Introducing L2Ork, World's First Linux Laptop Orchestra

Agram writes "Take a netbook, Wiimotes, Nunchuks, and hemispherical speakers (which were once IKEA salad bowls), toss it up with some Ubuntu goodness and what you get is Virginia Tech's L2Ork, the world's first Linux-based laptop orchestra. With its affordable design and support from the Linux community, L2Ork hopes to bring laptop orchestras to K-12 education and beyond. So, regardless whether you wish to hear how L2Ork might sound or to learn how to build your own Linux-based *Ork infrastructure, perhaps this is a good opportunity to reopen the age-old debate: is Linux finally ready for some serious audio work?"

How is this better than a Theramin?

[JSBiff]

Ok, if we want electronic musical instruments, wouldn't getting schools one or two Theremins [wikipedia.org] be a lot cheaper and more reliable?

I'm sure lots of schools are already struggling to fund basic music programs, and get kids 'ordinary' instruments (although, I suppose it's quite possible these Laptop Virtual instruments may be cheaper than some analog instruments). Do we really need to be complicating things and making them more expensive, and for what benefit? If the goal is to educate kids about music, I don't think you need some gee-whiz laptop-and-wiimote based 'virtual instruments' to introduce more points of failure? Aren't broken reeds and strings enough for kids to worry about?

Now get off my lawn. *grin* (Seriously, I hate to be a cranky old geezer - I mean, I'm only 31, but even after scanning the article, I'm not really sure I see any advantage to this)?

Re:When I tried

[Cheile]

A general rule of thumb is that latency and throughput are opposing interests. For the majority of users and uses occasional latency is a good tradeoff for higher overall throughput and performance. For realtime applications low latency is a good tradeoff for lower overall performance. Linux can be tuned in many ways for many different use cases and may not suit your needs right out of the box.

Re:How is this better than a Theramin?

[PopeRatzo]

Do we really need to be complicating things and making them more expensive, and for what benefit?

I've designed and built the computer music lab at a prestigious university, and our experience is that technology is actually a big money saver for a music program.

Plus, it's a boon for the students. If you're studying composition or arranging, do you think it's easier to go out and find an orchestra or chamber group to try out your work or fire up a computer with a (virtual) rack of samplers? There's no substitute for hearing your composition played in real time by a real (if sampled) orchestra. Plus, a fair number of students will seek work in the music industry, and there they will be expected to be familiar with the technology of producing music. Now, if you're talking about a grammar school music program, it might be different, but I'm guessing the same efficiencies apply there. Of course, if you want to learn the violin, you need a violin. But if you want to learn music, a MIDI keyboard and PC can be a great way to go.

If it's cheaper for film composers and music production facilities to use technology, why wouldn't it be for schools' music programs?

Re:When I tried

[sowth]

In some ways this is a good point. Distros intended to have desktop users should probably include options to use low latency kernels. I don't see how it would be too difficult to include a few kernels compiled for different uses.

• High latency (100Hz), no preemption for server / intensive cpu jobs.
• 300 Hz with Voluntary Kernel Preemption for gamer / media playing. I noticed videos were not jumpy at all with this setting. The default would sometimes cause mplayer to stall for a several frames, then play them all at once. I didn't think about this or consciously notice it on video before I did this, but I really can tell now.
• One with the lowest latency setting and complete preemption, etc for more time critical and reaction sensitive work such as music and running robots and such.

All a distro maintainer needs to do is compile them with a few minor option adjustments each time, then either put each kernel in a separate package or just install them all and config the bootloader to run the correct one. Not a huge amount of work, relatively.

This seems to be a decent article about preemption [affenbande.org].

As for an app which takes in MIDI and plays sounds, would Zynaddsubfx [sourceforge.net] be it? I seem to remember seeing that feature when playing with it, though I have no MIDI devices, so I couldn't use it or say if it worked...

Too Steep a Curve

[b4upoo]

So far Linux does have musical capabilities but not enough and not easy enough either.
            For example converting a treble clef trumpet part to a bass clef trombone part should be a no brainer and printing it out should be a breeze as well. Further converting a melody played through the system into a score should also be simple.
            Having to use different programs to print a score, convert a score and play a score will not attract many musicians.
            So far one of the very, very few excuses for running Windows is the production of music by musicians. Online DJs may fare better as they may alter music but do not usually create it.

Re:Ardour Sucks

[makohund]

Huh? Change the gain of what? Why?

Is a firetruck responsible for maintaining proper water pressure at the fire hydrant it is hooked up to? No. That's not its job.

It's a DAW, not a preamp. If you have to change the "gain" of something in Ardour, you're already doing it wrong. If it could... all it'd be doing is allowing you to continue to do it wrong.

If you are feeding it a weak signal, fix the damn signal.
If you are feeding it an overly strong signal, fix the damn signal.

"Fix the damn signal" means adjust your outboard equipment (preamp, mixer, noise generator, or whatever) and/or your soundcard's mixer to get a proper signal happening to begin with. Don't feed your DAW a crappy signal, unless you WANT to record a crappy signal. Get it right as early in the chain as possible, because once you are in the DAW, it's too late. You're already past the AD conversion... any gain is nothing more than a little math to make it louder, which can be done at any time. It'll never make it "better".

If it is too late fix the signal, and you are dealing with pre-recorded audio that is too weak (or outboard equipment that isn't up to snuff), fix it right (permanently) with an audio editor. That way you aren't wasting CPU doing that same digital math over and over again every time you hit play.

And if you absolutely, positively can't do it the right way, and insist on changing your gain in the DAW itself... yes, you can do it with a plugin. (Not exactly hard to find... I found no less than three already installed, and I've never installed any but the most common/basic plugin packages. And that's on a plain vanilla Ubuntu machine, not a studio machine with lots of goodies on it.) In the end, doing this with a plugin is no different than having it built in. It is better, because it gets it off the damn screen for most people, who will not and should not need it 99% of the time.

Sounds like a lack of either minimal effort or minimal understanding of recording in general. Go get a Mac and use Garageband. It was made for you.

(And hey, I'm not knocking it. I do have a Macbook, and have used GB for sketching around with. It's not bad. Inability to export midi is probably the only really major drawback.)

If you wanna knock Ardour, say something like "it lacks MIDI functionality found in most other DAWs", which would be true. Until v3 comes out, at least.