As I am approaching the writing of some review of Audio performances of different Linux operating systems it is better to give some definition. Refer to this page for any word in bold inside posts in the category “Audio Linux OSes Reviews”.
Compatible hardware: a piece of audio hardware that is reported as compatible “out of the box” by the official relevant software support or by previous documented users experience.
Simplicity, Code-correctness, User-Centric, Openness, Freedom: given accordingly to the Arch Linux KISS principles.
Official software sources: Any repository or database of software directly maintained by the particular distribution developers or community.
Common audio software: the audio software that is more commonly used under Linux. Lacking any statistic at this moment, I will just define it as the software included by default in the latest Ubuntu Studio release.
System: for Audio Linux OSes reviews the system is a whole computer with its internal or external hardware and software machinery. For example, my Compaq Presario CQ61, with my Focusrite Scarlett 2i4 and ArchBang Linux on top of it, is a system. An eventual mixer supplying the signal for the sound-card of the computer is not part of the system.
Standard user system: the system that a pro audio user, on average, is supposed to aim to. It is made by a computer composed of compatible hardware comprehensive of a sound device and a Linux operating system configured at least with a window manager, the pro audio tweaks required and the common audio software.
Lowlatency system: a system that conforms to any of the definitions given here. I will make clear when relevant if the system is practically lowlatency or psychoacustically lowlatency.
Effective documentation: when referring to the documentation of a distribution, we define it effective for the user when the documentation is easy to navigate and the required information easy to find and extrapolate.
Audio stability: as defined here.
The CrocoDuck’s Official Rate: a score in tenths given to the following subjective attributes of a Standard user system in the specified ways:
- Packages and Package management: _/10. A judgment of how many packages are ready to install, if they are up to date and how effective is the package manager. Scale pivots: 1/10: the official software sources are lacking software, especially in the audio realm (also among the common ones). 5/10: all the common audio software is in the official software sources, they are at least at the latest stable release and the installation/upgrade/downgrade processes are simple and straightforward. 10/10: all the audio software to date is in the official software sources, it is bleeding edge and the installation/upgrade/downgrade processes are simple and straightforward.
- Beginner friendly: _/10. A judgment of how much the system will be easy to use and tune for a inexperienced Linux user with the goal to be able to start making music as soon as possible. Scale pivots: 1/10: the inexperienced user can’t do anything with it. 5/10: the inexperienced user will be successful if he/she puts the required effort into it. 10/10: the inexperienced user will be able to make music out of the box.
- Documentation/Support: _/10. A judgment of how explanatory, effective, up do date and easy to use is the official documentation and forum/official support from a distribution. Scale pivots: 1/10: the documentation is poorly written, lacking, out of date and almost no support is offered through forums or other means. 5/10: the documentation covers all the most important aspects of the operating system installation, configuration and maintenance, it is at least updated to the last stable version (if any) and the community is kind and helpful. 10/10: the documentation covers basically all the system in a effective way for the user, it is daily updated and the community is kind and helpful.
- Audio stability: _/10. A judgment of the audio stability of the system after all the eventual required configuration has been done. Grades given using the xruns frequency as a variable with respect this reference pivot: 5/10: the system respects the definition of stability in terms of frequency of xruns given here.
- Audio ready after install: _/10. A judgment of the out of the box audio capabilities of the system. 1/10: nothing audio related can be done on the system just after install. 5/10: after install the system is able to work with compatible audio interfaces and all the basic “desktop user” audio (playback, recording, browsing, VoIP…) is running smoothly. 10/10: the system is already capable of lowlatency for practical purposes highly audio stable workflow out of the box.
- Stability: _/10. A judgment of the system stability intended as the capability to run over long times without encountering hard to solve troubles. It is given involving the effort required by the user to maintain the system stable. Scale pivots: 1/10: the system becomes sluggish and prone to crashes in less then a year of usage. Upgrades are likely to harm the system. 5/10: the system needs at lest 3 years to become sluggish and prone to crashes. Upgrades may introduce temporary regressions. 10/10: the system always feels responsive as the very first day after installation, upgrade cannot introduce regressions.
Any system that will score more than 7/10 in Stability and Audio stability will win an Official CrocoDuck’s Pro Audio Seal of Approval!
As specified, the attributes are subjective. Other user may experience differences, maybe just because they are running on a different hardware. Also, for me a score of 10/10 somewhere could be very positive. For another user the opposite. These attributes are unappealing from the scientific point of view and are to be taken carefully. For sure many objective attributes could be defined or already they are. However, I think that this scheme is a good compromise as it meets my need to be able to give to the reader a clear point. All this while avoiding a too technical speech. Also, it is not to be forget that objective properties of systems will translate to objective properties of signals and then sound, but those are still related to subjective attributes as soon as we listen to the sound. Making previsions of these phenomena would require psychoacoustical models and over-complication or even real experiments to be performed on a listening panel. A thing I can’t really do. Lastly, as the Linux world evolve quickly, it would be quite pointless to carefully measure and benchmark a system, as in a matter of months anything could change and a lot of time would result as being wasted.
To sum up: those are just definitions given by me to make clear my arguments when I talk about something.