-
Notifications
You must be signed in to change notification settings - Fork 12
Sound SoundObject
Author: Steven Yi
One of the more curious soundObjects in blue is the Sound SoundObject, which allows for the direct writing of instruments in the score. It's origin came a long time ago now when I was first beginning to create more soundObjects for blue. I was thinking of what exactly should be possible for a soundObject, thinking that if someone were to make a soundObject, they should be able to also embed instruments and f-tables that they'll know will always be generated and available. The case which I was thinking about at the time were non-generic soundObjects like a drum machine, where the soundObject would not only have a GUI to develop pattern tracks for different drum sounds, but also would be able to furnish the instruments and f-tables needed to generate those sounds, the idea being that one could release a soundObject that a user could use "straight out of the box", no instrument writing or tables required. That design decision eventually lead me to the conception of the Sound SoundObject, which uses those mechanisms that were put in place for all soundObjects. Theoretically I was very fascinated with the possibility of writing instruments within the main scoring area, mixing note blocks with pure sound blocks. It became possible to really compose with the sound in a very direct manner; saying things like "I want a sine wave here with this envelope, then a triangle wave here, and on top of that i want a processed sample sound to come in here" now had a direct translation. You didn't need to write the instrument, then write a note for it, and then have to work with both of them as separate entities. You could just add a Sound SoundObjectblock on the timeline, write your sound using standard orc code, and move it around in time on the timeline. After having thought through that possibility to express my musical goals by using the Sound SoundObject, I found that it opened up a lot of how I saw blue's timeline as well as how I thought about ways to work with blue.
As a Csound user or general electronic musician, it might seem strange to think of directly writing instruments in a score, especially in context to the existing music tools and musical models that you've probably worked with. However, by having the Sound SoundObject, I've found that interesting possibilities have opened up, and since it's creation and inclusion into blue it has been used and has played a part of just about every piece that I've worked on.
Note: It's not that it's impossible to implement this any other way. Really, it's just a single instrument with a single note, but, that's a technical issue, an implementation issue. The interesting thing is that when working with it, it's really a different thing altogether conceptually. It's the sound soundObject in the context of the timeline that makes its usage interesting and useful(well, for me at least!).
The following sections will go over how to use Sound SoundObject, what happens when it gets processed by blue, as well as some usage scenarios and patterns that have arose while using it in my own work.
First, insert the Sound SoundObject as you would any soundObject by rt-clicking (for Mac users, hold down the apple key and click) on the main timeline of the score area and chose "Add New Sound" from the popup menu.
A Sound SoundObject will be inserted wherever you clicked on the timeline. You'll notice that this soundObject acts and behaves like any other soundObject: you can move it around in time, drag to change the duration, change it's name in the soundObject property dialog, and click on it to edit it. You won't be able to add any noteProcessors to it, however, but that will be explained more in detail in the "What happens when it gets processed" section.
If you click on the soundObject to pull up its editor, you'll see a text box with a default message "insert instrument text here". In the editor is where you write your instrument definition, but without writing any "instr 11" or "endin": the number of the instrument and the correct formatting of the instrument is handled by the soundObject. For a simple example, you might try:
aout oscili 30000, 440, 1
outs aout, aout
where the 1 at the end of the oscili line is an ftable numbered 1, defined in the tables editor under the tables tab. (For our example, let's go ahead and definte the ftable as "f1 0 65536 10 1", which is a sin wave). At this point you might want to try listening your work file with the play button (assuming you've set up the command line under project properties to a command line that will output to speaker; for more information on this, please consult the blue user's manual). What you should hear is that where you've put your Sound SoundObject on the timeline (i.e. starts at .5 seconds) should play the sound you've written, in this example case, a sine wave at 440hz at 30000 amp, playing at equal volume out both left and right channels.
And that's pretty much it! From here you might want to try copying the block and pasting it a few times, changing some parameters, embellishing your instrument definitions, etc. Also, you might want to try mixing this with other soundObject, i.e. write some instruments in the orchestra manager, then write some genericScore soundObjects to play those instruments, as well as use some Sound SoundObjects directly on the timeline.
If your confused as to what's going on, it'll probably help to know exactly what happens when blue processes Sound SoundObjects. blue, whenever it goes to create a .CSD file to use with a commandline or to generatoe out to a file, has different stages of its compilation. The relevant parts to know here are that all instruments from the orchestra manager are first generated, but not yet put into the .CSD. Next, all soundObjects are called to generate any instruments they might have. This is where the Sound SoundObject would generate an instrument from your text input. The generated instruments from Sound SoundObjects at this point are assigned an instrument number. After all instruments are generated from soundObjects, all score text is then generated. At this point, the instrument number assigned in the earlier pass is now used by the Sound SoundObjectto generate a note for your instrument. The note generated by the Sound SoundObjectconsists of only three p-fields: the instrument number, the start of the soundObject, and the duration. No other p-fields are generated (so your instrument should not use any other p-fields).
For example, let's say you have a Sound SoundObject with a start time at 0.5 seconds and a duration of 2 seconds.When blue goes to get it's instrument, let's say it is assigned instrument number 2. The generated note will be:
i2 0.5 2
Perhaps the best way to see it is to do the simple example from the "How to use the Sound SoundObject" section, generate a CSD file (from the Project menu, select "Generate CSD to file"), and inspect what got generated. Comparing the soundObject's representation on the timeline as a sound and seeing how it got generated out might explain things better.
(NOTE: because the Sound SoundObject only write out the three p-fields, using a noteProcessor really doesn't have any purpose, which is why the sound soundObject does not support noteProcessors)
Ultimately in the lowest level implementation, there is a separation of a note as well as an instrument, but within blue, that separation is hidden from the user. From the user's point of view, all they have to do is write their sounds on the timeline, and they don't have to worry about numbering the instruments or creating notes for that instrument.
Most of the time I've found myself using the Sound SoundObject at the start of a project when I'm creating new sounds for a piece. I find it's good a prototyping tool, initially working with sounds on the scoreTimeCanvas (this is the name of java object that is the main score timeline area). Usually, when I write instruments in the Sound SoundObject, I define all of the i-time variables in a way that will facilitate easy conversion to full-fledged instruments should I later want to do so. This is a general instrument writing pattern of mine that I would do anyways even before I had blue to use. For example, I might be designing an sound on the timeline with the something like the following at the top of the text:
ipch = cpspch(8.02)
iamp = ampdb(80)
ispace = .2
Later, when I get to a point after sketching out some sounds and finding I like how things are beginning to flow in the piece, I find that i usually want to start working with the sounds then as instruments. At this point, I normally convert the Sound SoundObject to a genericScore object (done by rt-clicking on the soundObject and picking "Convert to Generic Score" from the popup menu), which automatically takes the the instrument from the soundObject and adds it to the orchestra under the orchestra manager, and also leaves me with a single three p-field note, which also shows me what instrument number the instrument was assigned. After that I'll go to the orchestra manager and edit the instrument to now take in more p-fields, changing the top text to something like:
ipch = cpspch(p4)
iamp = ampdb(p5)
ispace = p6
I usually find myself making two or three different sounds, then copying a bunch of them and changing a few parameters to try out things in time, then converting them into instruments. It's a nice separation to have the ability to work just with sounds at the start of a piece for me, as really, that's what I'm concerned with at the beginning of a piece, finding the sounds and initially sculpting the sound space that the piece will take on. After I find what I'm looking for, it's easy for me to convert all the sounds into instruments and then proceed from there.
Sometimes when you're wanting to just to build a single sound or texture, maybe to use as a sound effect in project, you might not really be thinking in terms of scores, notes, and instruments but rather in terms of sounds in time. In situations like this, the Sound SoundObject would be the first thing I would use, and might really be the only soundObject I would use. Notes, as a concept, somtimes really don't play a part of the musical model for a piece. It's not that you have all these instruments being played everywhere, but rather you have sounds going on here and there. It might seem like I'm being a little to theoretical here, but I really think it does play a part in the work process.
In the situations I've been in when I've been asked to make a sound for a friend's website or game, I've found it nice to fire up blue, set the timeline to a really close-up zoom on time, and just work from there to craft a sound. I would add a Sound SoundObject here and there, maybe use some global variables so I can make Sound SoundObjects that might just function as an lfo or other control instrument, moving things around just slightly around in time to sculpt the sound. An oscillator here, maybe blending it into an fm sound, throw in a noise generator with some formants and a notch filter sweep...
Sometimes I find myself just making sounds with blue and Csound. It might be because I'm just curious to try something out, I might be working on really getting to know a synthesis technique, trying to learn how to express a sound in my mind or maybe to better train my imagination to know what a sound will really sound like when I write it down, etc. Sometimes its just that I want to try out some new opcodes I haven't really ever used.
It's times like this when I find myself just using the Sound SoundObject, as I'm not interested in the note-instrument paradigm, it's the furthest thing from my mind. I'm focused on achieving a sound, or on experimenting to see what is the sound of instrument code I've just written. And I want the flexibility to add more sounds on the timeline: I don't want to break my concentration to go and think about numbering instruments, writing notes, moving the note around in time. I want to see and work with it all of it in one place., as that's what's going on in my mind.
It's also great practice too, just writing alot of instruments. I'd imagine that the Sound SoundObject would be a useful tool for a person new Csound, as it allows just working with instrument code. (Note: you would still need to know the basics of how Csound works, what is a CSD, and understand how things in blue map the different parts if itself to the different parts of a CSD file).
Thanks for reading the tutorial! I hope this tutorial has helped to show how to use Sound SoundObject in blue, as well as helped show some ways in which you might want to use it.
If you have any comments, suggestions for improving this tutorial, or questions, please feel free to email me at stevenyi@gmail.com.
Thanks and good luck!