Slashdot Mirror
Ask Slashdot: Linux and the Home Recording Studio?
wjcofkc writes: Somewhere between IT jobs I found myself spending 2 1/2 years employed pretty deeply in the local music industry. It was a fantastic experience. Left and right I saw people using very expensive proprietary software. I never saw anything that a similar Linux counterpart, or a suite of Open Source counterparts could not do. Needless to say, I preached the good word. Unfortunately, I never exploited any opportunities to provide a demo. One thing concerned me. If you have a full DAW setup, it's not just software; there is always some sort of hardware interface of varying complexity involved and playing through an amp into a microphone connected to a computer is not an acceptable way to record. I recently purchased a Lexicon Alpha 2-Channel Desktop Recording Studio interface based on vague mentions that it might work with Linux. After plugging it in for the first time, I fired up Audacity and Ardour. The device was available to select as an interface with zero configuration and it works perfectly. My question to the music geeks among us: what is your take on the state of Open Source pro audio software? And what successes and failures have you had with studio hardware?
'Recording Studio' -> Get a Mac. Seriously. CoreAudio is probably *the* best audio subsystem on the planet, currently, Proprietary or Open Source.
I tried Ardour with my FireWire mixer (32ch in & out). Even with a realtime kernel, JACK would instantly eat 100% of a CPU core when connecting to the mixer, and I'd get dropouts. You have to manually tweak buffer sizes & sample rates to target a specific latency. Not to mention the entire JACK core would crash mid-recording sometimes. Ardour itself is amazing for an open-source DAW, but it's hobbled by broken subsystems as dependencies.
PulseAudio is like WDM on Windows - you're going to get absolute crap for latency, next to no control over sample rates and buffers unless you start digging through the Pulse configs, restarting the service yourself... it's meant as a 'consumer' audio playback engine, at *most* you'll want to use its recording side for, say, voice chat, livestreaming with Open Broadcaster Suite, that sort of thing. It also has a bad habit of assuming 'multichannel' means 'surround'.
The only other system that you can get by with would be ASIO on Windows, but then you're dealing with the typical Windows issues, hoping your mixer/interface manufacturer has created Windows drivers for your system, those drivers actually work and don't crash, etc. Your interface is a few years old? Only has drivers for WinXP/Vista? Good luck!
This is *the* one big place (video editing being the other) where Macs are still King. CoreAudio is literally plug-and-play, and can handle sample rate conversions, clock sync master/slave settings, etc. (Audio MIDI Preferences.app for details) - you can even merge multiple disparate interfaces into one combined virtual interface (though you risk timing issues with wildly disparate hardware).
You can assume almost every Pro Audio hardware manufacturer designs and tests on Macs & OS X *first*, and then hacks together a Windows ASIO & (maybe) WDM driver as an afterthought. All your 'regular' Windows apps will be using WDM - games, Skype, what have you, and all your Pro apps will be using ASIO. Two different drivers for two different sets of applications. On a Mac? CoreAudio is the same for everyone. Games and 'regular' apps use the same backend as Pro Tools or Ableton or Logic.
tl;dr: I tried the whole 'Linux DAW' thing years ago, gave up, got a Mac, actually spent my time getting stuff done and not fighting broken systems/drivers.
Choosing a DAW is much like choosing a programming language & IDE in one, all your knowledge becomes domain specific, and as such, as soon as you get serious it's difficult to consider open source options seriously.
No.
The only thing that is partly stuck to the program is the workflow. If you just know in which order to press the buttons and don't know what they do or why they do it, then sure, you knowledge applies only to that program. But if you know what the tools are doing then you can jump to pretty much any DAW on the market and make it work. You'll be slightly less efficient because of changes in workflow while getting used to the new workflow. And the bigger the difference, the weirder it is for the first couple of hours. Jumping from ProTools to Ableton Live is quite weird, for example.
I've jumped between ProTools, Ableton Live, and Reason with no problems. I've also played around with Logic, Reaper, and Cubase.
Reaper is quite nice, and works well. Also very easy to grok the workflow as it is a simple one that is quite like ProTools in that it relies heavily on an old workflow that would fit an audio engineer from the 1980s or 90s.
Reaper rocks for that price!
Lots of experience with this.
1) Real Time Kernel (compile it if you have to)
2) Ardour (My best results are with Ardour 2 so far)
3) Jack Audio
4) Rosegarden (for midi)
5) Linuxsampler (convert midi to audio using professional samples)
It doesn't hurt to have 8 cores, 32 GB RAM and tons of HD.
With 2 40" 4K screens I can display just the ardour mixer with 30-50 tracks across one screen.... and have room for the other apps.
I've done many tracks.... easily up to 70 tracks... with plugins... low latency....
I use Ardour on Fedora, connected to a Focusrite Saffire Pro 40, and heavily using the great and opensource Calf Studio Gear Audio plugin suite. Everything works really well, and the setup could be used to put together a really high quality album. We almost exclusively use it for recording church services, which doesn't exercise the full potential of the setup. One of these days I'll have time to put together a project that takes advantage of more of the capabilities we have.
Linux is not yet usable for professional-quality home recording.
The article, written by a guy who claims to have "spent 2 1/2 years in the local recording industry" tells us that he recently bought a "Lexicon Alpha 2-channel" interface. Just so you know, the Lexicon Alpha 2-channel interface is a $49 POS USB interface that you wouldn't use for anything more critical than a low bit-rate podcast. Let me know when I can use an Apogee or Apollo or Avid interface with a Linux box.
Linux is a fine platform for use in a recording studio. I use it all the time, but never as a main production platform. It can be used to stream samples, for storage, and even to off-load audio processing and rendering chores. The implementation of VSTs and VSTi's in Linux is still so wonky as to be unusable. And nobody ever used Audacity in their home recording studio to produce anything professional.
I hope Linux eventually does become a viable platform for music production. I've been waiting a long time for that to happen.
You are welcome on my lawn.
I don't even know where to begin with this.
First off, Windows has continued to be non-standards compliant, inconsistent in its directions, and has haphazardly dealt with security updates. I shouldn't have to wait for months to have a security vulnerability patched. Have they gotten better? Yes. But this licensing garbage that's tied to the firmware for 8 and up, and Office licensing that's tied to email is a horrendous endeavor to maintain.
If you want to consider Linux a failure of a desktop on the account of Gnome, Pulse Audio, Unity, and SystemD, you fail to realize that each one of those are simply software packages running on top of the operating system. Most of the issues you describe are either anecdotal, or simply a regurgitation of what you've heard. I've never had problems with Pulse Audio, nor has anyone that I know personally. Gnome I've had issues with when attempting to integrate with an existing installation, but that was during 3 Beta, Unity sucks, and is is sluggish, but it isn't buggy. SystemD, a lot of people complain about it, and I understand why, it's a complete overhaul, overbearing/far reaching, but since it's adoption out of Beta, I've yet to have ANY workstation or server have any reliability/stability problems. There have been minor configuration issues occasionally, but nothing that's atypical from normal software upgrades.
I feel bad responding to a first post troll but wrong is just wrong. I'm an amateur musician in the Olympic sense (I've never been paid for it in a meaningful way). It turns out you can do incredible things with a Linux desktop and just as much (if not more) with a Linux recording setup. Now, would I suggest using a Linux set up when trying to collaborate with other musicians? No. Sounds like a terrible plan. Can you record decent sounding multitrack stuff on your own with the right equipment? Absolutely.
http://soundcloud.com/chuck-bl...
It's not music that most people will like but it was recorded completely using a fedora studio spin. It can be done.
"Watch your cornhole, bud."