Hi,
At the beginning of this thread ppl were saying Linux is too hard and I was with them on that, based on my previous experience.
My how things have changed in a few years !
I just downloaded the latest Ubuntu 10.10 and it installed very quickly and immediately recognised all my devices and didn't require any help or drivers or serial key etc. So easy. I was browsing the web and playing WAVs through the Onkyo soundcard in less than five minutes. It's actually far easier than Windows to install and took a staggeringly shorter time too.
I have just one issue - how to connect into my Buffalo NAS. This only comes with software for Windows and Mac. Can anyone direct to a site that has this info ?
Anyway, I'm extremely pleased with Ubuntu. Now I have to re-read this thread to find out how to tweak it !
Are there any "Linux for Dummies" resources out there ?
Thanks,
Tom ( a new Linux believer ! )
At the beginning of this thread ppl were saying Linux is too hard and I was with them on that, based on my previous experience.
My how things have changed in a few years !
I just downloaded the latest Ubuntu 10.10 and it installed very quickly and immediately recognised all my devices and didn't require any help or drivers or serial key etc. So easy. I was browsing the web and playing WAVs through the Onkyo soundcard in less than five minutes. It's actually far easier than Windows to install and took a staggeringly shorter time too.
I have just one issue - how to connect into my Buffalo NAS. This only comes with software for Windows and Mac. Can anyone direct to a site that has this info ?
Anyway, I'm extremely pleased with Ubuntu. Now I have to re-read this thread to find out how to tweak it !
Are there any "Linux for Dummies" resources out there ?
Thanks,
Tom ( a new Linux believer ! )
There are numerous Wikis written and maintained by the Ubuntu crowd.
As usual -- Google is gonna be your best friend -- also in Ubuntuland.
Hi,
Actually, there are a lot resources so it's an embarrassment of riches and hard to know where to start. But I found a 'pocketguide' that seems to cover enough to get started. It's a bit like the manuals that came with MS DOS all those years ago.
Main page - Ubuntu Pocket Guide and Reference
I couldn't find any help about getting the Buffalo NAS to work. Lots of people asking and either no response or responses that didn't resolve any issues and brought up more (irrelevant) reading. Google wasn't playing friendly.
So instead I connected to the NAS login 192.168.0.102 in Firefox and played around with the ftp share settings. Then I got lucky ! In Ubuntu's dropdown menu there is a "connect to server" and I tried putting in the address for the Buffalo in an ftp connection and it worked.
So it's up and running, sweet as a nut. Boots very fast, shutdowns like a light bulb. Sound seems to be only very marginally better than Windows but I expect I need to tweak to get it at its best ?
Just one problem now. My remote keyboard has a key on the bottom left with the Windows flag. Where can I get an Ubuntu sticker ? ;-)
Tom
Actually, there are a lot resources so it's an embarrassment of riches and hard to know where to start. But I found a 'pocketguide' that seems to cover enough to get started. It's a bit like the manuals that came with MS DOS all those years ago.
Main page - Ubuntu Pocket Guide and Reference
I couldn't find any help about getting the Buffalo NAS to work. Lots of people asking and either no response or responses that didn't resolve any issues and brought up more (irrelevant) reading. Google wasn't playing friendly.
So instead I connected to the NAS login 192.168.0.102 in Firefox and played around with the ftp share settings. Then I got lucky ! In Ubuntu's dropdown menu there is a "connect to server" and I tried putting in the address for the Buffalo in an ftp connection and it worked.
So it's up and running, sweet as a nut. Boots very fast, shutdowns like a light bulb. Sound seems to be only very marginally better than Windows but I expect I need to tweak to get it at its best ?
Just one problem now. My remote keyboard has a key on the bottom left with the Windows flag. Where can I get an Ubuntu sticker ? ;-)
Tom
Last edited:
Hi folks.
Just to let you know. I won't give up on Linux.
I installed my Ubuntu Natty (11.04) Beta the third time now. I've been trying this and that.
The third time I wanted to do things right. 100% focus on audio first.
After installing my basic stuff, I wanted to challenge myself to get my
RME HDSP9632 working with pulseaudio kept in place. Usually I remove pulseaudio.
But that usually ends up with a broken Gnome audio system.
I was aware of the broken hdspmixer that's part of the "officially released" alsa-tools 1.0.24.1. I knew I had to use amixer (PITA) instead.
I thought don't make it too difficult, don't set the target too high.
Just 44.1/16 stereo on analog out would be a major achievement for the beginning. In the past I never managed to get things working under pulseaudio.
Now.
Let me summarize what I've been doing:
1. Searching and finding the tools required to load HDSP firmware,Tools,Tool-GUis, was required. Otherwise nothing will work.
The HDSP mixer tool won't work currently anyhow. That's why amixer must be used.
2. I had to figure out the channel routing ( connect the right input with the right output) on that card to be able to set gain and channel routing with amixer. I wrote a script to figure out channel routing with amixer some time ago ( can be found at RME forum).
3. I wrote and added another setup script to the autostart dialog, to do the card setup at boot. (Only this way I can work without hdspmixer. This of course limits me to a kind if static setting for that card)
4. I had to do other basic settings with hdspconf/amixer (e.g. samplerate)
5. I had to add hdsploader (firmware) to rc.local. Without firmware nothing will work.
6. Make sure that users and groups are correctly connected to group audio.
7. Setting up realtime and nice rights in limits.conf for group audio
pulseaudio requires heavy priority to work without XRUNS . ( similar to jack)
8. Find out about and install certain pulseaudio tools
9. Figuring out the need to implement certain pulseaudio workarounds (e.g. famous tsched=0 tweak)
10. Change pulseaudio setup within configuration files (which was one of the major challenges)
11. Introduce an .asoundrc to define my 9632 as default soundcard.
All that took me while. Guess what? I managed.
ToDos:
1. I experienced XRUNS when streaming music over NFS share. pulseaudio seems to put high demands on the environment.
2. Checkout Skype
3. Get more insight into pulseaudio
I might write all that - more detailed - down on my blog sooner or later.
Summary:
Folks, I really enjoy the new look and feel of Ubuntu Natty. It's getting better
and better. Though I'd love to see Amarok and Digikam ported to a Gnome environment.
Everything seems to evolve in Ubuntu Linux - one thing though gets more and more complex and frustrating ""Linux Audio"".
Above exercise can't be done by 99% of the average Ubuntu/Linux user.
I'm sure something has to be done about Linux Audio. The question still is who feels responsible to do that. The question really is. Is anybody aware of that mess?
It probably should be Debian/Ubuntu/Redhat or similar project level, that brings all parts together and has the power to drive such a project into a
certain direction..
Let see how things are going to evolve.
Enjoy.
Cheers
Just to let you know. I won't give up on Linux.
I installed my Ubuntu Natty (11.04) Beta the third time now. I've been trying this and that.
The third time I wanted to do things right. 100% focus on audio first.
After installing my basic stuff, I wanted to challenge myself to get my
RME HDSP9632 working with pulseaudio kept in place. Usually I remove pulseaudio.
But that usually ends up with a broken Gnome audio system.
I was aware of the broken hdspmixer that's part of the "officially released" alsa-tools 1.0.24.1. I knew I had to use amixer (PITA) instead.
I thought don't make it too difficult, don't set the target too high.
Just 44.1/16 stereo on analog out would be a major achievement for the beginning. In the past I never managed to get things working under pulseaudio.
Now.
Let me summarize what I've been doing:
1. Searching and finding the tools required to load HDSP firmware,Tools,Tool-GUis, was required. Otherwise nothing will work.
The HDSP mixer tool won't work currently anyhow. That's why amixer must be used.
2. I had to figure out the channel routing ( connect the right input with the right output) on that card to be able to set gain and channel routing with amixer. I wrote a script to figure out channel routing with amixer some time ago ( can be found at RME forum).
3. I wrote and added another setup script to the autostart dialog, to do the card setup at boot. (Only this way I can work without hdspmixer. This of course limits me to a kind if static setting for that card)
4. I had to do other basic settings with hdspconf/amixer (e.g. samplerate)
5. I had to add hdsploader (firmware) to rc.local. Without firmware nothing will work.
6. Make sure that users and groups are correctly connected to group audio.
7. Setting up realtime and nice rights in limits.conf for group audio
pulseaudio requires heavy priority to work without XRUNS . ( similar to jack)
8. Find out about and install certain pulseaudio tools
9. Figuring out the need to implement certain pulseaudio workarounds (e.g. famous tsched=0 tweak)
10. Change pulseaudio setup within configuration files (which was one of the major challenges)
11. Introduce an .asoundrc to define my 9632 as default soundcard.
All that took me while. Guess what? I managed.
ToDos:
1. I experienced XRUNS when streaming music over NFS share. pulseaudio seems to put high demands on the environment.
2. Checkout Skype
3. Get more insight into pulseaudio
I might write all that - more detailed - down on my blog sooner or later.
Summary:
Folks, I really enjoy the new look and feel of Ubuntu Natty. It's getting better
and better. Though I'd love to see Amarok and Digikam ported to a Gnome environment.
Everything seems to evolve in Ubuntu Linux - one thing though gets more and more complex and frustrating ""Linux Audio"".
Above exercise can't be done by 99% of the average Ubuntu/Linux user.
I'm sure something has to be done about Linux Audio. The question still is who feels responsible to do that. The question really is. Is anybody aware of that mess?
It probably should be Debian/Ubuntu/Redhat or similar project level, that brings all parts together and has the power to drive such a project into a
certain direction..
Let see how things are going to evolve.
Enjoy.
Cheers
Four 'Powered by Ubuntu' stickers
and Four 'Ubuntu Key' stickers9
And they are free, just a self addressed stamped envelope
rgds
jms
and Four 'Ubuntu Key' stickers9
An externally hosted image should be here but it was not working when we last tested it.
And they are free, just a self addressed stamped envelope
rgds
jms
About pulseaudio, sure in the early stages I had problems too like crackling and such but since pulseaudio matured all my problems went away so now it just works, and is mighty convenient =)
I'm not using any of the hacks like using the legacy scheduling, only customization from default ubuntu config is that I've chosen a costlier resampling algorithm (src-sinc-medium-quality), and chosen sample-format, default fragments and default-fragment-size-msec to correspond with my sound card.
But even without the calculated fragments and fragment size it works without glitches nowadays.
Might as well throw in the link of where I found how to calculate them: Linux Mint Forums • View topic - PulseAudio - Tune your setting and prevent glitching
In short:
default-fragments = buffer_size / fragment-size
default-fragment-size-msec = fragment_size / sample-rate * sample-format (bits) * sample-channels
Which in my case for my Xonar-card is:
I: sink.c: device.buffering.buffer_size = "352768"
I: sink.c: device.buffering.fragment_size = "176384"
=>
default-fragments = 2 # buffer_size / fragment_size
default-fragment-size-msec = 63 # since the card wants s32_le
I'm not using any of the hacks like using the legacy scheduling, only customization from default ubuntu config is that I've chosen a costlier resampling algorithm (src-sinc-medium-quality), and chosen sample-format, default fragments and default-fragment-size-msec to correspond with my sound card.
But even without the calculated fragments and fragment size it works without glitches nowadays.
Might as well throw in the link of where I found how to calculate them: Linux Mint Forums • View topic - PulseAudio - Tune your setting and prevent glitching
In short:
default-fragments = buffer_size / fragment-size
default-fragment-size-msec = fragment_size / sample-rate * sample-format (bits) * sample-channels
Which in my case for my Xonar-card is:
I: sink.c: device.buffering.buffer_size = "352768"
I: sink.c: device.buffering.fragment_size = "176384"
=>
default-fragments = 2 # buffer_size / fragment_size
default-fragment-size-msec = 63 # since the card wants s32_le
Four 'Powered by Ubuntu' stickers
and Four 'Ubuntu Key' stickers9
An externally hosted image should be here but it was not working when we last tested it.
And they are free, just a self addressed stamped envelope
rgds
jms
Well, I was joking but.... nice stickers ! Thanks for the link ! I'll be proud to add them to the keyboard and case !
don't know that particular NAS, but usually you don't really need any special software to connect to them. Usually there are Web, Telnet and/or SSH interfaces to manage them. Check your NAS user manual.
(Oh, to access the files on the NAS, from Linux it would be better to use NFS rather than SMB - if the NAS has support for it).
trying a Beta is definitely not the best thing to do if you're looking for a smooth experience. For that I would not even try a freshly released version, but one which have been released since at least 3 months.
why do you want to do that?
The RME is by no means a normal "desktop" sound card.
Pulseaudio have never been intended to be used for "hi-end" or "pro" audio (and related hardware). Its sole purpose it to make life easier for the average desktop user: play sound through the typical sound card, plug in an USB headset and be able to use it, etc.
Perhaps you should rather try Ubuntu Studio instead. That should handle it gracefully, and it would actually be a bug if it doesn't.
trying a Beta is definitely not the best thing to do if you're looking for a smooth experience. For that I would not even try a freshly released version, but one which have been released since at least 3 months.
why do you want to do that?
The RME is by no means a normal "desktop" sound card.
Pulseaudio have never been intended to be used for "hi-end" or "pro" audio (and related hardware). Its sole purpose it to make life easier for the average desktop user: play sound through the typical sound card, plug in an USB headset and be able to use it, etc.
Perhaps you should rather try Ubuntu Studio instead. That should handle it gracefully, and it would actually be a bug if it doesn't.
I'm trying to build a desktop system ( for serious audio playback I currently use my SB Touch).
The system should handle satisfactory
1. pictures
2. video
3. audio playback
4. audio file tagging
5. ripping (optional)
at a reasonable quality level !!
And I'd like to use the newest kind of desktops.
Everything needs to work smooth on a home network, with multiple clients.
Ubuntu server is used. That server is running mysql for pictures,
audio (squeezebox).
Unfortunately Amarok won't seem to work under Gnome anymore.
It's seems to be so deeply integrated into Plasma that it fails under
Gnome. Perhaps only due to that I need to give KDE another try.
I had a look also at all other Gnome compatible players. They all are pretty basic apps. Nothing I'd use or would recommend.
Cheers
I am afraid ubuntu studio 10.04 uses pulseaudio just the same way regular ubuntu does.
But it is trivial to remove pulseaudio in a minute, just a matter of copying/pasting a few commands to the terminal. I have used the tutorial Elegantly Disabling PulseAudio in Ubuntu 10.04/10.10 | www.jeffsplace.net many times, always with 100% success. Even a beginner can acomplish that easily.
I would not recommend using the betas between LTS releases either. Their lifespan is miniscule, upgrading painful, changes too deep. LTS 10.04 has gradually become a fine distribution, with the major bugs already fixed or reasonable workarounds known. There are lots of useful and up-to-date ppa's for lucid available.
But it is trivial to remove pulseaudio in a minute, just a matter of copying/pasting a few commands to the terminal. I have used the tutorial Elegantly Disabling PulseAudio in Ubuntu 10.04/10.10 | www.jeffsplace.net many times, always with 100% success. Even a beginner can acomplish that easily.
I would not recommend using the betas between LTS releases either. Their lifespan is miniscule, upgrading painful, changes too deep. LTS 10.04 has gradually become a fine distribution, with the major bugs already fixed or reasonable workarounds known. There are lots of useful and up-to-date ppa's for lucid available.
1.
I btw had this "elegant" way to running for quite a long period of time.
It's not what I call elegant. You break Gnome packages.
And you need to rely on that ppa and related packages to fix these.
I don't like that.
Elegant would be just a button in a menu: "Pulseaudio On/OFF"
On one hand you recommend LTS releases on the other hand you recommend any kind of hacker ppa's. Hmmh.
2.
LTS releases are usually not up2date. Those are much behind newest developments.
Look at the new 2.6.38 kernel, 1.0.24 alsa asf. asf.
Linux is always late - much too late - with drivers in particular.
And this is exactly the reason why rolling releases (Distros) become more and more popular. Afaik there are plans to prepare a rolling release version
for Ubuntu.
For those who havn't heard about rolling releases: With a rolling release there is no need for major upgrades anymore. Packages are continuously upgraded on the fly.
The rolling releases can work since it seems that Linux got a lot more stable than it used to be. Ubuntu even seems to do a better job then good old Debian.
My current Ubuntu Natty Beta proves that things run quite stable even at such an early beta phase. Since I installed it I'havn't had much problems.
I even enjoy it.
The only problem is the broken Alsa-Tools package (HDSP mixer) , which is not an Ubuntu problem!
Cheers
Liquorix Kernel
Hi folks.
Yesterday I tried the Liquorix Kernel.
It's based on 2.6.38.
The kernel had been suggested over here some time back.
Just run below 5 commands and reboot into the new kernel and you're set.
I can tell you it made a noticeable improvement on my system. Give it a try.
Enjoy.
Hi folks.
Yesterday I tried the Liquorix Kernel.
It's based on 2.6.38.
The kernel had been suggested over here some time back.
Just run below 5 commands and reboot into the new kernel and you're set.
Code:
sudo su
echo "deb http://liquorix.net/debian sid main" >> /etc/apt/sources.list
apt-get install '^liquorix-([^-]+-)?keyring.?'
apt-get update
apt-get install a*2.6-liquorix-686I can tell you it made a noticeable improvement on my system. Give it a try.
Enjoy.
Last edited:
Creating a fake pulseaudio package is no major problem. I have never encountered any trouble with it. It is how the packaging works.
I have no problem with it. The LTS will not be upgrading to newer pulseaudio package with incompatible features, therefore the fake package will not need any upgrade.
You have all the power to implement it, if you desire it. I am fine with the above procedure. Ubuntu is a desktop distribution and as such Canonical will not put resources into disabling pulseaudio which is getting better each version.
On one hand you recommend LTS releases on the other hand you recommend any kind of hacker ppa's. Hmmh.
I recommend what I find the best option at this stage of matters.
And it never will. You either want stability and no hassle, or "newest development"
Then use something else, there are numerous options.
I wonder which rolling releases you have used actively and for how long. I do not mean installing and wiping the next day/week as is usually the case for reviewers and distro hoppers. I do not know ANY rolling release distro I would recommend for production use. Arch being most popular is still too much edgy. Just try running it for a few months and use it on a daily basis. I did. It is a wonderful distro for developers and linux enthusiasts though.
For me the hassle-free upgrade to new version is basically equivalent to rolling-release - I never have to reinstall and do the configuration again. And that is the case with our servers running debian stable - we have not reinstalled for over 8 years, just dist-upgrades (while serving) and continuous HW upgrade/replacement. For the desktop I have come back to ubuntu LTS which I find a good compromise between stability and novelty. We will see what they do with 12.04, whether they keep the classical gnome desktop and Xorg. Most probably not, ubuntu is purposefully heading away from the classical desktop to touchscreen devices (unity, wayland). And I understand, that's where the only money is, in the app stores mobile users are accustomed to pay for. That is another reason why I do not recommend the new betas after 10.04.
These recommendations last for weeks, maybe months. Most people do not have the time and patience to keep reinstalling/upgrading/solving upgrade problems which inevitably happen unless the upgrade process is thouroughly tested and perfected. And that takes A LOT of work and time.
But I still believe there is a market opportunity for corporate desktop linux distribution. I hoped Canonical would fill the niche, apparently they are moving away to consumer space.
I believe companies would be willing to pay reasonable yearly fees for a reasonably modern linux desktop with well tested upgrade route. At least mine would. Something like the debian desktop/debian CUT, but with a commercial entity behind offering paid support and proper care of packaging (i.e. LOTS and LOTS of testing, perhaps developing the much needed state-of-the-art automated package testing infrastructure). Perhaps RedHat will eventually expand seriously into corporate desktop with viable pricing.
I believe companies would be willing to pay reasonable yearly fees for a reasonably modern linux desktop with well tested upgrade route. At least mine would. Something like the debian desktop/debian CUT, but with a commercial entity behind offering paid support and proper care of packaging (i.e. LOTS and LOTS of testing, perhaps developing the much needed state-of-the-art automated package testing infrastructure). Perhaps RedHat will eventually expand seriously into corporate desktop with viable pricing.
No company - except Google - is gonna manage to make Linux available to a wider consumer market. It's the money which makes things moving.
Not some idealistic Linux hackers. Others will for sure piggyback on that.
But even with Google in the lead you see 100s of different variants popping up. It's a known fact that Google has to get this under control.
Corporate Distros will be used as special purpose animals. Linux as a Desktop system is IMO unacceptable in professional Desktop environments.
You can reduce testing efforts and increase quality if somebody would take the overall integration lead, if you'd get structured, if you'd avoid redundant work, if you'd have a common goal.
And -- you can't count on support by hobbyists, if you take all this erious. Linux would be dead by now without long-term industrial cross-financing. The active fulltimers are paid by the industry.
If all the wasted design time and efforts in Linux land world could be used in a productive manner, not any other OS would be able to compete.
Example:
1. Ubuntu and Debian should better merge. That would avoid lot's of redundant work and would increase quality.
2. The audio layer should become integrated alsa/oss/pulseaudio/jack/gstreamer/xine ..... all in one hand. On top of that proprietary audio driver handling should be introduced to get the manufacturers on board. And somebody should start working on an OSX compatibility layer. In this case OSX drivers could be used.
Last edited:
actually, I don't see why bothering with all that. At least on 10.10 (but IIRC also on 10.04) you can simply purge pulseaudio altogether!
Code:
root@lucid:~# aptitude purge pulseaudio
Reading package lists... Done
Building dependency tree
Reading state information... Done
Reading extended state information
Initializing package states... Done
The following packages are BROKEN:
indicator-sound libcanberra-pulse pulseaudio-esound-compat pulseaudio-module-bluetooth pulseaudio-module-gconf
pulseaudio-module-x11 ubuntu-desktop
The following packages will be REMOVED:
pulseaudio{p}
0 packages upgraded, 0 newly installed, 1 to remove and 0 not upgraded.
Need to get 0B of archives. After unpacking 4,567kB will be freed.
The following packages have unmet dependencies:
ubuntu-desktop: Depends: pulseaudio but it is not installable
pulseaudio-module-bluetooth: Depends: pulseaudio but it is not installable
pulseaudio-module-x11: Depends: pulseaudio but it is not installable
pulseaudio-module-gconf: Depends: pulseaudio but it is not installable
indicator-sound: Depends: pulseaudio but it is not installable
libcanberra-pulse: Depends: pulseaudio but it is not installable
pulseaudio-esound-compat: Depends: pulseaudio but it is not installable
The following actions will resolve these dependencies:
Remove the following packages:
indicator-sound
libcanberra-pulse
pulseaudio-esound-compat
pulseaudio-module-bluetooth
pulseaudio-module-gconf
pulseaudio-module-x11
ubuntu-desktop
Leave the following dependencies unresolved:
gnome-session-canberra recommends libcanberra-pulse
gnome-settings-daemon recommends pulseaudio
indicator-applet recommends indicator-sound
Score is -2223
Accept this solution? [Y/n/q/?]answer yes and pulseaudio will be gone for good ("ubuntu-desktop" is just a metapackage which is used to "pull in" other packages; removing it does not change anything on your system). Then you only need to run:
gstreamer-properties
to tell gstreamer to use ALSA directly instead of PA and install/use some different mixer. I personally like best the text mode alsamixer, but you may want some GUI based one.
I hope nobody with the decision-making power will share this idea. Merging alsa/pulseaudio/jack is ok, but gstreamer/xine are at least one level above and do not belong to the low-level libraries.
Well, good luck with that.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.