20 Oct
2014
20 Oct
'14
9:19 a.m.
Hi,
Is it possible to install JACK etc. and for instance Pianoteq on a device
like this, to create your own instrument module?
http://www.rikomagic.com/en/product/showpro_id_58_pid_22.html
Regards,
Dirk
20 Oct
20 Oct
9:32 a.m.
On 10/20/2014 11:18 AM, rosea grammostola wrote:
Hi,
Is it possible to install JACK etc. and for instance Pianoteq on a device
like this, to create your own instrument module?
http://www.rikomagic.com/en/product/showpro_id_58_pid_22.html
Regards,
Dirk
Hi Dirk,
It's possible but not on this specific device because of the Rockchip
chipset. I've done a great deal of getting JACK and all to run on
Rockchip based devices but it turned out to be a bit of a
disappointment, I couldn't get it to work in a stable manner because of
the way the peripherals are interconnected. You might want to try a
device based on an Allwinner chipset (A2x or A3x), an Allwinner chipset
is also used as the base of the MOD Duo.
Pianoteq won't run though, AFAIK there are no ARM binaries available.
Jeremy
9:34 a.m.
On 10/20/2014 11:18 AM, rosea grammostola wrote:
Hi,
Is it possible to install JACK etc. and for instance Pianoteq on a device
like this, to create your own instrument module?
Dunno about jack, but it seems interesting! However I can't seem to find
any prices...
--
Atte
http://atte.dk http://a773.dk
11:59 a.m.
On 10/20/2014 11:34 AM, Atte wrote:
Dunno about jack, but it seems interesting! However I
can't seem to find
any prices...
These devices are fairly cheap starting from about $50:
http://www.dx.com/p/ditter-u20-dual-core-android-4-2-google-tv-hd-player-w-…
That's one of the cheapest I could find with an Allwinner A20 chipset.
Better would be to get something like a Cubieboard.
Jeremy
2:47 p.m.
On Mon, 20 Oct 2014, Jeremy Jongepier wrote:
On 10/20/2014 11:34 AM, Atte wrote:
Atom MB (very small) with 12v power can be had for ~$100 (I have seen them
as low as $69) And run anything any other intel chip will run. They do
have a 16bit audio out/in, though latency on jack is at best 64/3 (my wife
has one in her desktop. Video chipset is not real useful for gaming or
videos, but fine for desktop use (and not relevent for headless). There
are serial ports, so there is a possibility of having native MIDI porting
with a little HW/SW glue. Some of these do not have the 12v input but will
run from 12v with a very small adaptor. An old laptop ps could fit in the
box meaning no lump in the power cord.
Some of them also have a par port... which could be used for switches on
the box... but it would be easier to pull the board from a cheap ($3 or
so) keyboard and wire switches to that.
--
Len Ovens
www.ovenwerks.net
Dunno about jack, but it seems interesting!
However I can't seem to find
any prices...
These devices are fairly cheap starting from about $50:
http://www.dx.com/p/ditter-u20-dual-core-android-4-2-google-tv-hd-player-w-…
That's one of the cheapest I could find with an Allwinner A20 chipset.
Better would be to get something like a Cubieboard. 
3:10 p.m.
On Mon, October 20, 2014 10:59 pm, Jeremy Jongepier wrote:
On 10/20/2014 11:34 AM, Atte wrote:
If you want something that is relatively cheap that can run pianoteq,
linuxsampler, etc... then you should look into the Intel NUC range.
They provide a very nice piece of gear for a stand alone solution and
decent responsive support via the OTC forums. I requested a change to the
BIOS and they did it for me as an update.
I have had one for several months as a head for my home entertainment
center and it is working very well. Yesterday I did some edits with
ardour2. All I had to do was "apt-get install ardour".
It's not $50 but it is less than $300 for a complete 64 bit x86 solution
including RAM/HDD/SSD. The case is brushed aluminium and it is very
robust. If you get an SSD it is nearly 100% solid state apart from the CPU
fan which is almost noiseless.
My (cheap) home stereo speaker system generates more noise than the fan
and I have it running 24/7 for net streaming to TV with full 1080p over
hdmi.
Saves all the hassle of learning how to cross compile and the general
issues associated with developing for ARM.
I tried several other options before settling on this one. They were
cheaper but also more frustrating.
Intel are catching up on ARM these days with the 16/14/12nm fab process.
Dunno about jack, but it seems interesting!
However I can't seem to find
any prices...
These devices are fairly cheap starting from about $50:
http://www.dx.com/p/ditter-u20-dual-core-android-4-2-google-tv-hd-player-w-…
That's one of the cheapest I could find with an
Allwinner A20 chipset.
Better would be to get something like a Cubieboard.
--
Patrick Shirkey
Boost Hardware Ltd
4:04 p.m.
On Tue, 2014-10-21 at 02:10 +1100, Patrick Shirkey wrote:
My (cheap) home stereo speaker system generates more
noise than the
fan
I dislike fans. I believe that even a more expensive home audio system
could cause more noise, than a brand new fan, but fans suffer from
time ;). Unfortunately small noiseless fans usually become as loud as a
sonic bang after a while. IMO experiences just big fans of power
supplies last for a very long time (knocking on wood ;). YMMV!
4:22 p.m.
PS: To be fair, HDDs nowadays are usually much louder than a fan, and
usually even those small fast fans last longer than a HDD. Anyway, if
possible avoiding a fan is better, than using a fan.
21 Oct
21 Oct
9:27 a.m.
2014-10-20 17:10 GMT+02:00 Patrick Shirkey <pshirkey(a)boosthardware.com>om>:
On Mon, October 20, 2014 10:59 pm, Jeremy Jongepier wrote:
If you want something that is relatively cheap that can run pianoteq,
linuxsampler, etc... then you should look into the Intel NUC range.
They provide a very nice piece of gear for a stand alone solution and
decent responsive support via the OTC forums. I requested a change to the
BIOS and they did it for me as an update.
I have had one for several months as a head for my home entertainment
center and it is working very well. Yesterday I did some edits with
ardour2. All I had to do was "apt-get install ardour".
It's not $50 but it is less than $300 for a complete 64 bit x86 solution
including RAM/HDD/SSD. The case is brushed aluminium and it is very
robust. If you get an SSD it is nearly 100% solid state apart from the CPU
fan which is almost noiseless.
My (cheap) home stereo speaker system generates more noise than the fan
and I have it running 24/7 for net streaming to TV with full 1080p over
hdmi.
Saves all the hassle of learning how to cross compile and the general
issues associated with developing for ARM.
I tried several other options before settling on this one. They were
cheaper but also more frustrating.
Intel are catching up on ARM these days with the 16/14/12nm fab process.
> That's one of the cheapest I could find with an Allwinner A20 chipset.
> Better would be to get something like a Cubieboard.
--
Patrick Shirkey
Boost Hardware Ltd
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user(a)lists.linuxaudio.org
http://lists.linuxaudio.org/listinfo/linux-audio-user
That about Intel NUC range sounds really interesting. Which device do you
have, Patrick? I guess the possibilities are one of these:
http://techpartner.intel.com/campaigns/nuc/
Apart from running Ardour, what about RT and IRQs and that stuff?
Thanks Patrick.
--
C. sanchiavedraZ:
* NEW / NUEVO: www.sanchiavedraZ.com
* Musix GNU+Linux: www.musix.es
On 10/20/2014 11:34 AM, Atte wrote:
http://www.dx.com/p/ditter-u20-dual-core-android-4-2-google-tv-hd-player-w-…
Dunno about jack, but it seems interesting!
However I can't seem to find
any prices...
These devices are fairly cheap starting from about $50:
9:37 a.m.
On Tue, October 21, 2014 8:27 pm, Carlos sanchiavedraz wrote:
2014-10-20 17:10 GMT+02:00 Patrick Shirkey
<pshirkey(a)boosthardware.com>om>:
I got the cheapest one at the time which was the DNY2820FKH. It has a dual
core celeron and 4GB RAM. I think it can take up to 16GB.
On Mon, October 20, 2014 10:59 pm, Jeremy Jongepier wrote:
http://www.dx.com/p/ditter-u20-dual-core-android-4-2-google-tv-hd-player-w-…
If you want something that is relatively cheap that can run pianoteq,
linuxsampler, etc... then you should look into the Intel NUC range.
They provide a very nice piece of gear for a stand alone solution and
decent responsive support via the OTC forums. I requested a change to
the
BIOS and they did it for me as an update.
I have had one for several months as a head for my home entertainment
center and it is working very well. Yesterday I did some edits with
ardour2. All I had to do was "apt-get install ardour".
It's not $50 but it is less than $300 for a complete 64 bit x86 solution
including RAM/HDD/SSD. The case is brushed aluminium and it is very
robust. If you get an SSD it is nearly 100% solid state apart from the
CPU
fan which is almost noiseless.
My (cheap) home stereo speaker system generates more noise than the fan
and I have it running 24/7 for net streaming to TV with full 1080p over
hdmi.
Saves all the hassle of learning how to cross compile and the general
issues associated with developing for ARM.
I tried several other options before settling on this one. They were
cheaper but also more frustrating.
Intel are catching up on ARM these days with the 16/14/12nm fab process.
> That's one of the cheapest I could find with an Allwinner A20 chipset.
> Better would be to get something like a Cubieboard.
--
Patrick Shirkey
Boost Hardware Ltd
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user(a)lists.linuxaudio.org
http://lists.linuxaudio.org/listinfo/linux-audio-user
That about Intel NUC range sounds really interesting. Which device do you
have, Patrick? I guess the possibilities are one of these:
http://techpartner.intel.com/campaigns/nuc/
On 10/20/2014 11:34 AM, Atte wrote:
> Dunno about jack, but it seems interesting! However I can't seem to
find
any
prices...
These devices are fairly cheap starting from about $50:
Apart from running Ardour, what about RT and IRQs and
that stuff?
Thanks Patrick.
The rt config is in place but I haven't really pushed it for audio
generation yet. It runs electricsheep and full 1080p video/audio without
any issues. I can get about 30fps with 1080p which is decent for viewing
but not fast enough for competitive playing.
Here is cat /proc/interrupts
CPU0 CPU1
0: 26 0 IO-APIC-edge timer
1: 2 0 IO-APIC-edge i8042
8: 1 0 IO-APIC-fasteoi rtc0
9: 0 0 IO-APIC-fasteoi acpi
10: 10801372 0 IO-APIC-edge ite-cir
12: 4 0 IO-APIC-edge i8042
103: 1018511 0 PCI-MSI-edge xhci_hcd
104: 7418719 0 PCI-MSI-edge ahci
105: 113014233 0 PCI-MSI-edge eth0
106: 216595760 0 PCI-MSI-edge i915
107: 373 1604714 PCI-MSI-edge iwlwifi
108: 1368299 0 PCI-MSI-edge snd_hda_intel
NMI: 48085 47493 Non-maskable interrupts
LOC: 383351941 317871495 Local timer interrupts
SPU: 0 0 Spurious interrupts
PMI: 48085 47493 Performance monitoring interrupts
IWI: 4051038 6083943 IRQ work interrupts
RTR: 0 0 APIC ICR read retries
RES: 69587463 71022804 Rescheduling interrupts
CAL: 427 487 Function call interrupts
TLB: 1192097 1158476 TLB shootdowns
TRM: 0 0 Thermal event interrupts
THR: 0 0 Threshold APIC interrupts
MCE: 0 0 Machine check exceptions
MCP: 2904 2904 Machine check polls
ERR: 0
MIS: 0
--
Patrick Shirkey
Boost Hardware Ltd
9:41 a.m.
2014-10-21 11:37 GMT+02:00 Patrick Shirkey <pshirkey(a)boosthardware.com>om>:
On Tue, October 21, 2014 8:27 pm, Carlos sanchiavedraz wrote:
If you want something that is relatively cheap that can run pianoteq,
linuxsampler, etc... then you should look into the Intel NUC range.
They provide a very nice piece of gear for a stand alone solution and
decent responsive support via the OTC forums. I requested a change to
the
BIOS and they did it for me as an update.
I have had one for several months as a head for my home entertainment
center and it is working very well. Yesterday I did some edits with
ardour2. All I had to do was "apt-get install ardour".
It's not $50 but it is less than $300 for a complete 64 bit x86 solution
including RAM/HDD/SSD. The case is brushed aluminium and it is very
robust. If you get an SSD it is nearly 100% solid state apart from the
CPU
fan which is almost noiseless.
My (cheap) home stereo speaker system generates more noise than the fan
and I have it running 24/7 for net streaming to TV with full 1080p over
hdmi.
Saves all the hassle of learning how to cross compile and the general
issues associated with developing for ARM.
I tried several other options before settling on this one. They were
cheaper but also more frustrating.
Intel are catching up on ARM these days with the 16/14/12nm fab process.
> That's one of the cheapest I could find with an Allwinner A20 chipset.
> Better would be to get something like a Cubieboard.
--
Patrick Shirkey
Boost Hardware Ltd
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user(a)lists.linuxaudio.org
http://lists.linuxaudio.org/listinfo/linux-audio-user
That about Intel NUC range sounds really interesting. Which device do you
have, Patrick? I guess the possibilities are one of these:
http://techpartner.intel.com/campaigns/nuc/
I got the cheapest one at the time which was the DNY2820FKH. It has a dual
core celeron and 4GB RAM. I think it can take up to 16GB.
Thanks!
It'll be great if it runs Debian/Musix and if it could be configured for RT
audio.
--
C. sanchiavedraZ:
* NEW / NUEVO: www.sanchiavedraZ.com
* Musix GNU+Linux: www.musix.es
2014-10-20 17:10 GMT+02:00 Patrick Shirkey
<pshirkey(a)boosthardware.com>om>:
>
> On Mon, October 20, 2014 10:59 pm, Jeremy Jongepier wrote:
> > On 10/20/2014 11:34 AM, Atte wrote:
> >> Dunno about jack, but it seems interesting! However I can't seem to
> find
> >> any prices...
> >
> > These devices are fairly cheap starting from about $50:
> >
>
http://www.dx.com/p/ditter-u20-dual-core-android-4-2-google-tv-hd-player-w-…
Apart from running Ardour, what about RT and IRQs
and that stuff?
Thanks Patrick.
The rt config is in place but I haven't really pushed it for audio
generation yet. It runs electricsheep and full 1080p video/audio without
any issues. I can get about 30fps with 1080p which is decent for viewing
but not fast enough for competitive playing.
Here is cat /proc/interrupts
CPU0 CPU1
0: 26 0 IO-APIC-edge timer
1: 2 0 IO-APIC-edge i8042
8: 1 0 IO-APIC-fasteoi rtc0
9: 0 0 IO-APIC-fasteoi acpi
10: 10801372 0 IO-APIC-edge ite-cir
12: 4 0 IO-APIC-edge i8042
103: 1018511 0 PCI-MSI-edge xhci_hcd
104: 7418719 0 PCI-MSI-edge ahci
105: 113014233 0 PCI-MSI-edge eth0
106: 216595760 0 PCI-MSI-edge i915
107: 373 1604714 PCI-MSI-edge iwlwifi
108: 1368299 0 PCI-MSI-edge snd_hda_intel
NMI: 48085 47493 Non-maskable interrupts
LOC: 383351941 317871495 Local timer interrupts
SPU: 0 0 Spurious interrupts
PMI: 48085 47493 Performance monitoring interrupts
IWI: 4051038 6083943 IRQ work interrupts
RTR: 0 0 APIC ICR read retries
RES: 69587463 71022804 Rescheduling interrupts
CAL: 427 487 Function call interrupts
TLB: 1192097 1158476 TLB shootdowns
TRM: 0 0 Thermal event interrupts
THR: 0 0 Threshold APIC interrupts
MCE: 0 0 Machine check exceptions
MCP: 2904 2904 Machine check polls
ERR: 0
MIS: 0
--
Patrick Shirkey
Boost Hardware Ltd
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user(a)lists.linuxaudio.org
http://lists.linuxaudio.org/listinfo/linux-audio-user
10:14 a.m.
On Tue, October 21, 2014 8:41 pm, Carlos sanchiavedraz wrote:
2014-10-21 11:37 GMT+02:00 Patrick Shirkey
<pshirkey(a)boosthardware.com>om>:
I have installed Debian but there were issues with the wifi chipset so I
switched to ubuntu which has better support for wifi ootb. Simply because
I didn't want to spend any more time on wifi issues at that point having
just ditched a rockchip device with similar problems. Now I don't use wifi
so I could switch back to debian but seeing as everything is already setup
there is not much motivation.
The realtime config is in place when I installed jack/ardour2. I have not
tested the actual best case latency with jack yet. It works fine with -p
1024 -r48000 -n2. Maybe someone else has latency details on the audio
chipset with other systems? (snd-hda-intel)
I'll try to run some tests when my daughter is not watching teletubbies.
--
Patrick Shirkey
Boost Hardware Ltd
On Tue, October 21, 2014 8:27 pm, Carlos sanchiavedraz wrote:
you
Thanks!
It'll be great if it runs Debian/Musix and if it could be configured for
RT
audio.
2014-10-20 17:10 GMT+02:00 Patrick Shirkey
<pshirkey(a)boosthardware.com>om>:
>
> On Mon, October 20, 2014 10:59 pm, Jeremy Jongepier wrote:
> > On 10/20/2014 11:34 AM, Atte wrote:
> >> Dunno about jack, but it seems interesting! However I can't seem
to
> find
> >> any prices...
> >
> > These devices are fairly cheap starting from about $50:
> >
>
http://www.dx.com/p/ditter-u20-dual-core-android-4-2-google-tv-hd-player-w-…
> >
>
> If you want something that is relatively cheap that can run pianoteq,
> linuxsampler, etc... then you should look into the Intel NUC range.
>
> They provide a very nice piece of gear for a stand alone solution and
> decent responsive support via the OTC forums. I requested a change to
> the
> BIOS and they did it for me as an update.
>
> I have had one for several months as a head for my home entertainment
> center and it is working very well. Yesterday I did some edits with
> ardour2. All I had to do was "apt-get install ardour".
>
> It's not $50 but it is less than $300 for a complete 64 bit x86
solution
> including RAM/HDD/SSD. The case is brushed
aluminium and it is very
> robust. If you get an SSD it is nearly 100% solid state apart from
the
> CPU
> fan which is almost noiseless.
>
> My (cheap) home stereo speaker system generates more noise than the
fan
> and I have it running 24/7 for net streaming
to TV with full 1080p
over
> hdmi.
>
> Saves all the hassle of learning how to cross compile and the general
> issues associated with developing for ARM.
>
> I tried several other options before settling on this one. They were
> cheaper but also more frustrating.
>
> Intel are catching up on ARM these days with the 16/14/12nm fab
process.
>
> > That's one of the cheapest I could find with an Allwinner A20
chipset.
>
Better would be to get something like a Cubieboard.
>
--
Patrick Shirkey
Boost Hardware Ltd
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user(a)lists.linuxaudio.org
http://lists.linuxaudio.org/listinfo/linux-audio-user
That about Intel NUC range sounds really interesting. Which device do have, Patrick? I guess the possibilities are one
of these:
http://techpartner.intel.com/campaigns/nuc/
I got the cheapest one at the time which was the DNY2820FKH. It has a
dual
core celeron and 4GB RAM. I think it can take up to 16GB.
Apart from running Ardour, what about RT and IRQs
and that stuff?
Thanks Patrick.
The rt config is in place but I haven't really pushed it for audio
generation yet. It runs electricsheep and full 1080p video/audio without
any issues. I can get about 30fps with 1080p which is decent for viewing
but not fast enough for competitive playing.
Here is cat /proc/interrupts
CPU0 CPU1
0: 26 0 IO-APIC-edge timer
1: 2 0 IO-APIC-edge i8042
8: 1 0 IO-APIC-fasteoi rtc0
9: 0 0 IO-APIC-fasteoi acpi
10: 10801372 0 IO-APIC-edge ite-cir
12: 4 0 IO-APIC-edge i8042
103: 1018511 0 PCI-MSI-edge xhci_hcd
104: 7418719 0 PCI-MSI-edge ahci
105: 113014233 0 PCI-MSI-edge eth0
106: 216595760 0 PCI-MSI-edge i915
107: 373 1604714 PCI-MSI-edge iwlwifi
108: 1368299 0 PCI-MSI-edge snd_hda_intel
NMI: 48085 47493 Non-maskable interrupts
LOC: 383351941 317871495 Local timer interrupts
SPU: 0 0 Spurious interrupts
PMI: 48085 47493 Performance monitoring interrupts
IWI: 4051038 6083943 IRQ work interrupts
RTR: 0 0 APIC ICR read retries
RES: 69587463 71022804 Rescheduling interrupts
CAL: 427 487 Function call interrupts
TLB: 1192097 1158476 TLB shootdowns
TRM: 0 0 Thermal event interrupts
THR: 0 0 Threshold APIC interrupts
MCE: 0 0 Machine check exceptions
MCP: 2904 2904 Machine check polls
ERR: 0
MIS: 0
--
Patrick Shirkey
Boost Hardware Ltd
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user(a)lists.linuxaudio.org
http://lists.linuxaudio.org/listinfo/linux-audio-user
10:28 a.m.
Hey all,
I have a HDA ( cat /proc/interrupts tells me so anyway: snd_hda_intel ).
Its my built-in laptop soundcard, and it works pretty well, down to about
44.1kHz, -p128 -n3. That's ~8ms, which is acceptable IMO.
I've read on mailing list / internet somewhere* that spec for the HDA is
pretty open-to-interpretation, so I think the exact hardware / chipset
would need to be tested to get true results.
-Harry
*and lost the link..
On Tue, Oct 21, 2014 at 11:14 AM, Patrick Shirkey <
pshirkey(a)boosthardware.com> wrote:
On Tue, October 21, 2014 8:41 pm, Carlos sanchiavedraz wrote:
Thanks!
It'll be great if it runs Debian/Musix and if it could be configured for
RT
audio.
I have installed Debian but there were issues with the wifi chipset so I
switched to ubuntu which has better support for wifi ootb. Simply because
I didn't want to spend any more time on wifi issues at that point having
just ditched a rockchip device with similar problems. Now I don't use wifi
so I could switch back to debian but seeing as everything is already setup
there is not much motivation.
The realtime config is in place when I installed jack/ardour2. I have not
tested the actual best case latency with jack yet. It works fine with -p
1024 -r48000 -n2. Maybe someone else has latency details on the audio
chipset with other systems? (snd-hda-intel)
I'll try to run some tests when my daughter is not watching teletubbies.
--
Patrick Shirkey
Boost Hardware Ltd
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user(a)lists.linuxaudio.org
http://lists.linuxaudio.org/listinfo/linux-audio-user
--
http://www.openavproductions.com
2014-10-21 11:37 GMT+02:00 Patrick Shirkey
<pshirkey(a)boosthardware.com>om>:
>
> On Tue, October 21, 2014 8:27 pm, Carlos sanchiavedraz wrote:
> > 2014-10-20 17:10 GMT+02:00 Patrick Shirkey
> <pshirkey(a)boosthardware.com>om>:
> >
> >>
> >> On Mon, October 20, 2014 10:59 pm, Jeremy Jongepier wrote:
> >> > On 10/20/2014 11:34 AM, Atte wrote:
> >> >> Dunno about jack, but it seems interesting! However I can't
seem
> to
> >> find
> >> >> any prices...
> >> >
> >> > These devices are fairly cheap starting from about $50:
> >> >
> >>
>
http://www.dx.com/p/ditter-u20-dual-core-android-4-2-google-tv-hd-player-w-…
> >
>
> If you want something that is relatively cheap that can run pianoteq,
> linuxsampler, etc... then you should look into the Intel NUC range.
>
> They provide a very nice piece of gear for a stand alone solution and
> decent responsive support via the OTC forums. I requested a change to
> the
> BIOS and they did it for me as an update.
>
> I have had one for several months as a head for my home entertainment
> center and it is working very well. Yesterday I did some edits with
> ardour2. All I had to do was "apt-get install ardour".
>
> It's not $50 but it is less than $300 for a complete 64 bit x86
solution
> including RAM/HDD/SSD. The case is brushed
aluminium and it is very
> robust. If you get an SSD it is nearly 100% solid state apart from
the
> CPU
> fan which is almost noiseless.
>
> My (cheap) home stereo speaker system generates more noise than the
fan
> and I have it running 24/7 for net streaming
to TV with full 1080p
over
> hdmi.
>
> Saves all the hassle of learning how to cross compile and the general
> issues associated with developing for ARM.
>
> I tried several other options before settling on this one. They were
> cheaper but also more frustrating.
>
> Intel are catching up on ARM these days with the 16/14/12nm fab
process.
>
> > That's one of the cheapest I could find with an Allwinner A20
chipset.
> > Better would be to get something like a
Cubieboard.
> >
>
> --
> Patrick Shirkey
> Boost Hardware Ltd
> _______________________________________________
> Linux-audio-user mailing list
> Linux-audio-user(a)lists.linuxaudio.org
> http://lists.linuxaudio.org/listinfo/linux-audio-user
>
That about Intel NUC range sounds really interesting. Which device do
you
have, Patrick? I guess the possibilities are one
of these:
http://techpartner.intel.com/campaigns/nuc/
I got the cheapest one at the time which was the DNY2820FKH. It has a
dual
core celeron and 4GB RAM. I think it can take up to 16GB.
Apart from running Ardour, what about RT and IRQs
and that stuff?
Thanks Patrick.
The rt config is in place but I haven't really pushed it for audio
generation yet. It runs electricsheep and full 1080p video/audio without
any issues. I can get about 30fps with 1080p which is decent for viewing
but not fast enough for competitive playing.
Here is cat /proc/interrupts
CPU0 CPU1
0: 26 0 IO-APIC-edge timer
1: 2 0 IO-APIC-edge i8042
8: 1 0 IO-APIC-fasteoi rtc0
9: 0 0 IO-APIC-fasteoi acpi
10: 10801372 0 IO-APIC-edge ite-cir
12: 4 0 IO-APIC-edge i8042
103: 1018511 0 PCI-MSI-edge xhci_hcd
104: 7418719 0 PCI-MSI-edge ahci
105: 113014233 0 PCI-MSI-edge eth0
106: 216595760 0 PCI-MSI-edge i915
107: 373 1604714 PCI-MSI-edge iwlwifi
108: 1368299 0 PCI-MSI-edge snd_hda_intel
NMI: 48085 47493 Non-maskable interrupts
LOC: 383351941 317871495 Local timer interrupts
SPU: 0 0 Spurious interrupts
PMI: 48085 47493 Performance monitoring interrupts
IWI: 4051038 6083943 IRQ work interrupts
RTR: 0 0 APIC ICR read retries
RES: 69587463 71022804 Rescheduling interrupts
CAL: 427 487 Function call interrupts
TLB: 1192097 1158476 TLB shootdowns
TRM: 0 0 Thermal event interrupts
THR: 0 0 Threshold APIC interrupts
MCE: 0 0 Machine check exceptions
MCP: 2904 2904 Machine check polls
ERR: 0
MIS: 0
--
Patrick Shirkey
Boost Hardware Ltd
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user(a)lists.linuxaudio.org
http://lists.linuxaudio.org/listinfo/linux-audio-user
8:17 p.m.
On Tue, 21 Oct 2014, Harry van Haaren wrote:
Hey all,
I have a HDA ( cat /proc/interrupts tells me so anyway: snd_hda_intel ).
Its my built-in laptop soundcard, and it works pretty well, down to about 44.1kHz, -p128
-n3. That's ~8ms, which is acceptable IMO.
I've read on mailing list / internet somewhere* that spec for the HDA is pretty
open-to-interpretation, so I think the exact hardware / chipset would need to be tested
to get true results.
I have gotten 48k -p64 -n3 on mine. Jack will crash if I try -p64 -n2, but
-p128 -n2 works.
From my reading, HDA is (like AC97) more of a bus spec
than a sound spec.
It seems to deal with connecting the sound from the HW to the
internal
bus. The HDA bus runs at 48K (like AC97) but does not expect the HW to be
48k (as AC97 did) and has methods for transporting audio with other sample
rates than 48K.
In my opinion, 128/2 is just on the edge of useful for live (guitarix for
example) use. 256/2 is annoying already. (No I do not play play really
fast or anything like that)
--
Len Ovens
www.ovenwerks.net
8:39 p.m.
~ Russell
On Oct 21, 2014, at 1:15 PM, Len Ovens
<len(a)ovenwerks.net> wrote:
I get about the same on the same hardware. Using KxStudio, I can occasionally get 64/2.
Unstable for any kind of serious live or recording activity. With - rt kernel can do a
little better.
Agreed... Playing guitar / bass / synth with anything less than 128/2 is not really usable
due to its anoying-ness factor for me personally. The 2nd you slap the headphones on
anyone else, usually the very first comment is 'um, there's a delay?!'.
R
On Tue, 21 Oct 2014, Harry van Haaren wrote:
Hey all,
I have a HDA ( cat /proc/interrupts tells me so anyway: snd_hda_intel ).
Its my built-in laptop soundcard, and it works pretty well, down to about 44.1kHz, -p128
-n3. That's ~8ms, which is acceptable IMO.
I've read on mailing list / internet somewhere* that spec for the HDA is pretty
open-to-interpretation, so I think the exact hardware / chipset would need to be tested
to get true results.
I have gotten 48k -p64 -n3 on mine. Jack will crash if I try -p64 -n2, but -p128 -n2
works.
From my reading, HDA is (like AC97) more of a bus spec than a sound spec. It seems to
deal with connecting the sound from the HW to the internal bus. The HDA bus runs at 48K
(like AC97) but does not expect the HW to be 48k (as AC97 did) and has methods for
transporting audio with other sample rates than 48K.
In my opinion, 128/2 is just on the edge of useful for live (guitarix for example) use.
256/2 is annoying already. (No I do not play play really fast or anything like 
22 Oct
22 Oct
1:59 a.m.
On 22/10/14 04:15, Len Ovens wrote:
On Tue, 21 Oct 2014, Harry van Haaren wrote:
This is also what I was lead to believe. snd_hda_intel
covers a fair
range of different chips, each of which has its own characteristics and
so there is no single answer to the original question. Pretty sure both
my last two laptops were reported as snd_hda_intel and know they were
definitely different sound chips! (Although offhand couldn't tell you
what...)
"Like AC'97, HD Audio is a specification that defines the architecture,
link frame format, and programming interfaces used by the controller on
the PCI bus <http://en.wikipedia.org/wiki/Conventional_PCI> and by the
codec on the other side of the link. Implementations of the host
controller are available from at least Intel
<http://en.wikipedia.org/wiki/Intel>, Nvidia
<http://en.wikipedia.org/wiki/Nvidia>, and AMD
<http://en.wikipedia.org/wiki/AMD>."
<http://en.wikipedia.org/wiki/Intel_High_Definition_Audio#cite_note-sb600-7>
[1]
Hey all,
I have a HDA ( cat /proc/interrupts tells me so anyway: snd_hda_intel ).
Its my built-in laptop soundcard, and it works pretty well, down to
about 44.1kHz, -p128
-n3. That's ~8ms, which is acceptable IMO.
I've read on mailing list / internet somewhere* that spec for the HDA
is pretty
open-to-interpretation, so I think the exact hardware / chipset would
need to be tested
to get true results.
I have gotten 48k -p64 -n3 on mine. Jack will crash if
I try -p64 -n2,
but -p128 -n2 works.
From my reading, HDA is (like AC97) more of a bus spec than a sound
spec. It seems to deal with connecting the sound from the HW to the
internal bus. The HDA bus runs at 48K (like AC97) but does not expect
the HW to be 48k (as AC97 did) and has methods for transporting audio
with other sample rates than 48K.
In my opinion, 128/2 is just on the edge of useful for live (guitarix
for example) use. 256/2 is annoying already. (No I do not play play
really fast or anything like that)
With no careful tuning I run my internal at 48k -p128 -n3 and get rare
xruns, mostly if I have my wifi turned on but still the very occasional
one otherwise.
Not sure if there is a mistake on the page or if my laptop actually has
both of these for something?? [2]
Intel 7 Series/C210 Series Chipset Family High Definition Audio
Controller
<http://www.ubuntu.com/certification/catalog/component/pci/8086%3A1e20/>
Intel Panther Point High Definition Audio Controller
<http://www.ubuntu.com/certification/catalog/component/pci/8086%3A1e20/>
[1] http://en.wikipedia.org/wiki/Intel_High_Definition_Audio
[2] http://www.ubuntu.com/certification/hardware/201206-11273/components/
3:38 a.m.
On Wed, 22 Oct 2014, Kazakore wrote:
This is also what I was lead to believe. snd_hda_intel
covers a fair range of different
chips, each of which has its own characteristics and so there is no single answer to the
original question. Pretty sure both my last two laptops were reported as snd_hda_intel
and know they were definitely different sound chips! (Although offhand couldn't tell
you
what...)
"Like AC'97, HD Audio is a specification that defines the architecture, link
frame
format, and programming interfaces used by the controller on the PCI bus and by the
codec on the other side of the link. Implementations of the host controller are
available from at least Intel, Nvidia, and AMD." [1]
Ok, that sounds prettier than what I said :) the reason I called it a bus
though is that one HDA interface seems to be able to handle a number of
physical audio devices. Each of which can have a different bit depth and
sample rate at the same time (say what?) so that the user could have a
telephone handset hooked up to the interface at 32k/8bits while listening
to/watching a high definition video. I don't know that Linux can handle
this though. (or windows for that matter) I am just repeating some of the
stuff I remember from reading though some of the very long and involved
Intel Doc.
Not sure if there is a mistake on the page or if my
laptop actually has both of these
for something?? [2]
Intel 7 Series/C210 Series Chipset Family High Definition Audio Controller
Intel Panther Point High Definition Audio Controller
It may have both on the same PCI IF even. cat /proc/interupts should tell
I think. As a note. On my netbook, the audio IF can run at a number of
sample rates with no problems, but my internal mic which shows up on the
same ALSA device, can only run at 48k. That is the ADC for the internal
mic only runs at 48k and if the rest of the internal audio (internal
speakers, line in/out plugs, are set to 44.1k, those i/os will be fine,
but the internal mic will have a periodic click as the 44.1 and 48k sample
rates align and cycle past each other. The click goes away if the AI is
set to 48K.
Be sure to look at the variety of HDA IF types the ALSA driver handles
and all of the settings that can be given to the HDA driver. It is quite a
long list.
--
Len Ovens
www.ovenwerks.net
4:30 a.m.
On 22/10/14 11:36, Len Ovens wrote:
On Wed, 22 Oct 2014, Kazakore wrote:
I did a very small amount of research after my last post and it seems
the HDA chipset controller is often the same as the South Bridge. Only
my old P4 desktop it was the ICH6 I think, (or was that ICH5 and AC97??
can't remember now...) and on this laptop I guess the 7 relates to
something similar to the ICH7 but for laptops.
But yeah there is a ~250 page pdf on the v1 docs which can be found
online (in fact it's linked from the Wikipedia article I linked to
earlier.) Not read through it myself. But from the desctription I can
understand why you say Bus. But a bus wouldn't generally include codecs
and things, so it's more than just the bus really isn't it? Nearly a
whole protocol really...
This is also what I was lead to believe.
snd_hda_intel covers a fair
range of different
chips, each of which has its own characteristics and so there is no
single answer to the
original question. Pretty sure both my last two laptops were reported
as snd_hda_intel
and know they were definitely different sound chips! (Although
offhand couldn't tell you
what...)
"Like AC'97, HD Audio is a specification that defines the
architecture, link frame
format, and programming interfaces used by the controller on the PCI
bus and by the
codec on the other side of the link. Implementations of the host
controller are
available from at least Intel, Nvidia, and AMD." [1]
Ok, that sounds prettier than what I said :) the reason I called it a
bus though is that one HDA interface seems to be able to handle a
number of physical audio devices. Each of which can have a different
bit depth and sample rate at the same time (say what?) so that the
user could have a telephone handset hooked up to the interface at
32k/8bits while listening to/watching a high definition video. I don't
know that Linux can handle this though. (or windows for that matter) I
am just repeating some of the stuff I remember from reading though
some of the very long and involved Intel Doc.
Not sure if there is a mistake on the page or if
my laptop actually
has both of these
for something?? [2]
Intel 7 Series/C210 Series Chipset Family High Definition Audio
Controller
Intel Panther Point High Definition Audio Controller
It may have both on the same PCI IF even. cat /proc/interupts should
tell I think. As a note. On my netbook, the audio IF can run at a
number of sample rates with no problems, but my internal mic which
shows up on the same ALSA device, can only run at 48k. That is the ADC
for the internal mic only runs at 48k and if the rest of the internal
audio (internal speakers, line in/out plugs, are set to 44.1k, those
i/os will be fine, but the internal mic will have a periodic click as
the 44.1 and 48k sample rates align and cycle past each other. The
click goes away if the AI is set to 48K.
Be sure to look at the variety of HDA IF types the ALSA driver handles
and all of the settings that can be given to the HDA driver. It is
quite a long list.
11:47 p.m.
On Wed, 22 Oct 2014, Kazakore wrote:
HDA Audio:
But yeah there is a ~250 page pdf on the v1 docs which
can be found online
(in fact it's linked from the Wikipedia article I linked to earlier.) Not
read through it myself. But from the desctription I can understand why you
say Bus. But a bus wouldn't generally include codecs and things, so it's more
than just the bus really isn't it? Nearly a whole protocol really...
I don't think the codecs themselves are a part of it. But the way they
talk to it is, and so most codecs are "made for HDA" chips. I think any
codec can be used with the right glue.... just like with a PCI(e) AI the
glue has to work with the PCI bus.
The main thing though, is that HDA audio is not a name of an audio
interface that is specific, so much as a group of similarily accessed AIs.
I think the bus structure or protocol does limit how low latency can go. I
am not sure if running higher sample rates (96/192k) would give lower
latency or not as the transport speed is still 48k in any case. However
filling up the buffer faster may give lower latency. I think the HDA spec
is designed to have the same latency as the rest of the intel system with
no special tweaks. That is, still using all the selling point features,
such as "Boost technology" or "Hyper threading", both of which
increase
latency.
"Pros" are expected to be rich and willing to spend money on special
interfaces and computing engines, not try to get $10k plus ($100k is not a
lot for some of this stuff) performance from $1k minus studios :) Though
I can't see as that has ever stopped anyone. My point though is that what
we are doing is not what Intel has designed desktop CPUs and related chip
sets for. It is something we do in spite of the design.
--
Len Ovens
www.ovenwerks.net
21 Oct
21 Oct
9:48 a.m.
On 10/20/2014 05:10 PM, Patrick Shirkey wrote:
On Mon, October 20, 2014 10:59 pm, Jeremy Jongepier wrote:
Nice devices but are they fanless? They also seem to be fairly big, I
guess a Cubieboard is 4x as small as a NUC board. But I haven't really
compared the measurements yet so I could be wrong.
Jeremy
On 10/20/2014 11:34 AM, Atte wrote:
If you want something that is relatively cheap that can run pianoteq,
linuxsampler, etc... then you should look into the Intel NUC range.
They provide a very nice piece of gear for a stand alone solution and
decent responsive support via the OTC forums. I requested a change to the
BIOS and they did it for me as an update.
I have had one for several months as a head for my home entertainment
center and it is working very well. Yesterday I did some edits with
ardour2. All I had to do was "apt-get install ardour".
It's not $50 but it is less than $300 for a complete 64 bit x86 solution
including RAM/HDD/SSD. The case is brushed aluminium and it is very
robust. If you get an SSD it is nearly 100% solid state apart from the CPU
fan which is almost noiseless.
My (cheap) home stereo speaker system generates more noise than the fan
and I have it running 24/7 for net streaming to TV with full 1080p over
hdmi.
Saves all the hassle of learning how to cross compile and the general
issues associated with developing for ARM.
I tried several other options before settling on this one. They were
cheaper but also more frustrating.
Intel are catching up on ARM these days with the 16/14/12nm fab process.
Dunno about jack, but it seems interesting!
However I can't seem to find
any prices...
These devices are fairly cheap starting from about $50:
http://www.dx.com/p/ditter-u20-dual-core-android-4-2-google-tv-hd-player-w-…
That's one of the cheapest I could find with
an Allwinner A20 chipset.
Better would be to get something like a Cubieboard.
--
Patrick Shirkey
Boost Hardware Ltd
10:12 a.m.
2014-10-21 11:48 GMT+02:00 Jeremy Jongepier <jeremy(a)autostatic.com>om>:
On 10/20/2014 05:10 PM, Patrick Shirkey wrote:
If you want something that is relatively cheap that can run pianoteq,
linuxsampler, etc... then you should look into the Intel NUC range.
They provide a very nice piece of gear for a stand alone solution and
decent responsive support via the OTC forums. I requested a change to the
BIOS and they did it for me as an update.
I have had one for several months as a head for my home entertainment
center and it is working very well. Yesterday I did some edits with
ardour2. All I had to do was "apt-get install ardour".
It's not $50 but it is less than $300 for a complete 64 bit x86 solution
including RAM/HDD/SSD. The case is brushed aluminium and it is very
robust. If you get an SSD it is nearly 100% solid state apart from the CPU
It seems they do have some fan, Jeremy. On that link I posted they announce
another device fanless made for thin clients:
http://techpartner.intel.com/campaigns/nuc/thin-client/
BTW, I'm still looking forward to have or test a Cubieboard for audio, it
seem they can be quite cheap. I think you've made good progresses with them.
--
C. sanchiavedraZ:
* NEW / NUEVO: www.sanchiavedraZ.com
* Musix GNU+Linux: www.musix.es
On Mon, October 20, 2014 10:59 pm, Jeremy Jongepier wrote:
> On 10/20/2014 11:34 AM, Atte wrote:
>> Dunno about jack, but it seems interesting! However I can't seem to
find
>> any prices...
>
> These devices are fairly cheap starting from about $50:
>
http://www.dx.com/p/ditter-u20-dual-core-android-4-2-google-tv-hd-player-w-…
fan which is almost noiseless.
My (cheap) home stereo speaker system generates more noise than the fan
and I have it running 24/7 for net streaming to TV with full 1080p over
hdmi.
Saves all the hassle of learning how to cross compile and the general
issues associated with developing for ARM.
I tried several other options before settling on this one. They were
cheaper but also more frustrating.
Intel are catching up on ARM these days with the 16/14/12nm fab process.
Nice devices but are they fanless? They also seem to be fairly big, I
guess a Cubieboard is 4x as small as a NUC board. But I haven't really
compared the measurements yet so I could be wrong.
Jeremy
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user(a)lists.linuxaudio.org
http://lists.linuxaudio.org/listinfo/linux-audio-user
That's one of the cheapest I could find with
an Allwinner A20 chipset.
Better would be to get something like a Cubieboard.
--
Patrick Shirkey
Boost Hardware Ltd
10:46 a.m.
On Tue, October 21, 2014 8:48 pm, Jeremy Jongepier wrote:
On 10/20/2014 05:10 PM, Patrick Shirkey wrote:
No, but they are very quiet.
On Mon, October 20, 2014 10:59 pm, Jeremy Jongepier wrote:
Nice devices but are they fanless? On 10/20/2014 11:34 AM, Atte wrote:
If you want something that is relatively cheap that can run pianoteq,
linuxsampler, etc... then you should look into the Intel NUC range.
They provide a very nice piece of gear for a stand alone solution and
decent responsive support via the OTC forums. I requested a change to
the
BIOS and they did it for me as an update.
I have had one for several months as a head for my home entertainment
center and it is working very well. Yesterday I did some edits with
ardour2. All I had to do was "apt-get install ardour".
It's not $50 but it is less than $300 for a complete 64 bit x86 solution
including RAM/HDD/SSD. The case is brushed aluminium and it is very
robust. If you get an SSD it is nearly 100% solid state apart from the
CPU
fan which is almost noiseless.
My (cheap) home stereo speaker system generates more noise than the fan
and I have it running 24/7 for net streaming to TV with full 1080p over
hdmi.
Saves all the hassle of learning how to cross compile and the general
issues associated with developing for ARM.
I tried several other options before settling on this one. They were
cheaper but also more frustrating.
Intel are catching up on ARM these days with the 16/14/12nm fab process.
Dunno about jack, but it seems interesting!
However I can't seem to
find
any prices...
These devices are fairly cheap starting from about $50:
http://www.dx.com/p/ditter-u20-dual-core-android-4-2-google-tv-hd-player-w-…
That's one of the cheapest I could find with
an Allwinner A20 chipset.
Better would be to get something like a Cubieboard.
--
Patrick Shirkey
Boost Hardware Ltd They also seem to be fairly big, I
guess a Cubieboard is 4x as small as a NUC board. But I haven't really
compared the measurements yet so I could be wrong.
This particular model has a slightly larger case. The other models are not
as tall. The physical size is approx 10x10x4cm so they are very portable
and the solid aluminium casing is very robust.
Cubieboards are also very good units but they have some drawbacks too.
They are not as powerful as a Celeron and they require cross compiling
everything. IIUC, they also don't support the same amount of memory.
As a stand alone device the NUC's can do very well as a relatively low
cost processing unit. For the same form factor you can also get an i5
chipset. I'm not sure if they have released the i7's yet but they are
probably just around the corner. The top end i5's are very close to the
same speed as an i7.
--
Patrick Shirkey
Boost Hardware Ltd
11:39 a.m.
On 10/21/2014 12:46 PM, Patrick Shirkey wrote:
<snip>
Bummer, I wouldn't trust on a device that has fans for on stage use.
Not entirely true, most of the software is readily available in the
repositories. If you need unpackaged software or newer versions of
software you need to recompile, yes. I've got a toolchain set-up for
this (an ARM chroot with everything installed needed for Debian
packaging) and it would also be possible to do this on the device
itself. But indeed, on a x86 machine it's not as trivial as building
stuff for x86 itself.
IIUC, they also don't support the same amount of memory.
No idea ;) Never checked that actually since most devices I know don't
come with memory slots or memory chips that can be piggybacked.
Nice devices but are they fanless?
No, but they are very quiet.
They also seem
to be fairly big, I
guess a Cubieboard is 4x as small as a NUC board. But I haven't really
compared the measurements yet so I could be wrong.
This particular model has a slightly larger case. The other models are not
as tall. The physical size is approx 10x10x4cm so they are very portable
and the solid aluminium casing is very robust.
Cubieboards are also very good units but they have some drawbacks too.
They are not as powerful as a Celeron and they require cross compiling
everything. As a stand alone device the NUC's can do very well
as a relatively low
cost processing unit. For the same form factor you can also get an i5
chipset. I'm not sure if they have released the i7's yet but they are
probably just around the corner. The top end i5's are very close to the
same speed as an i7.
Yeah, I'm definitely going to check them out, such a device would be
awesome as a multimedia player. So thanks for the info!
Jeremy
1:12 p.m.
On Tue, October 21, 2014 10:39 pm, Jeremy Jongepier wrote:
Not entirely true, most of the software is readily available in the
repositories. If you need unpackaged software or newer versions of
software you need to recompile, yes. I've got a toolchain set-up for
this (an ARM chroot with everything installed needed for Debian
packaging) and it would also be possible to do this on the device
itself. But indeed, on a x86 machine it's not as trivial as building
stuff for x86 itself.
IIUC, they also don't support the same amount of memory.
No idea ;) Never checked that actually since most devices I know don't
come with memory slots or memory chips that can be piggybacked.
The ARM platform shouldn't have a problem with larger memory but the
cubietrucks and all the portable/low end ARM devices I have seen do not
provide support for large amounts of RAM. The highest I have seen is 2GB.
--
Patrick Shirkey
Boost Hardware Ltd
On 10/21/2014 12:46 PM, Patrick Shirkey wrote:
<snip>
Bummer, I wouldn't trust on a device that has fans for on stage use.
Every DJ I know uses a laptop with a fan so it depends on how you intend
to use them. As a stomp box it's probably not a good idea but for a
linuxsampler/pianoteq/realtime synth module which sits on a stable dry
surface it shouldn't be a problem.
Nice devices but are they fanless?
No, but they are very quiet.
They also seem to be fairly big, I
guess a Cubieboard is 4x as small as a NUC board. But I haven't really
compared the measurements yet so I could be wrong.
This particular model has a slightly larger case. The other models are
not
as tall. The physical size is approx 10x10x4cm so they are very portable
and the solid aluminium casing is very robust.
Cubieboards are also very good units but they have some drawbacks too.
They are not as powerful as a Celeron and they require cross compiling
everything. As a stand
alone device the NUC's can do very well as a relatively low
cost processing unit. For the same form factor you can also get an i5
chipset. I'm not sure if they have released the i7's yet but they are
probably just around the corner. The top end i5's are very close to the
same speed as an i7.
Yeah, I'm definitely going to check them out, such a device would be
awesome as a multimedia player. So thanks for the info!
6:50 p.m.
New subject: kernel rt-patches future
Dear list,
some of you may know this:
http://lwn.net/Articles/572740/
has there already been any thread about the future of the rt-patches?
How to build a realtime audio system without rt-patches? Is rt-patches
near dead, as there is no release since kernel 3.14? Guess this would
have a very big impact on linux audio. Please tell if I'm missing something.
Best regards,
Uli
6:55 p.m.
New subject: kernel rt-patches future
On Tue, 2014-10-21 at 20:46 +0200, Ulrich-Lorenz Schlüter wrote:
Dear list,
some of you may know this:
http://lwn.net/Articles/572740/
has there already been any thread about the future of the rt-patches?
How to build a realtime audio system without rt-patches? Is rt-patches
near dead, as there is no release since kernel 3.14? Guess this would
have a very big impact on linux audio. Please tell if I'm missing something.
Calm down :) and subscribe to
http://vger.kernel.org/majordomo-info.html :).
9:07 p.m.
New subject: kernel rt-patches future
On Tue, 21 Oct 2014, Ulrich-Lorenz Schlüter wrote:
Dear list,
some of you may know this:
http://lwn.net/Articles/572740/
has there already been any thread about the future of the rt-patches?
How to build a realtime audio system without rt-patches? Is rt-patches
near dead, as there is no release since kernel 3.14? Guess this would
have a very big impact on linux audio. Please tell if I'm missing something.
What it also says is that most of the RT bits for the Linux Kernel are
already included in the mainline kernel. How much of what is left affects
audio?
In my opinion, Much of the use for RT kernels in audio is to try and fix
problems with system HW tuning. In my trials, I have found that using only
a "lowlatency" kernel, I can get clean latency as low as my card is able
to be set anyway (32 samples). The only xruns I get are from applications
that do not shut down correctly or take to long starting up (that is they
enable ports in jack before they are stable).
PC MB are not designed for low latency work, but rather large throughput
and good graphics speed. Getting good low latency performance out of a PC
requires tuning it, both SW and HW. Too many people see the RT kernel as
some sort of magic bullet that will fix everything. The PC MB is designed
for audio latency of 192 samples of 48k audio as minimum latency with the
odd xrun being ok. (we are talking games and youtube here) Yes it does
mean being picky about which MB is chosen, which CPU, which GPU, which GPU
drivers (and it's settings), which slot the audio board is in, performance
over ondemand, hyperthreading off, ... just to name a few.
The other thing the RT developers have pointed at is that changes to HW
(multi-cores in particular) have opened up a whole new way of dealing with
latency (for better or worse). A core or cores can be set aside for RT
work only. With some processors other resources as well as just the
processor core can be set aside from the OS such as memory.
But yes, the standard RT keneral development is seeing the money dry up.
--
Len Ovens
www.ovenwerks.net
22 Oct
22 Oct
8:32 a.m.
New subject: kernel rt-patches future
On Tue, 21 Oct 2014 14:05:23 -0700 (PDT)
Len Ovens <len(a)ovenwerks.net> wrote:
In my opinion, Much of the use for RT kernels in audio
is to try and
fix problems with system HW tuning. In my trials, I have found that
using only a "lowlatency" kernel, I can get clean latency as low as
my card is able to be set anyway (32 samples). The only xruns I get
are from applications that do not shut down correctly or take to long
starting up (that is they enable ports in jack before they are
stable).
IME/O, the one thing the -rt kernel brings to audio processing on
linux is reliable kernel scheduling. Provided that all the hardware
plays ball (which most does), you can expect a scheduling delay of
under 100usecs with a -rt patched kernel.
With vanilla kernels (even low latency ones), you can expect to see the
occasional latency peak into the msec range, each such peak will mean
an xrun if the kernel scheduling and audio processing latencies exceed
the deadline as given by samplerate and buffer sizes. They might be
just occasional and not a problem, but they will be there.
But yes, the standard RT keneral development is seeing
the money dry
up.
There never was any money to dry up. Most of the development have been
carried by https://www.linutronix.de, Red Hat and private individuals.
--
Joakim
9:21 a.m.
New subject: kernel rt-patches future
On Wed, 2014-10-22 at 10:32 +0200, Joakim Hernberg wrote:
Provided that all the hardware plays ball (which most
does)
It's easy to get issues caused by the hardware :(. I cared about getting
hardware that should play ball, my RME card doesn't work ok. Already
many years ago I experienced issues with hardware that isn't related to
audio. I e.g. got a 56k modem praised as working with Linux and indeed
most revisions of the modem did, but even the vendor forgot to mention,
that just one of several revisions doesn't and I exactly got this
revision. Most of the times I had good luck and send back the hardware
and got my money back, but for e.g. the RME card I didn't test it
extensive within the time, when it was possible to get the money back.
Hardware combinations come with a lot of pitfalls. Your hardware might
be ok now, but if you need to replace something, the trouble could
start. However, I experienced the RT patches most of the times as the
best solution for audio and MIDI, but until now, the last patch after
the 3.8.13-rt patches, the only working patch is 3.10.47-rt50, all other
newer patches already lock my machine at startup.
30 Oct
30 Oct
10:38 p.m.
New subject: kernel rt-patches future
On Wed, 22 Oct 2014 11:21:07 +0200
Ralf Mardorf <ralf.mardorf(a)rocketmail.com> wrote:
On Wed, 2014-10-22 at 10:32 +0200, Joakim Hernberg
wrote:
Is it still sharing an interrupt with nvidia and other devices?
Provided that all the hardware plays ball (which
most does)
It's easy to get issues caused by the hardware :(. I cared about
getting hardware that should play ball, my RME card doesn't work ok. the trouble could start. However, I experienced the RT
patches most
of the times as the best solution for audio and MIDI, but until now,
the last patch after the 3.8.13-rt patches, the only working patch is
3.10.47-rt50, all other newer patches already lock my machine at
startup.
The rt patch set was just rebased on a few kernel versions, let's hope
your AMD issue has gone away :)
--
Joakim
10:48 p.m.
New subject: kernel rt-patches future
On Thu, 30 Oct 2014 23:38:29 +0100
Joakim Hernberg <jhernberg(a)alchemy.lu> wrote:
On Wed, 22 Oct 2014 11:21:07 +0200
Ralf Mardorf <ralf.mardorf(a)rocketmail.com> wrote:
No shared interrupts with newer kernels, with older kernels it shares
the interrupt with the graphics. It doesn't matter, if I use the
internal ATI or a PCIe NVIDIA. The two PCIe slots on the mobo seem to
be one split thingy. This might explain the xruns even with much too
long latencies for music production. Another issue is that only 2 of
the 8 ADAT channels are available by the jack ports.
On Wed, 2014-10-22 at 10:32 +0200, Joakim
Hernberg wrote:
Is it still sharing an interrupt with nvidia and other devices? Provided that all the hardware plays ball (which
most does)
It's easy to get issues caused by the hardware :(. I cared about
getting hardware that should play ball, my RME card doesn't work ok.
2 Nov
2 Nov
1:58 a.m.
New subject: kernel rt-patches future
On Thu, 30 Oct 2014 23:38:29 +0100
Joakim Hernberg <jhernberg(a)alchemy.lu> wrote:
The rt patch set was just rebased on a few kernel
versions, let's hope
your AMD issue has gone away :)
Good news :)
http://archaudio.org/pipermail/archaudio-discuss/2014-November/000371.html .
22 Oct
22 Oct
9:30 p.m.
New subject: kernel rt-patches future
On Wed, 22 Oct 2014, Joakim Hernberg wrote:
Red Hat was paying at least some people's salary and I think some of the
companies who used RT allowed their employees to help out on company time
as well (perhaps no official funding though).
--
Len Ovens
www.ovenwerks.net
On Tue, 21 Oct 2014 14:05:23 -0700 (PDT)
Len Ovens <len(a)ovenwerks.net> wrote:
I seem to be able to stress the cpu, use the desktop, and run jack at 16/2
for hours on end with no xruns at all. Then load an app that uses jack and
see xruns only while loading or exiting. sure, I don't do much synth work,
more effects at that latency, but even still, I have been very impressed
with just the low latency kernel. (i5 at 3.4G BTW) I think too, there are
more switches in the kernel that can be un/set to make it more RT than the
average lowlatency version.
I have not done that much system tuning either. I did choose the i5 on
purpose for no HT, chose to use a pci AI and which slot it should be in.
No wifi, performance mode. Mouse and kb are not USB (in fact no USB use at
all, now that I think about it) Video is Intel based.
In my opinion, Much of the use for RT kernels in
audio is to try and
fix problems with system HW tuning. In my trials, I have found that
using only a "lowlatency" kernel, I can get clean latency as low as
my card is able to be set anyway (32 samples). The only xruns I get
are from applications that do not shut down correctly or take to long
starting up (that is they enable ports in jack before they are
stable).
IME/O, the one thing the -rt kernel brings to audio processing on
linux is reliable kernel scheduling. Provided that all the hardware
plays ball (which most does), you can expect a scheduling delay of
under 100usecs with a -rt patched kernel.
With vanilla kernels (even low latency ones), you can expect to see the
occasional latency peak into the msec range, each such peak will mean
an xrun if the kernel scheduling and audio processing latencies exceed
the deadline as given by samplerate and buffer sizes. They might be
just occasional and not a problem, but they will be there. But yes, the
standard RT keneral development is seeing the money dry
up.
There never was any money to dry up. Most of the development have been
carried by https://www.linutronix.de, Red Hat and private individuals.
30 Oct
30 Oct
10:43 p.m.
New subject: kernel rt-patches future
On Wed, 22 Oct 2014 14:30:21 -0700 (PDT)
Len Ovens <len(a)ovenwerks.net> wrote:
I seem to be able to stress the cpu, use the desktop,
and run jack at
16/2 for hours on end with no xruns at all. Then load an app that
uses jack and see xruns only while loading or exiting. sure, I don't
do much synth work, more effects at that latency, but even still, I
have been very impressed with just the low latency kernel. (i5 at
3.4G BTW) I think too, there are more switches in the kernel that can
be un/set to make it more RT than the average lowlatency version.
Well in the end that is all that counts. I'm happy it works well for
you.
I have not done that much system tuning either. I did
choose the i5
on purpose for no HT, chose to use a pci AI and which slot it should
be in. No wifi, performance mode. Mouse and kb are not USB (in fact
no USB use at all, now that I think about it) Video is Intel based.
I mainly use an i7, I did however disable HT :) I did not see many
xruns caused by it in my testing, but they were occasionally there.
The cost was noticably longer build times for wine and the kernel :)
--
Joakim
31 Oct
31 Oct
8:33 p.m.
New subject: kernel rt-patches future
On Thu, 30 Oct 2014, Joakim Hernberg wrote:
On Wed, 22 Oct 2014 14:30:21 -0700 (PDT)
Len Ovens <len(a)ovenwerks.net> wrote:
I am not sure exactly how HT works. That is I do not know the mechanism
for switching threads. I would assume though, that it is cycle based,
after so many cycles we switch to another thread. That being the case, a
faster processor would allow lower latency with HT on. My experience with
a 2.4Ghz P4 was that jack at 64/2 was not really a problem with HT on but
anything lower was. I would expect an i7 running at 3.6Ghz would at least
be able to run jack at 32/2 with no problems, but at some point HT will be
a barrier to lower latency. In audio, it would seem that so long as there
are no xruns at the latency being used there is not problem. So long as
the sound "card" is providing the media clock that is true. If someone was
trying to run a stable PTP based sw media clock, I think jitter would be
a problem. I think that is one of the reasons most AoIP setups suggest
using their NIC/AI which would have it's own HW based PTP server. (a
number of NIC chips have one built in, even cheap $5 ones)
So far as I know, HT thread switching is controled by firmware outside the
control of the OS. If it could be controlled by the OS, low latency might
be enhanced with HT on as a high priority thread could be left in CPU
memory while servicing a lower priority thread. I don't think a single
core would be helped with this, but a 4 core cpu might. It would depend on
the total number of threads and how many of them had raised priority.
--
Len Ovens
www.ovenwerks.net
I have not done that much system tuning either. I
did choose the i5
on purpose for no HT, chose to use a pci AI and which slot it should
be in. No wifi, performance mode. Mouse and kb are not USB (in fact
no USB use at all, now that I think about it) Video is Intel based.
I mainly use an i7, I did however disable HT :) I did not see many
xruns caused by it in my testing, but they were occasionally there.
The cost was noticably longer build times for wine and the kernel :)
8:41 p.m.
New subject: kernel rt-patches future
On Fri, 31 Oct 2014 13:32:54 -0700 (PDT)
Len Ovens <len(a)ovenwerks.net> wrote:
On Thu, 30 Oct 2014, Joakim Hernberg wrote:
[snip]
Have you ever tried hpet for jackd audio?
For ALSA queue timer the HR timer reduces MIDI jitter on my machine. My
machine doesn't have the resources to use it for audio too.
8:51 p.m.
New subject: kernel rt-patches future
On Fri, 31 Oct 2014, Ralf Mardorf wrote:
On Fri, 31 Oct 2014 13:32:54 -0700 (PDT)
Len Ovens <len(a)ovenwerks.net> wrote:
I wonder how HT and ondemand governing affect hpet.
--
Len Ovens
www.ovenwerks.net
On Thu, 30 Oct 2014, Joakim Hernberg wrote:
[snip]
Have you ever tried hpet for jackd audio?
For ALSA queue timer the HR timer reduces MIDI jitter on my machine. My
machine doesn't have the resources to use it for audio too.
8:59 p.m.
New subject: kernel rt-patches future
On Fri, 31 Oct 2014 13:50:39 -0700 (PDT)
Len Ovens <len(a)ovenwerks.net> wrote:
Your machine seems to fulfil the requirements to test it. One day
you're perhaps bothered by deadly boredom, then remember it and check
it out :).
On Fri, 31 Oct 2014, Ralf Mardorf wrote:
> On Fri, 31 Oct 2014 13:32:54 -0700 (PDT)
> Len Ovens <len(a)ovenwerks.net> wrote:
>> On Thu, 30 Oct 2014, Joakim Hernberg wrote:
> [snip]
>
> Have you ever tried hpet for jackd audio?
>
> For ALSA queue timer the HR timer reduces MIDI jitter on my
^ MIDI
;)
machine. My
machine doesn't have the resources to use it for audio
too.
I wonder how HT and ondemand governing affect hpet.
21 Oct
21 Oct
6:51 p.m.
On Wed, 2014-10-22 at 02:35 +0800, Kazakore wrote:
On 21/10/14 21:12, Patrick Shirkey wrote:
And dinosaurs like me prefer to listen to vinyl (at home) (or to very
good DAT masters) (I dislike to visit locations where DJs are needed).
Btw. usually I don't use a slipmat ;), but I like some DJ's who know how
to use a slipmat :).
Old school rules! A "DJ" using a Laptop isn't a DJ.
Every DJ I know uses a laptop
But real DJs use vinyl No fans there!! ;-)
7:06 p.m.
On Tue, 2014-10-21 at 20:51 +0200, Ralf Mardorf wrote:
On Wed, 2014-10-22 at 02:35 +0800, Kazakore wrote:
My turntable has got an odd mini wheel instead of a fader to adjust the
speed, no MK II here, so using the turntable as an instrument doesn't
work that good with my thingy, two turntables anyway would be better.
On 21/10/14 21:12, Patrick Shirkey wrote:
And dinosaurs like me prefer to listen to vinyl (at home) (or to very
good DAT masters) (I dislike to visit locations where DJs are needed).
Btw. usually I don't use a slipmat ;), but I like some DJ's who know how
to use a slipmat :). Every DJ I know uses a laptop
But real DJs use vinyl No fans there!! ;-)
Old school rules! A "DJ" using a Laptop isn't a DJ.
7:13 p.m.
On Wed, October 22, 2014 5:35 am, Kazakore wrote:
On 21/10/14 21:12, Patrick Shirkey wrote:
It's rare to find a DJ that doesn't use a laptop at least some of a set
these days. Nearly everyone has one.
In my experience there's a greater risk of overheating without a fan and
the ARM (allwinner) chipsets are prone to that. My bet is a low power x86
processor/unit with a (quiet) fan will out perform and outlast an ARM
chipset without.
But the difference is around $200 so you get what you pay for.
IMO, it's still a big ask to expect a super cheap ARM device from a cheap
Chinese manufacturer to perform to the same levels as a $300 x86 unit from
Intel. The cheap chipsets are just not designed for professional level
abuse especially from manufacturers like AllWinner/RockChip/MediaTek. The
cheaper you go the less reliable/powerful they become. For example the
a20's have to be clocked down if you don't want them overheating. They
have an auto shutoff at 70degC which is surprisingly easy to trigger in a
warm climate/room. You might get good results from Qualcomm or TI but I
haven't used their chipsets. Some other people round here work for those
companies though so they might be able to offer some insight.
If I was performing and needed a linuxsampler/pianoteq/realtime synth unit
I would go for a $500 i5 unit (incl fan), with a really fast SSD and 16GB
DDR3 RAM as a baseline. I used a similar setup for a few years and never
had any problems with fans. I also spent a lot more money and time on it
too. It would have been great to have a portable quad core with 16GB of
RAM and dual 1080p hdmi back then. That would have enabled me to achieve a
lot more. Back then those kind of specs costs several thousand dollars
and were definitely not portable. IIRC 8GB was the max for RAM too.
So, spend a little more and get something really killer. Be glad you don't
have to spend thousands any more to get something that can do almost
everything you can throw at it.
The difference between $50 and $500 is a few weeks work for most people
earning a western income so it's worth the wait.
--
Patrick Shirkey
Boost Hardware Ltd
Every DJ I know uses a laptop
But real DJs use vinyl No fans there!! ;-)
8:47 p.m.
On Wed, 22 Oct 2014, Kazakore wrote:
On 21/10/14 21:12, Patrick Shirkey wrote:
On McCurdy turntables?
--
Len Ovens
www.ovenwerks.net
Every DJ I know uses a laptop
But real DJs use vinyl No fans there!! ;-)
22 Oct
22 Oct
6:11 a.m.
On Tue, 2014-10-21 at 13:45 -0700, Len Ovens wrote:
On Wed, 22 Oct 2014, Kazakore wrote:
DJs for good reasons usually use the MK II from Technics.
On 21/10/14 21:12, Patrick Shirkey wrote:
On McCurdy turntables? Every DJ I know uses a laptop
But real DJs use vinyl No fans there!! ;-)
9:11 p.m.
On Wed, 22 Oct 2014, Ralf Mardorf wrote:
On Tue, 2014-10-21 at 13:45 -0700, Len Ovens wrote:
The McCurdys are older from what I can tell. When the Technics came out,
McCurdy used them in their equipment as well. Generally though, the
McCurdy TT would require the DJ to have roadies. When I was in broadcast,
there were still a lot of McCurdys around (I do live in Canada BTW) but
anything new was Technics. So kept a McCurdy 16in even after they replaced
most with newer technics 12 in units. Most of the highly automated places
used Cart machines with a big autoloader. There was a wall full of carts
that the cart machine was moved in front of the cart slot so it could be
loaded. Must have been more than one play slot. (I only saw one once, so I
don't remember very clear) McCurdy made mixers as well and from what I can
tell, the Mic pres are quite well thought of by those who have them.
McCurdy custom made things for CBC (like marcony did for BBC) and a lot of
early radio stations in Canada as well. I saw so many of them in the
places I toured I thought they must be everywhere, but I think in the US
the RCA was more widespread. Europe would have had their own favourite
brands too.
--
Len Ovens
www.ovenwerks.net
On Wed, 22 Oct 2014, Kazakore wrote:
DJs for good reasons usually use the MK II from Technics. On 21/10/14 21:12, Patrick Shirkey wrote:
On McCurdy turntables? Every DJ I know uses a laptop
But real DJs use vinyl No fans there!! ;-)
21 Oct
21 Oct
8:31 p.m.
On Tue, 21 Oct 2014, Patrick Shirkey wrote:
How small should they be? In the end one wants to put robust connectors on
it anyway. Robust connectors are in general, big. I can understand wanting
to be able to carry ones kit all in one go, but if a box is 2x2x1 inch, or
6x6x1.5 inch doesn't affect that really. Something that can survive being
trod on is more important.
--
Len Ovens
www.ovenwerks.net
On Tue, October 21, 2014 8:48 pm, Jeremy Jongepier
wrote:
Cooling solutions can always be changed out to something fanless. The Atom
board that I have, is fanless and the laptop harddrive I have put with it
is also very quiet. This is as used for a desktop. As stage box I would
use SSD as there is a socket on board for the mini version. If I was using
it as a stage box, I might put an older (pre-AC97) sound card on a
90degree adaptor for audio/MIDI as the PCIaudio card I have seems to do
-p32 -n2 quite well and having internal MIDI (well, through a game port)
is a bit more solid than USB. (if the keyboard is USB already this is no
use)
Nice devices but are they fanless?
No, but they are very quiet. They also seem to be fairly big, I
guess a Cubieboard is 4x as small as a NUC board. But I haven't really
compared the measurements yet so I could be wrong.
This particular model has a slightly larger case. The other models are not
as tall. The physical size is approx 10x10x4cm so they are very portable
and the solid aluminium casing is very robust.
20 Oct
20 Oct
2:14 p.m.
On Mon, 20 Oct 2014, rosea grammostola wrote:
Is it possible to install JACK etc. and for instance
Pianoteq on a device
like this, to create your own instrument module?
http://www.rikomagic.com/en/product/showpro_id_58_pid_22.html
It "looks" like it "should" be possible. It uses Ubuntu's Arm
repos which
if they do not include jack, do include all the build tools to make it
with. There may be no FPU, so that may limit what sound modules will
build.
What I am saying is that this is a "set top box" made to do TV kinds of
things. The Ubuntu repos for the Arm will include those things people have
spent the time to do for the Arm achitecture. So I would expect to have to
build some of the things I wanted. It is impossible to know how well the
audio HW is set up, if it has it's own irq, if it is 16bit/48k, etc. Is
there a lowlatency kernel ready made or will you need to roll your own?
(Is the hw put together in a way it will work without?) Control would have
to be USB or ethernet as there is no serial port to play with. The WIFI
may be more of a liability than a help for music (it has to be turned off
on this netbook).
Basically, it comes down to, what is the cost? Is it cheap enough I can
use it for something else if it doesn't work? The only way to find out for
sure is to buy and try.
For stage use, it may need to be "reboxed" with good road worthy
connectors, USB stuff always plugged in or at least wired out to good
quality pannel USB plugs. 1/8 wired to 1/4 inch plus etc.
--
Len Ovens
www.ovenwerks.net
2:36 p.m.
On Tue, October 21, 2014 1:12 am, Len Ovens wrote:
On Mon, 20 Oct 2014, rosea grammostola wrote:
Unfortunately the Rockchip series are completely crap for realtime audio.
Allwinner chips provide a better all round developer experience.
Is it possible to install JACK etc. and for
instance Pianoteq on a
device
like this, to create your own instrument module?
http://www.rikomagic.com/en/product/showpro_id_58_pid_22.html
It "looks" like it "should" be possible. It uses Ubuntu's Arm
repos which
if they do not include jack, do include all the build tools to make it
with. There may be no FPU, so that may limit what sound modules will
build.
What I am saying is that this is a "set top
box" made to do TV kinds of
things. The Ubuntu repos for the Arm will include those things people have
spent the time to do for the Arm achitecture. So I would expect to have to
build some of the things I wanted. It is impossible to know how well the
audio HW is set up, if it has it's own irq, if it is 16bit/48k, etc. Is
there a lowlatency kernel ready made or will you need to roll your own?
(Is the hw put together in a way it will work without?) Control would have
to be USB or ethernet as there is no serial port to play with. The WIFI
may be more of a liability than a help for music (it has to be turned off
on this netbook).
Basically, it comes down to, what is the cost? Is it cheap enough I can
use it for something else if it doesn't work? The only way to find out for
sure is to buy and try.
For stage use, it may need to be "reboxed" with good road worthy
connectors, USB stuff always plugged in or at least wired out to good
quality pannel USB plugs. 1/8 wired to 1/4 inch plus etc.
--
Patrick Shirkey
Boost Hardware Ltd
2:49 p.m.
On Tue, 21 Oct 2014, Patrick Shirkey wrote:
On Tue, October 21, 2014 1:12 am, Len Ovens wrote:
Good to know.
--
Len Ovens
www.ovenwerks.net
It "looks" like it "should" be possible. It uses Ubuntu's Arm
repos which
if they do not include jack, do include all the build tools to make it
with. There may be no FPU, so that may limit what sound modules will
build.
Unfortunately the Rockchip series are completely crap for realtime audio.
Allwinner chips provide a better all round developer experience.
3469
days inactive
3482
days old
linux-audio-user@lists.linuxaudio.org
48 comments
12 participants
participants (12)
-
Atte -
Carlos sanchiavedraz -
Harry van Haaren -
Jeremy Jongepier -
Joakim Hernberg -
Kazakore -
Len Ovens -
Patrick Shirkey -
Ralf Mardorf -
rosea grammostola -
Russell Hanaghan -
Ulrich-Lorenz Schlüter