Rosegarden is a useful program for producing music; it has all kinds of features, and a lot of flexibility in how you want the music to come out. It is a combination MIDI sequencer and DAW, and has support for a number of synthesizer and sampler plugins, as well as multiple views for editing music.
Now, I’m not a musician, and if you heard what I have made for myself… well, I wouldn’t need to kill you, because the sounds loosely termed as “music” would already have done my work for me. But, from time to time, I do render MIDIs I download from the net, just to enjoy a little practice of singing to them, or just listening to stuff with the words taken out, so I have had a bit of practice with MIDI using Rosegarden. And while I have no musical talent per se, I do understand enough music theory that I can probably make a decent… article writer.
There are a lot of features to this program, and I don’t really have them all down, but a good starter might be useful for those who want to bang out a quick bit of music. The one thing I do want to mention, since I’ve heard a number of people dismiss this as a musical score printer, is that this is much more than just a printer; it is a full, multitrack sequencer and DAW program, capable of using MIDI instrument plugins and external synths and samplers to generate music.
The basic program, fresh from startup, looks like this:
There are two major toolbars, consisting of several different groups of related commands.
This is the most obvious group, consisting of “New Project,” “Load Files,” “Save Project,” “Print” and “Quick Print” (for printing sheet music using LilyPond), Undo, Redo, Cut, Copy, and Paste. No real rocket science here, right?
One thing to note; the “Load Files” button can load a project, but it also serves to import a source file as well; this will allow you yo actually open a MIDI file as a base for a Rosegarden project (which is much more than just a MIDI file; it handles MIDI and analog audio).
These options are for adjustment of track segments (subsections of a track containing audio or MIDI data). The first option selects an existing segment, the second creates a blank segment, and the third deletes an existing segment. The fourth and fifth are move and resize, respectively.
Following the segment editing commands are two types of segment split. The first will split a segment vertically, cutting it into two halves, one half before the split, and the second half after the split. This is useful if you want to detatch a bar from a larger segment, and then, perhaps assigning it to repeat, making a loop.
The second (called “split by pitch”) will split the segment horizontally, the first making one segment with the higher notes, and the other with the lower notes. This way, you can move the lower-note segment to a different track, and perhaps give it another instrument to play that better emphasizes the area in which the notes reside.
Finally, there is quantize. Now, when someone plays a keyboard, their fingers will generally press and release the keys fast enough to make the music sound… roughly like it should. However, there are small imperfections in the timing that can bring notes juuuust out of sync.
Quantize theoretically should be able to fix those slight timing issues; the point is that each note is brought to the exact length of the grid in which it resides. However, sometimes the timing of a note is just a little too long or a little too short, and the quantized result would end up making it sound even worse (makes a long note even longer, or a short note even shorter).
Unless you’re ready to deal with the consequences, you might just want to pass on quantize for now.
Rosegarden is a multitrack application, so obviously, you’ll need to manage the tracks. These controls are what you’d use. The first button allows you to make a new track. The second deletes the selected track. The third and fourth moves the selected track up and down in the listing.
The final two buttons affect all tracks; the dark dots will mute all tracks, and the lit ones (surprise!) unmute all tracks. No real mystery here.
This next group is extremely important toward the placement of music; these allow you to edit the segments directly, whether it is to input notes, percussion, or sounds that will be used in your composition. Each will open a special window, which will be discussed later in the article.
The first control will open the “Matrix Editor” window, commonly known as the “Piano Roll.” This is named after the rolls of music that were used with the classic player pianos, which allowed them to play published music without anyone actually at the keyboard. This is the most popular method of music entry in most common music programs today.
The second control will open the percussion matrix editor. This has a very similar appearance to the piano roll editor, and for good reason: why mess with something that works? The actual name of the drums, however, will replace the visual representations of the piano keys in the roll, allowing you to know which instruments you’ll be working with.
The third, the notation editor, is great for those who are more familiar with sheet music than piano rolls, as it will allow you to edit music using classic sheet notation, complete with many different forms of marking. In a way, this can increase the flexibility of the music, since there are features that can be added to the music in this view that would require additional settings in the piano roll view, such as note slurring and volume control.
The fourth one is for MIDI experts, as it is essentially a view of the raw MIDI data in the track; when each event occurs, what duration it has, what type of effect is has on the music, if it’s a note, what pitch it has, and at what velocity the key is pressed, and so on. If ever you wanted to know what actual MIDI looks like, this is the place to look!
The fifth one is the pitch tracker. Unfortunately, this is one of those areas where I haven’t a clue how to use; it is almost identical to the notation editor, but it seems to have some form of tuning information at the bottom. If anyone can clarify the use of this display, I would appreciate it.
Onto number six. This is an audio file manager; in here, you store actual wav files to use as segments in the audio tracks. Just load the file into this list, and then drag it to the audio track of your choice. This manager will show you the exact duration of the audio, the waveform of the audio, the sample rate, channels, resolution, and the filename. Unfortunately (to a FLAC user like myself), this can only use WAV files.
MIDI Plugin/Connection Controls
The next two items are extremely important. Rosegarden is probably the only program where I would not recommend using the Jack connection graph to make the connections between MIDI devices; this is because Rosegarden has an internal group of settings that should also be set at the same time that you connect your keyboard, synth, sampler, or instrument plugin to Rosegarden; this is important to make track settings much easier, and to make the sound actually work.
The first button is the MIDI Device manager. In here, you create internal MIDI interfaces, and then assign MIDI devices to them. These internal interfaces are what the MIDI tracks actually send their musical data to, and when you select which interface you want for each track, it is the internal interface’s name you will see. You can also assign MIDI input devices as well, such as keyboards, drum pads, and control surfaces. Note: any synth or sampler you set up from this dialog will use its own internal audio jacks, rather than outputting it through Rosegarden’s master bus.
The second button is to set up internal synth plugins; where the MIDI device manager handles connections to external synths and samplers, this will allow you to use instrument plugins internally to Rosegarden. These include the DSSI softsynth plugins like Hexter and Calf Monosynth, but you can also use VST Instrument plugins, provided the software has been set up by the system with the support. Plugins in this dialog will be sent through the Rosegarden master audio jacks, rather than through the external synth’s own jacks.
No DAW would be complete without a mixer component, and Rosegarden is no exception. These two buttons control the mixers for the MIDI and audio data, respectively. There’s not much to these mixers, so I don’t use them for more than adjusting volume to prevent peaking; all the final mix is done in Ardour, since its track and filter plugin architecture is infinitely more flexible.
Another aspect to a DAW that would leave it incomplete otherwise would be the transport controls; the play, pause, stop, rewind, and so on and so forth. Most of these are so painfully obvious, that I don’t really need to tell you that they are “skip to beginning,” “rewind,” “play,” “fast forward,” “skip to end,” “stop” and “record.”
After record, however, are two more buttons that are not as obvious. The first is a setting; when selected, like you see here, the track display will scroll in order to keep the play line in the window, so you can follow the music as it plays.
The final button is important when dealing with MIDI. You see, MIDI plays a sound when it receives a “note on” command. However, if the playback is stopped before it receives the “note off” command, and something prevents it from receiving the “note off” command the stop button should be sending, and the note in question is in a non-decaying instrument (it doesn’t “fade off”), then it will keep playing forever. The “panic button” is placed here so that, when pressed, it will reset all the instruments to stop any constant note from continuing to play.
Finally, segment placement on the track display is important, in order to synchronize the playback, so you want to be able to zoom in on the exact time the segment should begin. The slider is pretty self-explanatory.
The first bar is the tempo bar, showing the tempo (beats per second) of the song in question. The second bar is the time signature, beats per measure, and type of note used for beat. The third and fourth are the measure and beat counters, and you can see them on both the top and bottom.
After the rulers, you start to see the tracks.
From the left, you can see the track number, the mute control (lit=not mute), record arming (lit = this track is recorded to), specific name of the whole track, and then the segments that are in the track.
Inside the segment, you will see the name of the segment (usually a MIDI instrument name or audio filename by default, but it can be relabeled), and a visual representation of the audio data, either showing a miniscule copy of the piano roll (for MIDI) or the waveform (for audio).
Each track and each segment has its own settings. Because of this, tracks and segments can be selected, and on the right-hand side, you will see the parameters for the current segment/track. (If you don’t see the parameters, it can be turned on by going to “View” -> “Show Special Parameters”)
The label is usually, by default, the name of the MIDI instrument chosen for this track. However, it can be changed to whatever the musician seeks to make it represent. Color is also a visual setting with no effect on the content.
“Repeat” is a looping command; essentially, the segment will be repeated until it reaches another segment in the track; if you just want to repeat the segment for a certain amount of time, you can always add a blank segment at the point after the last desired repeat.
Transpose is used when you want to adjust a track a certain number of semitones; in case the synth/sampler plays the note at the wrong pitch. You can adjust the pitch up to 48 semitones (4 octaves) from its original level.
Delay will adjust the track to play slightly behind the beat by either a certain number of milliseconds, or by a specific note type. This is very useful in the practice of syncopation, in which the energy or emotion of a song comes from the melody’s tempo being slightly different than the percussion’s beat.
Please note: this will only allow you to slow the beat behind the current track, so it will wait that specified time after the beat begins to play the note. If you want fast syncopation, you need to move the track one beat forward and use the delay to bring it to the desired moment before the beat.
As mentioned above, quantize will try to align the playback to the grid. It can be on or off. It can be useful, although I would avoid it if you’re trying for syncopation.
The first two are the reason that you need to use Rosegarden to set up the MIDI instead of the Jack connection graph. For every MIDI synth, sampler, or instrument plugin added and assigned in Rosegarden, a new entry will be added to the “Device” list. Then, you select the channel for that specific device to determine which instrument is played by this track.
The recording options are for recording audio and/or MIDI. By default, the armed channel will record all MIDI and/or audio input; which is useful if you just want to bang out one track or another, one at a time. However, if multiple people are sending signals to the program at once, it helps to ensure that all incoming MIDI and audio data is individually routed to different tracks, so you may want to adjust these to send each MIDI (or audio) Interface and channel to a different track.
The next two are for the Rosegarden LilyPond exportation notation. The size determines the size of the notation on the staff, while the bracket type determines the brackets used with the staffs. The latter is reported to be fragile, so it is probably safer to just leave as the default value.
The final set of options are a mirror of the segment parameters. These are the default values of all segments created in this track. One common change would be to adjust the color of this track; after all, if each track is a different color, it becomes easier to see the upcoming music at a glance without constantly checking the track name.
Each track has its own channel; all segments in the track make use of this channel to output their note information to the synthesizer or sampler. This box determines the settings of the device, in order to ensure that the sound being produced is mixed and configured properly.
Many such devices tend to support the General MIDI standard, in which the device actually contains one or more banks of programmed instruments that can be played at will. Additionally, one or more of the instruments, or sometimes even whole banks, will consist of a collection of percussion instruments that can be used to produce the backing beat to the music.
The first option determines whether or not the current instrument is a percussion collection or an actual tonal instrument. Checking the box will tell Rosegarden that the instrument is a percussion collection, and the Percussion Matrix editor will be the default editor for the track.
The second option determines the instrument bank to use; General MIDi is the default, consisting of 128 instruments defined by the General MIDI standard. Additional banks can be selected if the device supports them, with additional instruments that are not necessarily supported by the standard.
The third option is the instrument itself; it will drop down a list of available instrument sounds supported by the General MIDI standard. This list will be different, depending on the bank chosen.
The fourth entry allows external control devices to control the settings in this section. Not very useful for selecting an instrument, but essential if you want to have a control surface adjusting the other settings.
What settings? Well, the next four items determine the track’s volume, panning of the sound in a stereo field, the chorus settings (allowing for a wider-sounding instrument), and the reverb (for a deeper sounding environment).
MIDI Output Connections
Now that we have covered the track display, the next thing we need to focus on is the connections for synths (mathematical sound generators) and samplers (plays recordings of actual instrument tones)… after all, without these, all we have are little dots in colored strips.
MIDI Instrument Plugins
If you just want to start making music, and don’t care about mucking about with external synths or samplers, be they hardware or software, the plugin support for Rosegarden is quite good. The software can handle a maximum of 24 different software plugins at one time, many of which can handle multiple sounds simultaneously, meaning you can use one plugin on multiple tracks, if desired.
To set up a plugin, you need to go to the Manage Synth Plugins dialog, which can be accessed using the button pictured to the left. This will open a box similar to the one above, except that, at first, all entries will simply show “<none>”. By clicking on one of those “none” entries, you will see a list of the plugins available for use by Rosegarden. As you can see above, there are quite a few options, and VSTi plugins can be used in this way (as shown by the inclusion of Safwan Matni’s excellent “Matnitron Universal Strings” plugin in #5).
If you need to make any basic adjustments to the plugin, you simply need to click on the “controls” button, and a small dialog box will appear with the plugin’s available settings:
The top orange button will contain the name of the plugin (orange means that the plugin is active). When you click the button, you will see the control dialog appear. Underneath it is the “editor” button, which launches the plugin’s own UI.
The button to the right of the editor is the soundfield button, which can allow you to decide whether the track is stereo or mono; it is set to stereo right now, and the mono setting will just show one circle. To the right of the soundfield button is the panning control, which determines where in the stereo field the instrument will actually reside.
The next five buttons are for effect plugins; this is where, if you plan to solely use Rosegarden, you can apply effects such as delay, reverb, compression, and so on. As a side note, in the same way that Rosegarden can handle VST instrument plugins, it can also use VST effect plugins, and they will be available if you so choose.
In this case, I am using the “Ambience” reverb plugin. Just like the VST instruments, you can also choose to use their native UIs as well:
You might notice a little dot just above the orange instrument plugin button. By clicking this, you can set the name of the plugin to be something other than “Synth plugin #5”.
External MIDI Devices
Plugins are all well and good, but what if you have a soundfont, or a Gigasampler file, or wish to use a non-plugin synth? What if you have a hardware synth or sampler that makes the absolute perfect sound for your needs? Well, Rosegarden’s there for you!
As mentioned before, and should be mentioned again: You need to use these controls to set up MIDI in Rosegarden. Chaining things up in the Jack Connection graph is not enough; Rosegarden needs these internal settings in order to assign the instruments to its individual tracks. Failing to pay attention to this is just that: Fail. No Music for You!
That’s right, I didn’t say Jack, I said “ALSA.” Jack is primarily an audio chaining tool, but while it can have MIDI support, it doesn’t generally need it, as the ALSA driver tends to handle this internally. To be fair, there are reasons to prefer Jack MIDI over Alsa MIDI, and there are reasons to use the “A2J MIDI Daemon” over both, but these reasons are in the domain of harder-core musicians who use MIDI extensively; since I am not, and the point is moot for Rosegarden, anyways, we’ll just move on.
This is not to say that Jack doesn’t have a place, however. Synths and samplers are programs that receive MIDI instructions, and churn out actual audio data, so while their MIDI input connection is handled by Rosegarden, their audio output most definitely needs to be going somewhere.
In this example, I have four fluidsynth samplers (soundfont-based sound generators) running off of a QSynth GUI. I am not planning to do any recording at this time, so they will be output to my system speakers (set up using the technique from the article “Using Multiple Devices with Jack“)
Here is the connection graph as I have it set up.
The Rosegarden master tracks are where the MIDI plugins output. Rosegarden does have a setting to allow all tracks to have their own outputs, but by default, Rosegarden has 64 tracks, 32 audio and 32 MIDI, and I don’t want to go through the trouble of deleting them all. 😉
The “qs_RealFont1” entry is a FluidSynth sampler making use of the RealFont soundfont, which is an excellent-quality free soundfount that consists of all the best soundfonts out there in General MIDI format… an excellent starting point for any beginning musician’s toolkit.
In both cases, they are connected to the system speakers, so that they can be heard while working (without wearing the headphones attached to my “system” connections).
Next, we go into the “Manage MIDI Devices” window to make the devices and the connections. In this case, we don’t need to make any additional connections, although I’ll show you shortly how to add additional synths to the program.
The window you’ll see will have two parts, separated into two parts themselves. For the process of hearing MIDI, we’ll focus on the top section, “MIDI Playback.”
The default MIDI playback device in Rosegarden uses the “General MIDI” settings by default, so you can select the bank and program in the instrument drop-down, and the appropriate instruments will be displayed. In this case, it also correctly identified the RealFont1 sampler as the one we want to use.
However, if it was RealFont2 we wanted, we would just select the output on the right so that instead of 129:0, 131:0 would be attached to the General MIDI device, instead.
Now, what are all those buttons underneath the MIDI Playback list? These affect the settings in the list above them. The “New” and “Delete” buttons add and remove MIDI devices from the listing, Banks sets the names of the banks and their programmed instruments, and Controllers adds, removes, or adjusts additional effect controllers to the device, such as sustain, pitchbend, and chorus.
Now, suppose we don’t want just one sampler? What if we also wanted to use the ZynAddSubFX synth? For those who don’t know, unlike FluidSynth, which is essentially a sampler, ZynAddSubFX is an actual, mathematical sound synthesizer; sound samples do not apply.
Well, first, I’ll start ZynAddSubFX…
So now, we have two different devices connected to Rosegarden, each with a different name. I could also rename “General MIDI Device” to “RealFont” or “FluidSynth,” but it’s not necessary for the purpose of demonstration (and I don’t want to go completely nuts with the screenshots… well… any more than I am already… or maybe just because I’m getting lazy in my old age).
On the bottom half, there is the input devices used to add note data to your segments. These are simpler in that the inputs don’t need any special parameters; you just add the input, name it, connect it to an Alsa-recognized device, and off you go. For the sake of not repeating myself, I’ve already named the devices and linked them accordingly.
Tracks are empty by default. In order for any sound to be made, you have to create and edit musical segments in the tracks. Creation, manipulation, and deletion of segments are handled by the segment controls, shown before, while the actual entry of audio and/or notes are handled in the segment editors (or directly recorded by arming the track, setting the input device, and pressing the “record” button).
Creating a Segment
To begin, you need to make a segment. The pencil icon in the segment controls will change the mouse to “segment creation mode,” in which a new segment is made at the point the mouse is pointed. This segment will be, by default, one measure in length, and completely empty. It will look like this:
For clarification: The segment is the beige/brown item at the bottom with the words “Acoustic Grand” in it. If expanded, it will say “Accoustic Grand Piano,” which is the first instrument in a General MIDI collection.
Resizing a Segment
Let’s expand it, and you will see the rest of the word. To expand it, you will need to click the two-arrow icon to the right of the pencil, beyond the four-arrow “move” control. Once selected, the mouse will then resize any segment it comes into contact with. In this case, we drag the right end of the track… let’s say two measures further.
Moving a Segment
It is possible that we don’t want the segment to be there, at the very beginning of track 1. Maybe we clicked in the wrong spot, but didn’t realize it until later? Not to worry, the track can be moved around quite easily using the “Move” mode for the mouse, which is the four-arrowed icon mentioned earlier.
Splitting a Segment into Two Segments
However, suppose, after adding some notes, we want to split it, so that part of it can be looped? Well, in this case, we want to split the segment into two segments, so that one of the segments can then be repeated using the “Repeat” option in the Segment parameters. The mouse can be put into split mode by clicking the “torn-segment” looking icon. At this point, wherever in the segment you click, the split will then be there… and the segments can then be moved and adjusted individually.
Selecting a Segment for Adjusting Parameters
Now, I mentioned repeating. Rosegarden has a feature that allows you to loop a segment until it comes to another segment; an easy way to add loops to a mix without having to add every single iteration every time. But first, I want to move the segments around so that the smaller one is at measure 1, and there’s a measure between the smaller and the larger segments:
In any of the manipulation modes, the mouse will only select the segment as long as the button is held down, which is pretty useless if you want to adjust the parameters of that segment. The selection mode is the answer, and is the default state of the mouse when Rosegarden is started. To return to selection mode, simply click the “mouse arrow” item in the bar.
Setting a Segment to Repeat (Looping a Segment)
You’ll notice that the colored portion ends around the “P” in the name, but the name itself continues over a gray box. At this point, that segment is now repeating itself for one measure before it hits the second segment and stops. If we were to move the second segment… say… half a measure, the loop would continue until it reaches the second segment, basically repeating one and a half times.
Deleting a Segment
Now, what if we decide that the second segment doesn’t need to be there, that the first should just keep looping? Well, we can then delete the second segment, and let the first one roam free throughout the entire song. That’s where the erase mode for the mouse comes in useful.
Now, our happy loop can roam free! And if we want to constrain it, we just need to make another blank segment where we want the loop to end, so the loop knows where to end. Wait a minute… nothing’s happening!
The segment creation tool will not work when you click on top of an existing segment, and since the loop will fill up the track all the way to the end, that means you have to create a segment in another track and move it to the first. Since it’s blank, you don’t have to worry about its instrument type (and if the name is that important, you can change the label in the segment parameters).
Splitting a Segment by Pitch
Sometimes you play a segment and decide that you can apply different instruments to different levels in the playback… perhaps a lead guitar can handle the highs, and you want to hear a bass guitar play the lows. But you’ve already put all of this to the main track, and it would require a lot of work to separate them, right?
Now, as an example, I’ve created one of my killer (I’m serous… don’t listen to it) segments with notes in the 4-6 range, and another in the 0-1 range.
Unlike the previous items, “Split by Pitch” is not a mouse mode; in order to use this feature, you do need to make sure that the desired segment is selected. Once you click the button, a window appears, asking you where you want the notes to be split.
By default, C3 (middle C) is chosen, since it marks the traditional split between the treble and bass clefs. When you click OK, two segments are generated from the one; all notes higher than this note will be placed in the higher segment, and lower notes will be placed in the lower segment. Any non-note events (such as sustain, chorus, and reverb) will be duplicated for both segments by default, although you can turn this feature off if you choose. (I don’t know about the ranging up and down option… sorry.)
In the end, the segments will be placed one above the other on the same track:
Combining Separate Segments
Suppose that was not what we wanted after all? Or maybe we made two separate tracks, but now we want them to be handled together? Well, you can merge separate segments by selecting both segments (they must be on the same track) and using the “Join” control.
No windows will appear, but where two tracks were, one single track will be present, such as this re-merge of the two pltch-splitted segments performed above.
Quantization in MIDI is the process of aligning notes to the specific beat. Computers keep perfect time. However, if notes are input using a more analog style, such as playing a keyboard or pounding a drum pad, human timing variance will come into play; notes will start too soon, or too late, and possible last just a tiny bit longer than they should.
This can be a good thing in a lead or a solo, but it can cause a problem for backing tracks such as percussion and harmony, against which the variance of the lead determines the energy and/or emotion of the tune. So, we generally want to try and quantize the timing of the background so it can set the pace and allow the leads to do their thing.
Rosegarden’s quantization feature actually depends on how the music is presented in musical score; in essence, it will attempt to apply notation rules to the music, applying features like legato, tuplets, and other advanced musical features against a grid determined by the base note (all the way from whole to 64th-note triplets).
Pictured above, grid quantization simply tries to adjust notes to a more precise location on the grid, determined by the note size (indicated by the base grid unit, described below). This allows alignment of the rhythm to make the backing tracks more precise, although it helps to look them over after quantization to ensure that notes were not aligned with the wrong note.
The first three options determines the grid used by the quantization for alignment.
The “base grid unit” determines the size of the grid unit (the actual space between one edge of a note and the next for quantization). The size can stretch from a whole note all the way to a 64th triplet note.
“Swing” allows for the inclusion of a swing rhythm. How close to the full third you get is determined by the percentage of this value. I’ll leave the details up to those swing artists among you, as I’m getting a headache just trying to understand the rhythm. Different strokes, I suppose.
“Iterative amount” helps keep some organic life in the quantized result. By default, this is set to “full quantize,” which will line the note up to the nearest boundary. By reducing the percentage, the ends are adjusted closer to the boundary by the identified percentage.
By default, the quantization engine will only adjust the note’s beginning; the ending of the note, while moved, will not also be quantized. However, if “quantize duration as well as start times” is set, the note’s end will be quantized once the note’s beginning has completed its grid placement.
The next four options determine the final touches applied after the quantization; these are for the actual musical score printout generated by the program, but the latter two can in affect the sound of the music generated.
“Re-beam” adjusts the beams of small notes (the lines tying eighth notes or smaller) for appropriateness. In this way, eighth and smaller notes that are close together are printed with the beam for clarity.
“Add articulations” attempts to apply articulations based on the notes, such as staccato (hard, fast notes), slurs (played without separation), and tenuto (extra strength or duration).
“Tie notes at barlines etc” is pretty obvious; if a note lasts beyond the end of a bar, a tie is placed to show that it needs to be continued without a break. This can make a major change to what was once two notes to instead become one long-held note.
“Split and Tie Overlapping Chords” is used with counterpoint, to mark chords and the individual notes of a counterpoint. There’s more to this, I’m sure, but unfortunately, like the swing beat, it’s just a little deeper into music theory than I’m capable of following. Like the previous option, this can affect the musical output dramatically if an incorrect interpretation is reached by the quantizing engine.
This starts as a grid quantizer listed above, but instead of the notes being treated as individual, they are instead treated as continuing; the quantizer treats each note as a continuation of the previous note, rather than a note with a start and a stop. By default, this means that the quantizer will adjust both start and end notes to the grid, so that they can be tied.
There is only one grid parameter for this quantizer, the actual grid base unit.
The heuristic is the most complex of the quantizers, because it goes all out in interpreting the music to closely match the standard notation. It will analyze a segment, make its best attempt at deciphering just what is musically going on, and then adjust starts, ends, and other parameters to make the music match what it thinks is musically going on. When it works, it’s beautiful. When it doesn’t… well, good thing you can undo in this program.
The only notation parameter that this quantizer shares with the other two quantizers is the grid base unit size.
“Complexity” is the most important item to have set right; it should match the actual complexity of your segment, so the quantizer can make the correct assumptions. Trial and error is definitely going to be needed until you get the feel for what the program considers “complex.”
If you have any kind of tuplets (multiple notes in a beat) in your music, the “Tuplet” option will let the quantizer know this so that it can accurately determine the position of the tuplets in the music.
For the “Permit Counterpoint” option, I have not seen any documentation yet. I will assume for now that this will be treated by the Heuristic quantization engine as a way to determine if it will split and tie overlapping chords.
Now, we come to the important part of the process… actually putting notes into the tracks. For the examples, I’ll use a MIDI I have with the Star Wars main theme, since I don’t want to risk killing anyone trying to read the music I create. Or drive them mad, for that matter.
There are several tools for putting music together, and we will go over the primary ones, starting with the most popular means for putting music in a sequencer since the computer started performing musical tasks.
Matrix Editor (Piano Roll)
The matrix view is common in DAWs and MIDI Sequencers the world round. The principle is that it is supposed to resemble the holes punched in the rolled tapes used in the player pianos to allow the keys to automattically depress, playing the note punched in the roll. This is why this is also known as the “Piano Roll.” To emphasize the point, the notes are identified on the left side in the form of piano keys.
You probably notice a lot of familiar items in the toolbars. In an attempt to be as intuitive as possible, the icons that were used to edit segments in the main window are used here to edit individual notes. The create, delete, move, resize, and select options work identically here as they do there… as do the save, cut, copy, paste, undo, and redo buttons, as well as the transport controls.
However, there are a few new items here, so let’s look them over.
On a piano, the sound of a note will change slightly as you press harder on the key. The volume obviously increases, but also, the vibrations on the string will be harder, giving a sharper edge to the sound of the vibrating string. If you blow harder on a trumpet or clarinet, the sound of the note likewise will change along with the volume.
On the matrix, a default note is orange in color, indicating a default strength. However, you can increase the strength of the note to indicate force, or decrease it to indicate a light, gentle sound.
In order to make these changes, you need to use the “Note Velocity” mouse mode. In this mode, you move the mouse over the note in question, and then drag it up or down in order to increase or decrease its intensity. An increase will give the note a hotter color (closer to red), while a decrease gives it a cooler color (closer to green). Red is the most intense, and green is the least intense.
Now if you’re a concert-trained pianist who can play a perfect song on one sitting, then bully for you, you’ve got a serious future (or perhaps present, if you’re already doing this for money). However, I’m certainly not very good at playing music in good time, and I certainly am not to the point where I can always get the right notes all at once. So, this next tool is very useful for entering music without having to “get it right.”
This feature requires the track to be attached to your input device, be it a drum pad or a keyboard. To do so, you need to adjust the “Recording filters” track parameters to use your keyboard (or drum machine) as the device.
Once this mode is started, you just press the first note on the keyboard. As long as it’s held down, every additional note played on the keyboard will be placed in the same beat. Once let go, the insert point goes to the next beat, and you can repeat the process. How hard you press the keys determines the strength of the note; harder presses will make it hotter than light presses.
The same applies for a drum machine.
Note Selection Filter
Sometimes, you need to edit notes fitting a certain criteria. Perhaps all the high notes need to be strengthened, or the low notes weakened. Perhaps notes beyond a certain strength need to be pulled back a bit. Maybe you want to shorten some long-duration notes a bit. For whatever reason, you don’t want to make adjustments to all the notes, but the ones you want are so dispersed into the sequence that it would take forever to select all of them, missing none, and not selecting those notes not needing the edit.
Because of this, Rosegarden can, when all the notes are selected, reduce the selection to those notes fitting a certain description. The selection filter is the tool for this use.
To begin, you first need to select the group of notes that contain all the items you wish to correct using the selection tool. Once done, you’ll have have something that looks like this:
Now that the selection has been made, click on the button that looks like a brass funnel; this is the Event Filter. This will bring up the following dialog box that will allow you to select what you want to include or exclude from your final selection.
Once done, only some of the notes are highlighted; preferably the ones you selected:
The second toolbar shows the grid size (in the above, the grid is eighth-note sized), Velocity (beats per minute), and quantize (disabled) options.
The first two are explanatory, but the third needs a little explanation. When you highlight notes in the display, and then select a quantize value, they will be aligned to that note’s grid without needing to select any extra options in the Quantize area.
MIDI Effect Rulers
Three more buttons allow you to select special rulers to appear underneath the matrix editor. These will show and allow you to adjust certain parameters of the instrument, such as velocity (volume), pitch bend, reverb, panning, and so on.
Note, however, that the values have to be present for the editing to take place; the presence of a note is sufficient for the velocity, but pitch bend, panning, etc. all are non-note MIDI values that are part of the track as separate events. Some can be added (highlight note(s), Edit->Insert Pitchbend Sequence) in the editor, while others need to be added separately, such as during recording.
Below, you can see the matrix window with the velocity ruler selected underneath. Note the way the color applies to the strength of the note in question.
Percussion Matrix Editor
When you want to set up a percussion clip (or when you double-click on a segment in a drum track), the percussion matrix editor will appear. Note that it is almost identical with the above piano roll, with two important exceptions: Instead of the piano keys, you see a list of drums and the keys they are assigned to, and the “notes” are always the same size; to get things like a drumroll, you have to place a number of percussion “notes” immediately next to one another.
This is the other primary way to enter music. Because of the nature of musical notation, this entry method is more exacting, since things like vibrato, slurring, staccato, and velocity control are actually identified by various forms of notation, and automatically handled correctly.
You will notice that there are a lot more toolbars here than there are in any of the other entry views; you have toolbars for accidentals, clefs, notes and rests (switchable, both dotted and not), symbols, and more, in addition to the standard toolbars we’ve discussed so far. If you can read music, these are going to be no mystery to you.
Sheet Text Entry
The first item is the text entry item, marked as a letter “T”. This is for the printed score; the text is placed underneath the line in the classic location for lyrics; it is assumed that the note above the lyric is the sung note for the lyric. To use this, you need to enable the text entry mouse mode, and then click on the measure that the lyric will be applied to. When completed, you end up with something like this:
The next item is one that would be useful in guitar sheet music (although not so much for actually computer playing); the chord marker. This can be especially useful for having the sheet music handy when beginning to record guitar tracks (which even the best samplers simply cannot duplicate with any sincerity).
On the left top is the list of chord keys you have to choose from. In the center are the chord modifiers. On the right, you will see the fingerings needed to accomplish those chords. On the bottom-left, the selected fingering, and in the center, a selector for the chord complexity required, and the options to create, delete, and edit custom fingerings.
Once you have the desired chord selected from the list on the right, you click OK, and your chord is now just above the measure.
Audio File Manager
The first trick is to have pre-recorded vocals and/or loops available, in which case, they can be inserted into the song. The first thing you need to be sure of, though, is that these loops are available for insertion. This is where the audio file manager comes in.
The audio file manager is the project database, where the wav files used in the project are store in preparation to placement in the appropriate tracks. Once they are stored in the file manager, you can just drag them out to the track where they can be integrated into the mix.
Of course, some of the features will not be available for this track; it cannot be edited in the rolls or notation editor, and it cannot be quantized. It can, however, be looped, and it is recommended to have a blank recording to serve as an end for the loop.
Rosegarden has some additional editors available for more advanced users. Some are more useful than others, but if you understand the featureset and want to use it, they’re there for you. I’m not going to go into depth with these; if you know what they’re for, then you don’t really need any more depth, if you don’t know, then I may confuse you further, since my understanding is a little sketchy at the moment.
Event List Editor
If you understand MIDI at its most fundamental, or you want to see what MIDI looks like, then this is your editor. This is the list of MIDI events performed by the segment, from first to last, in order. You’ll notice that most of the events are note events, including pitch, velocity, and duration. This is what is sent to the sampler and/or synth to allow them to know when to make or stop notes in which instruments.
Next is the pitch tracker. Almost identical with the notation editor, but containing a sort of realtime visualization at the bottom that shows the pitch of the currently-playing note, as well as tuning information. How they work, or are used, however, is beyond me.
Now that you know what there is in general, it’s time to begin making some music with the software.
Start and Connect a MIDI Synthesizer
The first thing we want to do is connect the music rendering tool. Since I have not really written an article on any of the synths or samplers available in Linux (I’ve only just started learning how to use them), I will take some time now just for a basic setup.
For this example, we’ll use Qsynth with the RealFont soundfont mentioned earlier. Here are the quick setup instructions to get QSynth from zero to music in as short a time as possible. Before you begin, make sure Jack is running.
In the MIDI tab, make sure the options match what’s above. In the “Engine Name” box, I just typed in what I wanted the name of this MIDI device to look like in Rosegarden’s selector. Set the “Alsa Sequencer Client ID” to match.
Finally, in the Audio tab, you need to make sure that the program is set up to use Jack. In my case, since I use Jack at 48KHz, and it uses floating point resolution (internally), it seems like the best settings, and so far, I’ve yet to have a complaint.
The Jack Client Name ID can be manually entered; this is the name you will see in the connection graph. I use qs_ before my four tabs; two for RealFont (two in case I want one or two instruments to be filtered differently from the rest), one for drums (the Real Acoustic Drums Extra soundfont from the HammerSound library), and one meant for a clean sound (no effect filtering), which I name “Bypass.”
Once the software is set up to work with Jack, we click “OK,” and let the program restart the synth to make sure it is using the correct settings.
Now, we are done with the Qsynth setup, let’s move onto the Jack connection graph. You’ll notice that I added a Jack-Rack to the chain; I’ll explain it shortly. For now, just start a copy of Jack-Rack and connect your RealFont-enabled QSynth output to it, as well as your running copy of RoseGarden (the master bus).
Create a Track or Two
This is pretty obvious. At this point, you’ll notice that you can hear everything you’re doing, so making and editing tracks will be pretty simple. Just use the pencil button to create a blank segment, use the two-arrowed control to stretch out the segment to the desired length, then click on the selector (mouse pointer button) and double-click the segment to begin putting notes in it.
Place Notes in the Segments
Using the note-editing tools, create some notes in this segment that meets with your approval. If you are a concert pianist with excellent keyboard reflexes, you can probably do away with the need for the editor and just arm the track and begin recording off your keyboard.
Using Plugins to Make Things Sound Real
By default, MIDI instruments are dry. There isn’t a drop of reverb to them, and their included reverb isn’t very good. This is why I recommended the Jack Rack application; with this, you have much more flexible and excellent-sounding reverbs to make the instruments sound like they’re in a space with the desired acoustics.
Now, by itself, this would make a very bad reverb. However, that can be fixed by reducing the wet level (the volume of the reverb itself) to -20, while keeping the dry level (the original sounds) at its original level. Voilà! it’s not quite the same as being in a concert hall, but it is definitely closer to the mark! You might want to play some more with the settings for better results.
Hopefully, you will take all this and make great things. You certainly can’t do worse than me, so be strong, be creative, and above all, make noise! …I mean MUSIC!