On 08/06/2012 01:57 AM, Brent Busby wrote:
> On Sun, 5 Aug 2012, david wrote:
>
>> I didn't say it was either necessary or most common. It does depend on
>> the person doing the hearing. But I remember many people whose
>> response to singing or playing an instrument is, "Oh, I can't play
>> like that," so they won't even try.
>
> There are some players whose real no-overdubs playing is good enough to
> do that. Dave Weckl and Vinnie Colaiuta have both often made me wonder
> who I think I'm kidding about playing drums. :)
Yet if they weren't available on recordings, chances are very few people
would have ever heard their no-overdub playing. They'd only have
inaccurate descriptions by others to go on.
--
David
gnome(a)hawaii.rr.com
authenticity, honesty, community
http://clanjones.org/david/http://dancing-treefrog.deviantart.com/
On 06/08/12 19:46, Brent Busby wrote:
> On Mon, 6 Aug 2012, Simon Wise wrote:
>
>> On 06/08/12 02:16, Ralf Mardorf wrote:
>>
>>> Most people paying for listening to music aren't musicians theirselfes. They
>>> neither are that (self-)educated that they need (or even are able to listen)
>>> to more complex music. Mass media have the
>>
>> which is perhaps one of the huge changes in music in the last 60 or so years
>>
>> ... most popular music was played and sung by the listeners until then, the
>> ability to play something pre-recorded, here music that is not directly
>> connected to your own playing or singing, or at least someone in the room with
>> you, is really very recent in the development of music. Older popular music
>> was as much about the pleasure of reproducing it as it was about listening to it.
>
> That was true I think until the late 90's, and definitely has changed in the
> 2000's back to the way it was in the earlier period you're talking about. Pop
> music now seems to be about vocals, vocals, vocals...and by the way, did I
> mention the vocals? If there's any instrumental track to speak of at all, it
> usually seems to consist of some TR-909ish bass drum playing quarter notes, and
> maybe some sine wave synth bass on top of that, a few padding chords if you're
> lucky. And on top of that, it's literally all singing, all dancing. The bass
> drum part is only there to keep someone out there from proclaiming the emporer
> has no clothes when they notice there's actually no song really there.
>
> And the listeners seem to like it that way. If anything is too instrumentally
> complicated, they think it's "weird." Psychedelic music is now "creepy." We've
> re-entered the era of the campfire singalong. And let's not even get into the
> prevalence of karaoke games for consoles. Styles like rock and jazz are starting
> to feel to me, much as I love to play them, like something musicians do for
> eachother's enjoyment, sometimes in over-the-Net collaboration projects. It's
> almost become something where you have to be a player yourself to even
> appreciate it.
there is plenty of traditional popular music that is quite complicated, and also
meant to be played by many not just a few experts ... sometimes its more fun to
try something tricky than something easy ... that's the whole point of lots of
games, popular and widely practised does not always mean dumbed-down.
Simon
On Sun, August 5, 2012 8:01 am, Ralf Mardorf wrote:
> On Sun, 2012-08-05 at 07:12 -0700, Len Ovens wrote:
>> I have another idea... I make use of the system runlevels. Ubuntu only
>> uses runlevels 0,1,2 and 6.
>
> What runlevels will Ubuntu Studio have got, when Ubuntu makes the switch
> to systemd?
Both upstart (which Ubuntu uses now) and systemd are compatable with init.
I am in fact using upstart scripts now for my runlevel changes and because
many of the services I want to control are already upstart scripts (ie.
cron and mysql). Even with these new tools for doing what init used to,
the idea of operating modes is still there. It would be possible to change
RL 0 to halt.mode and RL 1 or S to single.mode, but even if that happens
it should be possible to create a new mode called audio.mode. However, I
don't see that happening because both upstart and systemd are created to
allow parallel processing and daemon startup rather than the serial
approach init uses. There seems to be no move to get rid of any of the
operating modes being used now. We still need an orderly startup and
shutdown.
> Writing a script and use it with a button for the panel IMO is the
> better way to go. Resp., everybody building customized kernels directly
> could disable anything, but performance.
Using a button from a panel is the way I like it, but I have talked to
people who feel differently and would like it to start when jackd does.
The way I am putting together can be used either way. The system side
things have to be separate anyway... so the systray script actually runs a
cli command that can just as easily be started by qjackctl or a session
manager at need. I actually have 4 RL set up normal/audio/graphics/video.
but they could just as easy be
normal/audiofullspeed/audiohalfspeed/audioanotherspeed. (and as far as I
know it would be possible to add RL7 if that was not enough.... though I
honestly don't know how upstart or systemd would handle those... init
would.
Rather than build a special kernel I would just change
/etc/init.d/ondemand to set performance... it would just take changing one
word in one file. However, because I have heard of (not experienced)
systems shutting themselves down due to overheating when run too hard I
don't want to suggest that as a blanket solution. I also feel that being
able to run at a lower speed that allows the machine to run with fans off
is valuable too. The recording can be at low speed and the post production
at full.
By the way, runlevel switching seems like a non-event to the user so long
as the runlevel chosen does not shut the session/X down. Of course if the
user is using something that gets shutdown (like wireless in my case) it
would be noticed.
My next step is to create a gui that allows the user to choose what to
shutdown when. This will take more time that everything I have figured out
so far.
> 2 Cents,
Don't know if anything I say has monetary value... The one worthwhile
thing I learned in tech school was:
The more you know, the more you know you don't know.
--
Len Ovens
www.OvenWerks.net
On Sun, August 5, 2012 1:54 pm, Robin Gareus wrote:
> On 08/05/2012 10:46 PM, Ralf Mardorf wrote:
>> On Sun, 2012-08-05 at 22:20 +0200, Robin Gareus wrote:
>>> I really wonder why you are all raving about this.
>>
>> Take a look at your own NB [1].
>
> That's why there's jackfreqd.
I did look at jackfreqd, which according to it's author has a high
hysteresis which in effect turns full speed on almost when jackd starts
and keeps it there till it ends... it even mentions that at jack startup
there may be one xrun as the speed ramps. So for having yet another
process running, I can get the same thing I get by having jackd run a
script at startup that puts the machine in performance mode and quites...
no extra process running while I am doing audio either.
really, any tweak is an as needed thing, not for everyone, but the
knowledge is there if needed.
"on modern machines" I have two things to say... my more modern year old
machine has more problems than my 8 year old machine does.... second, I
don't replace machines just because something new is out. there is enough
churn in the world as is. (end of rant- sorry)
> This is an edge case and very hard to reproduce and very unlikely to
> cause any issue to normal users (Sorry Nando, you and your setup at
> CCRMA don't qualify as /normal/. It's an awesome setup).
>
> NTL, you could have left the link to http://rg42.org/oss/jackfreqd/
> here. It's piece of software that can work around the described problem.
It is also true on lower end machines. Perhaps less of an edge case.
--
Len Ovens
www.OvenWerks.net
On Sun, 5 Aug 2012, Ralf Mardorf wrote:
> On Thu, 2012-08-02 at 09:51 -0400, Brett McCoy wrote:
>> AFAIK, if you have your cpu governor set to 'performance', you
>> shouldn't have an issue with frequency scaling anyway. I have so much
stuff consuming power in my studio as it is, I doubt Jack is going to
make that much of an impact on it. :-)
>
> I guess the governor needs to be set up to a fixed frequency.
> Performance or any lower fixed frequency. FWIW measurements with an
elCheapo tool resulted in nearly no difference for my dual-core 2.1GHz
set to performance and ondemand.
If you are changing the governor setting in Ubuntu (12.04 anyway), be
aware that OnDemand gets set 60 seconds after login. I assume the delay is
to allow a speedier login... but the governor has already been set by
that time anyway. At least cpufreq-info shows ondemand right after login.
The page http://wiki.linuxmusicians.com/doku.php?id=system_configuration
suggests setting performance in /etc/rc.local but this results in
performance when I first login and then it jumps back to ondemand. On the
same page there is another suggestion to use: /etc/init.d/ondemand stop
which doesn't work either as the stop command does nothing. However: sudo
update-rc.d ondemand disable
does get rid of the 60 second reset for ondemand and then setting
performance in /etc/rc.local will work.
I don't know that this is the best solution. I have had no problems
running at full speed even over night in my cheap atom based netbook, but
I have heard of some laptops that have heat problems even in ondemand (or
windows) if they are pushing it too hard. The fan on my netbook is noisier
too. The idea of running at a lower constant speed would help. I seem to
be able to do quite well at even 800Mhz and have a quiet machine.
I have another idea... I make use of the system runlevels. Ubuntu only
uses runlevels 0,1,2 and 6. I think most distros don't use all the
runlevels, last time I was running Slackware I think they used RL3 as
well. Anyway I use RL3 for audio. RL3 is very quiet, I shut off:
- cron and friends (this gets rid of auto updates and log rolling for
example)
- mysql and other daemons I don't need (netmanger is a good candidate to
if you are not using netjack)
- my wireless kernel module (ath9k) which gives me xrun/minute when
running and xrun/5sec with netmanger turned off...
- some people would find this a good way to shut off pulse as well. (if
they need it installed... for recording skype calls for podcasting for
example)
And I set cpufreq to performance.
I set up a bash script that has to be run as root to change runlevel to
either 2 or 3 and have added it to policy kit so it can be run with
pkexec without a password screen. (same way as we already can set
runlevel 0 and 6) This can be run as part of the jackd startup or, I have
an icon in systray for one click use.
It would be easy to set RL 3,4 and 5 to be three different cpu speeds
depending on need.
Len
--
Len Ovens
www.OvenWerks.net
Hi folks,
If you haven't seen it yet, someone started creating a whole new audio
system for Linux and BSD.
http://klang.eudyptula.org/
P.S. Yep, a system, not just a server. Just as we got ALSA, PA and
JACK working together nicely.... :)
Alexandre Prokoudine
http://libregraphicsworld.org
hi all,
bit of a niche question here, i can't get the mididings 'System'
function call to work as i understand it. i'm sure i'm doing something
wrong, i wondered if one of you could help? i wrote a python patch,
but couldn't get it to work. so, i made it simpler and simpler, to try
and understand it. in the end i got down to this, which i named
'mididings_test.py':
=========================
from mididings import *
run(
System('aplay /home/robin/autest.wav')
)
=========================
now, please note that this is trivially simple, it's not the command i
actually want to run, it is purely a way for me to understand how
'System' works.
yes, the autest.wav file exists, and it is in that location. and yes,
aplay is installed (part of alsa_base)
when i run the script ($ python mididings_test.py), it does nothing
so, what am i doing wrong here? once i figure that out, i can write
the real patch i require
cheers,
--
robin
http://fu.ac.nz - Auckland's Free University
Hi, I was wondering, what can we non-dev users do when our favorite
piece of FOSS software is lacking a little feature, but one that makes
the software for us of little or no use?
I'm thinking of seq24 not working, since a year or so, with JACK
Transport - which for an app like seq24 makes it almost useless. It
really itches me and I'd like to do what's in my power to help fix it.
I was thinking, since this would be (I think) a few hours of devs
work, couldn't I/we raise some money to pay a dev to do it? How would
I/we do it?
In long term thinking, we could have a site where users propose a bug,
devs name a price for it, and when the money raising reaches that
quantity, the dev starts to work on it and when he fixes it he recieves
the money. Basically a mix of kickstarter and amazon's mechanical turk.
I think this could work quite well for such bugs, which often appear in
awesome but semi-abandoned software (freewheeling and kluppe for
example come to [my] mind). I totally get that the original devs may
have lost interest and moved on in life, but being FOSS, the code is
there and maybe a few paid hours of a dev could get it working/add a
basic feature you really need.
Thoughts?
renato
Hi All,
I know I've posted an announcement about Praxis / Praxis LIVE here
before. Thought a few people might be interested in this blog entry
about Praxis' underlying architecture and the influence of the Actor
Model. Be really interested in any comments with thoughts, insights,
corrections, stupid mistakes, similar models in use elsewhere, etc.,
etc.
http://praxisintermedia.wordpress.com/2012/07/26/the-influence-of-the-actor…
Thanks in advance.
Best wishes,
Neil
--
Neil C Smith
Artist : Technologist : Adviser
http://neilcsmith.net
Praxis - open-source intermedia system for live creative play -
http://code.google.com/p/praxis
OpenEye - specialist web solutions for the cultural, education,
charitable and local government sectors - http://openeye.info
