Linux public beta (4408)

Official support for: u-he.com
Locked New Topic
RELATED
PRODUCTS

Post

Lump wrote:Wrong. I'm a very experienced old linux user, young padawan.
You are a current, active Linux user and musician who are not aware of any of the multimedia distros and still thinks he needs to compile his kernels to get anywhere? I am sorry but I find that hard to believe. But let's leave it at that, it's beside the point. I'm just so tired of FUD being spread on Linux being so complicated and hard to use that you even have to start compiling your stuff in order for it to run properly. That's just not the case - it hardly ever was, and definitely not now.
Lump wrote:Lot of glitches and crackles (with ALSA - BWS or Renoise) and thousands of xruns if I use JACK. It is no fun, it is a cramp.
My system is a i7-3820 w/16gb ram, not too different from yours it seems. I'm running Bitwig on Ubuntu 15.04 with a lowlatency kernel and a Focusrite 8i6 audio card.

Jack can be a bit tweaky to set up at first. I'll run through the essentials from my own setup:

- Run it with the parameter -S (for syncronous) in the server path field (ergo "jackd -S" instead of just "jackd") for a lot fewer Xruns in Jack2.

- Turn on "realtime".
- I use sample rate 48khz with a 256 buffer size (frames/period field), but optimal settings here heavily depends on your audio card. NB! Notice that it's essential that you set the buffer size in Bitwig audio setup to the same as in Jack.

If you still experience cracle and snaps after tweaking these settings (increase buffer and/or lower sample rate and they should go away) I'd see if there's any known issues with your audio interface on Linux.

Cause really, had what you describes been the standard on Linux you'd not find any of us other Linux musicians around here. We'd do our business on another OS, like we had to earlier.
Last edited by Beamboom on Tue Jun 02, 2015 12:38 pm, edited 2 times in total.

Post

I interpreted Lump's comment about kernel compiling,
as a bit of humorous sarcasm. Someone unaware of his history
in the linux audio community, might have have missed that. :wink:

Post

glokraw wrote:I interpreted Lump's comment about kernel compiling,
as a bit of humorous sarcasm. Someone unaware of his history
in the linux audio community, might have have missed that. :wink:

Yes in 2005, when I started the jacklab, there was no other chance then "roll your own" to get an acceptable audio performance. But still in 2015 someone gave me the serious advice, here at kvraudio, to use "hard realtime"

http://www.kvraudio.com/forum/viewtopic ... 4#p6076675
[del]AudioLinux sucks.[/del]

Post

Any chance of releasing Hive 1.0 on linux?

Post

I think the developers mentioned that it will ve released in 2nd of June. The Offizials Release Date.

Post

Lump wrote:But still in 2015 someone gave me the serious advice, here at kvraudio, to use "hard realtime"
First off, Lump, please accept my apologies for being so snappy at you. I really am frustrated about much of what I'd classify as FUD regarding this OS, but it's always wrong of me to let that vent out on anyone in particular, especially when I indeed have to do with a fellow Linux user. So I'm very sorry about that.

Now, to the quote above:
If there's one thing about the Internet I hate the most, it's how hard it is to separate misinformation from good advice. Either the info is not dated (and turns out to be very outdated), the information is per say correct but consist of a complex string of manual actions when in fact the right way is to use a tool in the distro to do that job for you (or even worse, is solved differently in your distro), or it just plain doesn't work and is based on misconceptions about the topic at hand. All presented with the same hardheaded "know-all" deliverance.

But there's one advice in that thread you posted there that I really find to be very good, it's at the end of page one:
The OP might also change the sample rate and period settings for the device. 48000 with a period of 3 works better with some USB audio devices
The core of this advise is that optimal solution really depends on your particular hardware setup. I experienced the same with my setup: At first the cpu-heavy U-He synths didn't run well despite the Jack settings being what was claimed to be the "best" configs. It was only after I closed the browser and methodically tweaked the settings and testing on my own I found my "sweet spot".

But yeah... I'm sorry. :)
Last edited by Beamboom on Tue Jun 02, 2015 12:35 pm, edited 2 times in total.

Post

Greetings,

I'm not complaining, but I was wondering why there's no mention of Linux availability on the official u-he site ? I've had so few problems with the u-he plugins, it's hard for me to think of them as beta releases, so shouldn't Linux get some mention-love too ? :)

Amazing synths, thanks again to everyone making this happen for Linux.

Best,

dp

Post

StudioDave wrote:I'm not complaining, but I was wondering why there's no mention of Linux availability on the official u-he site ? I've had so few problems with the u-he plugins, it's hard for me to think of them as beta releases, so shouldn't Linux get some mention-love too ? :)
IIRC, Urs said in an interview at Musikmesse that the Linux versions will be still in beta for a while as they first have to learn things like what kind of support cases they will have from Linux users and how to react to the questions of them.

That might explain why they are introducing the Linux versions slowly until the final release.

Post

Hello, I think that there is a Linux version for every u-he synthesizer by now? ;-)

Post

abique wrote:Hello, I think that there is a Linux version for every u-he synthesizer by now? ;-)
Wonderful, that's great news !

@Chapelle : I'm sure you're right. Being careful about the targets is just good business, saves everyone headaches on down the line.

Best,

dp

Post

Newer Linux-builds of Hive and Presswerk are found elsewhere on the forums now:

Hive 1.0 (release): http://uhedownloads.heckmannaudiogmb.ne ... e10.tar.gz
Presswerk 1.1: http://uhedownloads.heckmannaudiogmb.ne ... k11.tar.gz

Update first post maybe, abique?

Post

Hi abique, Urs and co!

zebralette 3466 is very crashy for me under both Tracktion 6.1.9 and qtractor 0.6.7. I 'downgraded' to 3305 and the older build runs fine.

Arch Linux
Linux 4.0.4-2-ARCH x86_64
glibc 2.21 / gcc 5.1.0

Also, I'd very much appreciate if someone could answer my thread in the main Uhe forum which has gone unanswered for almost two months now.

http://www.kvraudio.com/forum/viewtopic ... 1&t=436430

Thanks!

Post

Lump wrote:Lot of glitches and crackles (with ALSA - BWS or Renoise) and thousands of xruns if I use JACK. It is no fun, it is a cramp.
 
I've had crackles as well with my USB Audio Interface (not with the onboard though). Jack2 and a realtime kernel helped. On Arch it is really easy to install and compile (from the AUR). You can install it like any other package.

Beamboom wrote: - Run it with the parameter -S (for syncronous) in the server path field (ergo "jackd -S" instead of just "jackd") for a lot fewer Xruns in Jack2.
Could you elaborate on this? The manpages say nothing about that synchronous mode.

Post

danboid wrote:Also, I'd very much appreciate if someone could answer my thread in the main Uhe forum which has gone unanswered for almost two months now.
Sorry man, I can't help you there.
nonameyet wrote:
Beamboom wrote: - Run it with the parameter -S (for syncronous) in the server path field (ergo "jackd -S" instead of just "jackd") for a lot fewer Xruns in Jack2.
Could you elaborate on this? The manpages say nothing about that synchronous mode.
I can do better than that :)
I can refer to the documentation on the jack2 git hub page regarding the -S parameter (you can also use "--sync"):

https://github.com/jackaudio/jack2

Quote:
"Depending of the used interface and driver settings, the PortAudio layer may add additional buffering between the real card interrupt and the jack server callback. This usually result in *unregular* calls of the jack server callback (for example if jack server used a 256 frames buffer and the card used a 512 frames, the jack server callback will be called twice every card interrupt). For proper functioning of jack server and clients in this case, the jack server has to be started in "synchronous" mode, using the "-S" parameter."

In my personal experience I've always had better results (or at the very least never worse results) when running in synchronous mode, so I just always do that nowadays. This forum post on Linuxmusicians.com confirms that others too have good results with running jack2 in synchronous mode.
http://www.linuxmusicians.com/viewtopic.php?f=27&t=2683

Post

Beamboom wrote: In my personal experience I've always had better results (or at the very least never worse results) when running in synchronous mode, so I just always do that nowadays. This forum post on Linuxmusicians.com confirms that others too have good results with running jack2 in synchronous mode.
http://www.linuxmusicians.com/viewtopic.php?f=27&t=2683
Alright I now tried the synchronous mode. I got an insane amount of XRuns which were not audible (crackles and so on) and every couple seconds I had no sound at all. Not sure what's different with my setup. I have just installed jack2 and QJackCtl from my distro repo. Do I have to use the GitHub version or are there any special settings?

EDIT:

I think I have to take this back. This apparently only happens with Windows VST plugins (WINE and Airwave). Native ones run without XRuns at very low latencies.
Last edited by nonameyet on Tue Jun 09, 2015 3:19 pm, edited 1 time in total.

Locked

Return to “u-he”