to right sweep, and right to left sweep, and a center to right to left back to center sweep. In addition, you can create tracks for your favorite drum patterns or sequenced melodic patterns as well. And if your sequencer allows you to store MIDI and audio data together in the same file, you can set up tracks containing some of your favorite digital audio samples and loops too. This way you have all the tools you need right at your disposal for quick and easy drag-and-drop use.
5. The final step is to save your file with all of its settings so that when you open it again, everything will be perfect for your next session. If your sequencer supports it, you can also have the template file load automatically every time you load the software. For Steinberg Cubase on the Mac, name the file Autoload. For Steinberg Cubase on the PC, name the file DEF.ALL. And for Cakewalk Pro Audio and Sonar, save the project as a template file called NORMAL.TPL.

Today's digital audio sequencers provide you with a lot of power to accomplish your musical goals. Many times you can achieve the same results in a number of different ways. The tips we've covered here are mainly productivity boosters and techniques you might not have otherwise thought about.

One final word of advice is to go out and buy yourself a small binder to use as a studio journal. As you work on different projects, take a few moments to jot down any new techniques you may come up with in the process. It's also a good idea to reference what you read about in publications, so that you can easily go back and find the issue and page number where you initially saw it. Not only will this save you time on future projects, but eventually you'll have a nice collection of information to refer to.

While technology shouldn't govern the way you make your music, it can certainly help you create the songs of your dreams. The more you learn about how to utilize the tools at hand, the easier the process will be.
Tags: digital audio, midi, music production techniques, sequencers
  • Post a new comment


    default userpic

    Your IP address will be recorded