KXStudio – no looking back

Recently I posted how impressed I was with KXStudio and how it might find it’s way on my podcasting desktop.  Well, that moment has indeed arrived, KXStudio is now on my desktop and I’ve got no regrets – KX is the future of Linux audio and the future is now.

From what I can see, Falktx – the maintainer for the distro is working very hard testing and packaging the latest apps – Jack2, VST support,  Ladish and Ardour3, Mumble with Jack support are just a few examples.  He is also doing things like patching older applications here and there – an example being the gNinjam client which no longer compiled under Ubuntu Lucid.  I expect to see other projects such as Ubuntu Studio benefit from the experimental playground that is KXStudio.

There have been a few changes in KXStudio recently and you can read all about them in the atom feed here: http://feeds.launchpad.net/kxstudio/announcements.atom

Chiefly, the most recent change in direction to run audio from Alsa to Jack vs Alsa to PulseAudio to Jack as detailed here:


While it does appear that PulseAudio bridge is still running, it is not the preferred way to go unless you have to.

So, if Alsa is layer 1, and PulseAudio sits upon layer 2, then Jack is either on layer 2 (no PulseAudio – just Alsa then Jack) or layer 3 which would be Alsa -> PulseAudio -> Jack.  The more layers involved or the more complexity involved.

For those that don’t understand what all this means.  Alsa is one of the main sound servers running under Linux to handle audio, and if you have an Alsa supported sound card you could get sound with just Alsa.  Alsa however doesn’t handle situations where more then 1 application wants the audio card at the same time, therefore workarounds such as PulseAudio which is a layer above Alsa allows for the sound card to be accessed by many applications at the same time and offers volumes controls from each application in a PulseAudio mixer as well as the ability to stream audio on the network.

PulseAudio has come under fire (unjustly or justly – you decide) for not being a very good resolution to the common challenges that have plagued Linux audio for years.  Be as it may, PulseAudio under KXStudio is stable for me under the Generic and the Lowlatency kernel, so I can’t complain at all – it runs stable.

Which brings me to the part of the post I am most excited to announce.

As a co-host of The Open Source Musician Podcast separated from the host by geography, we had relied on a closed source application to handle our phone call to one another.   This was always a sore spot because all the other software we use is Open Source, however this was one last hurdle we couldn’t get over.  Recently this all changed and we did our first fully open broadcast Friday using the GPL’d VOIP client called Mumble.  The Mumble team has added experimental integration with JACK (Jack Audio Connection Kit).  This completely solves our VOIP issues.  The quality is excellent and Dan and myself can record a good track of local audio and each other’s VOIP call into its own track in the Ardour DAW software, we stream this all out on the internets and it all takes place with just 1 PC on each end – no double-ender, no extra hardware – it’s all in the box and its pure gold.

So life is peachy.  Our audio woes are virtually non-existant.  A big thanks goes to Falktx of KXStudio for packaging this experimental Mumble version,  and of course to the Mumble developers themselves for making this all possible!


  • Pingback: Bariman's Blog » Blog Archive » Installing KXStudio . . .

  • Congratulations on your first all-open episode! Your efforts are paving the way for generations of podcasters after you! Now if we could only get the generations of podcasters before you (yeah, I’m talking about the TLLTS guys) to use your setup. 🙂

  • Hahah no way! TLLTS forever… the shows charm would change I think if the audio was solid every episode. 😉 I have been thinking I need to consolidate how we do the show it into a proper blog post – it’s coming – I promise. 🙂

    Thanks for taking the time to comment.

  • Pingback: Current Projects, 10/10/2010 « Manysounds.com

  • I’ve used KXstudio for over a year and has been a very solid OS and set of apps. The integration is superb and especially the way audio is done using Jack2 to manage ALSA and Pulse. Very clever and functional. Once configured, I never need to switch drivers, etc. I have a newer HP 64-bit tablet PC and an older 32-bit Toshiba with the respective versions on KXstudio installed, both running very well.

    My only beef, and this is with all Linix distros and not exclusive to KX, is that updates can ruin configurations and getting things all back to full operation after upgrades can be daunting and even cause for reinstallation, once I make that call that I’ve wasted enough time trying. There is also an issue with blocked packages, where I’m offered an update but it won’t install for “legal” reasons, etc. Again, these are Linux and not KX issues, which I have with Ubuntu, Suse, etc. too. Often it’s just easier and more efficient to download the latest installation DVD iso, and do the install from scratch. I’ve even gone to maintaining a separate data partition to make this easier when necessary.

    Bottom line, for me, KSstudio is the best audio-supporting Linux OS I’ve found in about 5 years of Linux use. Kudos to Ubuntu Studio for effort, but Falk has done great things with KX and US should pay attention and adopt some of Falk’s better concepts where it makes sense. I see a lot of “mine is bettering” in the various Linux camps, so I hope Ubuntu can swallow some pride and work together for a better user experience, rather than succumbing to the popular stubborn pride driven oneupmanship that a lot of camps exhibit when something better comes along. Uforums moderators did let me post a comparison of the two, in which I tried to be objective. To me, the real test is whether a distro’s devs actually read and consider what us mortals post there. I would qualify myself as only a power user at best, definitely not a programmer, and therefore mortal. But if Linux is to ever beat Big Brothers Inc in user satisfaction, it has to be manageable by us mortals or it will remain “just another elitist OS”. This is what is so cool about Android (I love Android) and even iPod/Pad/Phone (oops, did I just praise Apple?), that mortals can use such software without any real technical ability because it just plain works (for the most part). The elegant simplicity is the key.

    Falk’s work with KXstudio is helping to make Linux, and especially Linux audio, mortal friendly! Kudos Falk! It’s not yet click-n-go, but it’s a huge improvement and very much a move in the right direction for Linux in general.

  • Steve,
    Very well put. I feel onside with a lot of your comments.
    KX is the best we have right now. I did get breakage here and there (less now on 10.04.3) and it shouldn’t happen. I will tell you that the breakage was almost always fixed by using the cmdline tools of: sudo apt-get update && sudo apt-get dist-upgrade. And when breakage occured: sudo apt-get install -f … Sadly the kde updater really does suck and you have to keep the distro managed via the command line tools for things to go smooth.

    Now I do get breakage from time to time on my Windows PC at the office, sometimes catastrophic, but I do agree, there ought to be a better way to protect the Linux from this plight.

    One thing I think could be considered is either an update safety level like in Linux Mint could be adopted – you set your level of update – updates are rated in terms of risk to stability and you choose your risk level. Core packages crucial to the operation of the distro are pinned so that they aren’t updated unless its from the KX repo – I think he does this to some degree already. But the big killer for me is the kernel updates – they always bust the wifi on my laptop – that’s a big annoyance for me (broadcom). Perhaps KX should not update the kernel unless its a bad security breach. Slackware is quite pragmatic in its kernel updates, they don’t happen unless its a total showstopper.

    Right now falktx is actually working on getting the KX experience ported over to non KDE environments. I think this when done is going to be a huge gift to the community that other distros can pickup and this will broaden the KX experience to all the GNOME users which should be an interesting development. Someone needs to hire this guy!! Canonical – get on it!.

  • I finally got the new version 10.10 its running smoothly on my asus 701sd im only running on 8 gigs and no problems I like all the changes but I do have some issues running pulse audio ..think its some simple driver issues with my sound cards … but otherwise all my random spurts here and there on forums and through anonymous emails the net-book studio is now here and KX studio made it simple

  • Sorry I missed your comment Mike. Wow – so you have it running on the 701! Are you running it on KDE or a minimal window manager like Fluxbox? That’s quite a feet since that particular eee model I think is a Celeron based processor. What are your favorite audio apps to run and what use case are you fulfilling with the Netbook? I know someone who was using it to stream their gigs.

  • hey! tertl3 here. I just want to give a +1 to KXstudio as it has become my main distro of choice also. 🙂

  • I too have taken the plunge into kxstudio on a donated laptop a family member didn’t have a use for. I personally feel many a musician will go linux. For me the documentation, was the biggest hurdle. I tried every program before discovering cadence. I still have an audio issue with the Toshiba laptop mentioned earlier. So, due to technical issues I use Propellerhead Reason in combination with Ableton Live. For VST, non-free modules like Trilian, BFD2 or any of Native Instruments you can’t go wrong. The exciting thing about having KXstudio is the amount of plug-ins and the Windows VST problems are becoming as thing of the past.

  • I think you are right – the documentation in a central location is lacking, and hardware can be an issue. On the flip side – it’s gotten to the point where if you have the right hardware and know-how, linux is a viable recording platform.

    I think that’s why I no longer podcast about it. I know what tools to use when and those tools are robust and mature. Now it’s an issue of finding/making time to use them. 😉


  • Upgrading was pretty easy, and PulseAudio works a lttile better still have to dork around a lot to get multiple sound sources playing at once. At least Civ4 works even though I’ve set it up to use OSS as the sound, and I can’t have Flash player playing songs through Pandora in the background. Oh well.

  • Sandeep – agreed if you have 2 separate audio interfaces it an get funky. I use Alsa_In for that. When I plug in my Fender Mustang amplifier via USB I have my main soundard on the laptop (built in) as the audio interface, then I have alsa_in run in a command line window to get the fender amp going and I route that using falktx’s patch bay utility.

    I think it’s pretty exciting that falktx has moved his repos to debian stable/testing as well. Things were working solid and reliable for me on that platform. I still keep that going in a virtual machine to provide feedback if I have any, but recently I moved to Arch Linux since I do a lot less work with digital media these days and in reality I use only Ardour and Hydrogen for the most part.

Leave a Reply