For all the wonderful tools and toys for sound out there, sometimes you want to find the couple of tools that, like a great kitchen knife, can accomplish the majority of what you actually need. (And as with the kitchen knife, while it may not eliminate your desire for all those other gadgets, it’s worth some sharpening.) So it is with something like Csound, the tested-and-tried, free synthesis tool. Jim Aikin looked at the QuteCsound front end recently, which puts the power of Csound in a more friendly work environment.

Via Synthtopia, there’s also now a screencast series that covers using QuteCsound, starting with digging into presets. (Yes, that’s right – presets. And here you thought you were going to have to do a lot of coding to have any fun.)

I find two YouTube users uploading how-to screencasts: (author of the series starting at top) (start with “Where to start?”

Also worth following is Jacob Joaquin’s excellent Csound Blog, hosted on Noisepages:
and on Twitter, follow @TheCsoundBlog

It’s very early in development (“alpha”), but Jacob is already doing amazing things integrating Processing, the non-coder-friendly, artist sketchbook-style coding language, with Csound, in a new library called Csoundo. That’s an ideal combination, because you can do logic and visuals quickly in Processing, then turn to Csound for audio. This is where I imagine work in two of Csound’s most popular rivals – the object-oriented, OSC-savvy SuperCollider and visual patching, Max-descendent Pure Data – may lead, as well. Check out Jacob’s roadmap for more.

By the way, I hear some folks are having some trouble building QuteCsound on Ubuntu, so I’ll see what the issue is, and write up some instructions and send them over to Jacob for his blog.

It’s a good time for Csound and free synthesis in general. With this work accelerating, I think doing a series of absolute-beginner tutorials will be very doable soon. And there’s no reason you can’t integrate a tool like this with your favorite host of choice, from Ableton to Cubase.

  • @Peter: Thanks for the mention!

    A couple of things. Csound would have greatly benefitted from having something like QuteCsound 10 years ago. It's so nice to have an integrated editor that is easy to get up and running almost instantaneously, with working examples, and GUI support. It literally took me 6 hours to produce my first sine wave with Csound back in 1996. Andrés and his team of contributors are doing a fantastic job, and QuteCsound is getting better with each release.

    I completely agree with you about things accelerating. A few years ago, I felt like Csound was on the wane. Not so much anymore. Here's a list of some of the things that are happening besides QuteCsound and Csoundo: Blue. Cecilia. The devs are working on parallelization. One team is exploring building Csound with CMAKE, which could lead to easier compiling for all platforms. The development infrastructure is slowly becoming more modernized, for such things as reporting bugs. The manual is constantly being improved. Csound Max and PD objects, etc…

    Also, I think you meant to link to this for Csoundo's roadmap. đŸ™‚

  • It looks like QuteCSound is badly out of date in Ubuntu's repositories, from the 0.4.x branch. So yes, it's better to have a package, but that project may need a new package maintainer.

  • Also, it's possible that at this point with so many CDM readers using Ubuntu, we should just go ahead and set up a PPA, collaborate on that, and then contribute to the main repositories when things are ready… open to suggestions.

    (That's not to be anti-Fedora, etc., but there are some advantages to Ubuntu's community support in general — and with Fedora well-covered by CCRMA, Ubuntu is what's hurting.)

  • Zach

    Forgive for stepping on some toes, but it seems to me that people should be ditching the archaic Csound in favor of the newer and better SuperCollider. Equally free, open source, and arguably more powerful — and also has pretty GUI elements as well.

    …why put effort into keeping Csound around?

  • Adrian

    Regarding the Ubuntu comment, I was able to install both CSound and QuteCSound straight from the graphical Synaptic Package Manager in Ubuntu 10.04. Building from source manually is so archaic!

  • @Zach: Personal preference.

  • @Peter: Also, thanks for all the attention!

    I use Ubuntu, so it should really be a matter of getting the right packages. You need qt-dev-tools, csound-dev and libsndfile-dev. If you use the csound from the package manager you will need to build using:

    qmake CONFIG+=build64

    Since the Debian package of Csound is compiled for 64-bit precision, and QuteCsound builds 32 bit by default.

    The debian package maintainer (Felipe Sateler) is actually very commited and 0.6.0 has already hit debian, but it takes a very long time for these packages to trickle down to Ubuntu, so if someone can set a PPA, that would be grand. All the work towards building a debian package has already been done.


  • Kim

    Good Info thanks

  • It’s not either or, either. You may use different tools for different jobs. You can learn more about synthesis by learning both Csound and SuperCollider. And some of the development work has multiple benefits. For instance, I’ve been following Peter Brinkmann as he’s worked on developing the ability to make Pd run on Android. Now, that work will have relevance to Pd on other platforms *and* to bringing other languages (like Csound) to Android. So just because effort is expended in one place doesn’t mean it’s taken away from somewhere else — in fact, sometimes exactly the opposite. There is more crossover, though, when using free tools in this ilk than hopping between proprietary platforms, at least for things like synthesis development. That’s at least in my experience, and I’m a great fan of some of the proprietary platforms.

  • Roet

    arguably more powerful?

    what parameters are you using to say this? Utter nonsense. SC has an interesting sound server but SCLang is a horrible language.

    As for Csound being archaic, you should try to get more informed about things first, before issuing sweeping statements like this

  • @Roet: I love Csound, but I partially agree with Zach that Csound has some serious archaic tendencies.

  • They all have archaic tendencies. Let's face it… we're an archaic crowd. đŸ˜‰ SCLang has some beautiful features from SmallTalk.

    I think CSound, Pd, and SC all make useful starting points. The ultimate goals here go across those platforms, which is to make them friendlier in music making and interaction.

  • Jim Aikin

    @Zach — YMMV, but I took a look at SuperCollider for about two hours once, and decided I prefer Csound. Csound's coding is almost entirely procedural, while SC is super-OO. In SC, if my memory isn't playing me tricks, you instantiate your filter as an argument to the output stream device, and then within that you embed the oscillator as an argument to the filter constructor. I like object-oriented programming a lot for writing interactive fiction, but SuperCollider just wasn't my cup of tea. The logic of a Csound code block is much easier for me to follow! Plus, I have little interest in real-time processing, which is SC's specialty, from what I understand.

  • I don't think there is anything wrong with the language, It's about having the right tools. api's can be a pain though unless you can get it to work with a portable install (so that an upgrade doesn't kill everything).. Looking at processing it could use a little grunt work i.e. changing the index page that is included with the package so that all the help for the sub-packages is available from the processing menu. I am excited by the possibilties of this one.