On 02/02/16 08:49, Ian Zimmerman wrote:
> On 2016-02-01 22:02 +0100, Didier Kryn wrote:
>

>>      If you do a lot of sound manipulation, this is an area in which I can't help.
>> Linux sound is a nightmare to me and I never found a sensible description of it all,
>> so using pulseaudio by default because it mostly works out of the box. But there was
>> a thread about it one or two weeks ago...

>
> I use straight ALSA and it's quite usable to me, if somewhat
> intimidating on first approach. What are the nightmarish aspects, maybe
> I can help?
>

Serious linux audio is very configurable, sophisticated and flexible, this makes
it complex.

Some hardware manufacturers don't care to support a smallish market, leaving the
end user to deal with integration into the system.

A lot of serious linux audio setups are within the context of a large
installation with serious in-house expertise and quite specific needs .. so DIY
as they wish, and deal with the complexity as the price of configurability.

Some very serious linux audio is sold as part of top-end integrated hardware and
carefully set up by those vendors, again they favour configurability over ease
of setup.

The software based smaller digital audio workstation end attracts many people
interested in the underlying algorithms, often programmers as well as musicians.
They are seeking the flexibility to at least script if not actually program
certain parts of their setup. They are often interested in a system putting
together a lot of small, specific tools of their choice which follow the unix
philosophy of do one thing and do it well as opposed to wishing for a ready to
play fully integrated audio application. Because of this they are prepared to
deal with complexity for the same reasons as the previous two types of serious
audio user.

Thus serious linux is a very attractive tool for a few, and best avoided for
many. If you know the stakes ... it is not a nightmare, it is a tremendous
opportunity.

The Linux Audio Users email list is a great resource if you want to take it up.

So are the lists of most of the tools you want to use.

Other musicians are often much more interested in choosing one big application
that caters for most of their needs (then all of them, once they prune ambitions
to match the workflow they learn). This is quite understandable and they are
very well catered for by commercial offerings, usually in commercial operating
systems, so they don't generally bother with linux. They find the cash required
if they can. There are options in linux but this does require actively seeking
them out, and the serious ones can be quite expensive if they are commercial
tools. The big businesses in this field depend heavily on actively chasing the
musicians ... to sell them stuff ... and focus on platforms that are also big
business, and familiar to the biggest number of potential buyers.

So serious linux audio ends up not at all suitable for plug and play desktop use.

Pulseaudio goes a long way to solving this, autoconfiguring everything to suit
the expected desktop usage patterns without asking any questions. It is much
more acceptable now that it will allow jack as a backend between it and alsa,
rather than insisting it controls alsa itself, since now the usual desktop audio
stuff can be fed into the serious audio chain, or just to the speakers when you
want to use skype or something, and does not need to be disabled.

Pulseaudio is sensible choice if you are after simple audio on the desktop, it
isn't a nightmare unless you try to do more than the standard stuff. It will of
course drag in all the autoconfigure black-box stuff that devuan tries to avoid
... it must, to integrate with that whole black box ... that is why it exists.
DIY is always a bit more work. You will need to learn about audio as well as
linux to get anywhere.


Simon