Does anybody know how to fix a PKGBUILD or what ever needs a fix, to
build muse2 2.1.1?
https://aur.archlinux.org/packages/muse2/ see comment by Ralf_Mardorf
2013-03-10 20:16.
On Sat, March 9, 2013 9:46 am, Ralf Mardorf wrote:
> I only want to be able to play at least a stereo track (a temporary mix
> of a production with much more tracks) and to record some tracks with
> microphones for drums. Perhaps it's needed to use a mixer and to record
> a stereo mix, I guess onboard surround devices can't record more than
> stereo, especially in duplex mode.
The HDA standard (in as much as there is such a thing) is stereo in at
16bits, though some have 20bit. The 20 bits seems to be a waste of time
as, in my experience, the s/n on even the line inputs is something like an
old cassette recorder. On my desktop it is not even uniform from one
channel to the other :P I have my internal audio disabled in bios. I also
have an old ensoniq 1370 which is much quieter than any internal audio I
have heard. The ART dualUSBpre would be a better bet (or something
similar)... it does have the USB problems but it is USB1.1 so bandwidth is
low and for recording you may get away with quite a high latency anyway.
It is very small to carry and USB powered. (remembering you also mentioned
a mixer :) )
Outputs are mostly 24bit though and normally reasonable quality
amplification.
--
Len Ovens
www.OvenWerks.net
On Sat, March 9, 2013 9:49 am, Peder Hedlund wrote:
> Quoting Len Ovens <len(a)ovenwerks.net>:
>
> The previous reply was due to a trigger happy Send finger, sorry about
> that...
been there...
>>> An obvious workaround would be having a cron job doing a 'touch
>>> TempFile' or something similar on the disk once every 5 minutes or so.
>>
>> Would that actually touch the disk or just the copy of the temp file in
>> the ram buffers?
>
> Initially it would only hit the cache but I think it's be written to
> disk within the remaining five minutes.
I would guess it would depend on how busy the system was. I don't think it
would be high priority on a system running real time audio tasks. I don't
know enough about background disk writing to really say though.
> You could issue a 'sync' but I'm not sure a forced sync during, say, a
> recording would be the proper thing to do, since it syncs all cached
> data on all disks.
Ya that was my thought too.
>
> Another option would be to do a 'dd if=/dev/zero of=TestFile bs=1k
> count=1 conv=fdatasync' to make sure the data gets written to disk
> immediately (http://romanrm.ru/en/dd-benchmark).
Run in a script with sleep, not from cron which might be turned off for
recording :) apt-get update running in the BG is enough to give me xruns
sometimes (I am not sure it is apt-get itself, but there is system disk
and network activity too)
--
Len Ovens
www.OvenWerks.net
On Sat, March 9, 2013 4:30 am, Ralf Mardorf wrote:
> an external hard drive does spin down too often for my taste. It's a
> http://www.wdc.com/global/products/specs/?driveID=762&language=1 and I
> wonder if there's a way to disable it.
So far as i can tell mine doesn't, but then mine is a case I added a drive
from an old laptop that stopped working. BTW this is not the drive itself
but the USB board... Or if it is the in the drive itself, it would be the
same if the drive was used internally.
> Regarding to a German PDF SmartWare doesn't support this particular
> external hard drive and even if it would, it seems to need Windows or OS
> X.
So all you need is to provide disk activity every 8 or nine minutes...
sync or something. I am not sure but I don't know how bad sync is at
grabbing system resources and if there are a lot of full buffers this
could take some time to complete.
> Is there a way to adjust this feature? I fear that spinning up and down
> again and again, will reduce the lifetime of the drive significantly and
> assumed I'll use it for recording, it might cause performance issues.
I don't know if this would cause an early fail, but from the
manufacturer's POV, so long as it is past warranty that is a good thing :P
> Is it possible to use such drives for recording, without any risk? If I
> would test it, I might not cover all worst case scenarios, think it
> wouldn't cause issues and then be surprised if it fails in an
> unfavorable moment.
This is a hard issue, because I have 2.5G ram. My OS and all the apps I
use for recording take up maybe about 1G (cause I was able to record when
I only had 1G). So there is 1.5G worth of ram that can buffer data going
to coming from disk. I might be able to record several tracks without ever
touching the drive. So to test this you have to have enough tracks to fill
ram buffers and force write to disk. Now when tracking, normally I am
reading a bunch of tracks and only writing one or two. Reading takes less
system resources than writing (I think... requires inode writes at least)
so if the system is smart it will off load the reading first. Once it does
that your drive will stay spinning, so the write part will not induce a
spin up... anyway to test you need something that is on the verge of
needing real disk access. Memory almost full and then record till the
drive works. I am assuming the drive is only used for audio and no system
commands would touch it.
> This drive should be used as a backup media only, but I'm thinking to
> buy an additional external hard drive for recording, to be able to visit
> a Windows user (Mac became seldom here :), but to boot Linux from the
> drive and to record on the drive.
Sounds like fun :) now the amount of ram in an unknown system comes into
play. Assuming you are taking a USB audio IF, then USB irqs have to be
looked at before deciding where to put the IF, USB stick and drive (DVD
might in this case be better than USB stick, though the USB sticks seem to
be more reliable). I personally have not had any problems a USB drive
being in a USB port that shares irqs while recording, but I can not say
for sure that a write or read was forced while I was recording.
--
Len Ovens
www.OvenWerks.net
I'm checking out Ubuntu Studio 12.04, 64-bit. When I try to play a
wave file with using aplay <name of file>, I am told it is an
unsupported format (and the file is identified correctly as 3 byte
signed little endian mono) and it lists the available formats of little
and big endian versions of 16 and 32 bit and 32 float, and even 8 bit.
What happened to 24 bit? Can I add it? Thanks.
It seems some nasty hacker has defaced ardour.org
I hate when they do that, never any thought about all the hard work
people have put in to make the website look nice..
I hope Paul has a recent backup so he can get the site back to it's
old looks soon.
(sorry if I spoiled anything Paul... :)
On Wed, March 6, 2013 11:46 am, Gabbe Nord wrote:
> Hello all, and thank you so much for your answers! I feel like I need to
clear up a few things here for my own sake, sorry for being slow: * You
are referring to an "I/O" USB card that connects to PCI-E as a solution,
right?
Yes. See my comments below as the two are related :)
> * I don't really know what IRQs are to which port etc, but I have my
soundcard plugged in to one port, and my USB-hub with everything else on
on
> yet another. I also have USB-ports on the front of my case, which are
connected to the motherboard directly. Can I somehow find out what's
connected to what inside the computer, and how do I modify my IRQ-files
to
> correspond to where I have my soundcard? Will unplugging/replugging and
using lsusb -t do that for me?
You posted your cat /proc/interrupts a few days ago. In there I could only
see two USB ports at all. USB1 and USB2. Not only that, but both USB ports
use the same IRQ. What this means is that, even though there are some more
physical USB ports on your machine, they are only there because the MB has
an internal USB hub (or two). The only way to get beyond that I can see is
to add a new USB interface on a fresh IRQ. I don't know what kind of slots
your MB has (PCI or PCIe or both) but USB cards should be available for
either... probably cheaper than a new audio card. Make sure it is USB2.0
or USB3.0 (+2 +1.1) and not just 1.1 ...
I normally just use dmesg to find out what the new device is:
[363472.113041] usb 1-7: new high-speed USB device number 3 using ehci-pci
That says the usb stick I inserted is in USB1 (a USB2.0 port) and that is
device 3 (out of 4 I am guessing)
lsusb -t gives this before I add the device:
$ lsusb -t
/: Bus 05.Port 1: Dev 1, Class=root_hub, Driver=uhci_hcd/2p, 12M
/: Bus 04.Port 1: Dev 1, Class=root_hub, Driver=uhci_hcd/2p, 12M
/: Bus 03.Port 1: Dev 1, Class=root_hub, Driver=uhci_hcd/2p, 12M
/: Bus 02.Port 1: Dev 1, Class=root_hub, Driver=uhci_hcd/2p, 12M
/: Bus 01.Port 1: Dev 1, Class=root_hub, Driver=ehci-pci/8p, 480M
|__ Port 8: Dev 2, If 0, Class=stor., Driver=usb-storage, 480M
(Dev 2 is my system disk right now)
and this after:
$ lsusb -t
/: Bus 05.Port 1: Dev 1, Class=root_hub, Driver=uhci_hcd/2p, 12M
/: Bus 04.Port 1: Dev 1, Class=root_hub, Driver=uhci_hcd/2p, 12M
/: Bus 03.Port 1: Dev 1, Class=root_hub, Driver=uhci_hcd/2p, 12M
/: Bus 02.Port 1: Dev 1, Class=root_hub, Driver=uhci_hcd/2p, 12M
/: Bus 01.Port 1: Dev 1, Class=root_hub, Driver=ehci-pci/8p, 480M
|__ Port 7: Dev 5, If 0, Class=stor., Driver=usb-storage, 480M
|__ Port 8: Dev 2, If 0, Class=stor., Driver=usb-storage, 480M
Hmm, now it is device 5, so I guess it is not out of 4. I don't know what
the port is though. My cat /proc/interrupts looks like this:
16: 1881681 IO-APIC-fasteoi uhci_hcd:usb2, uhci_hcd:usb5, nouveau
17: 1985740 IO-APIC-fasteoi eth0
18: 66775 IO-APIC-fasteoi ata_piix, uhci_hcd:usb4
19: 0 IO-APIC-fasteoi uhci_hcd:usb3
20: 210124 IO-APIC-fasteoi snd_ens1370
22: 25879534 IO-APIC-fasteoi snd_ice1712
23: 1281786 IO-APIC-fasteoi ehci_hcd:usb1
The interesting thing to note is that when I told the BIOS _NOT_ to give
my USB devices IRQs The kernel does so instead and USB1 (as you can see)
now has it's own IRQ, it used to share 16... Not so good as it is the only
high speed USB device I have.
> On a different note: If I were to get a new soundcard instead, does
anyone
> have recommendations of non-USB cards? There seems to be some issues
with
> PCI-cards and newer motherboards (I have a z77 intel chipset or
something)
> that uses emulated and not native PCI. I guess this resricts me from
using
> PCI-cards? (referring to
> http://www.linuxmusicians.com/viewtopic.php?f=6&t=10106 ).
> I'm using 48khz and 24bit right now, and I record mostly acoustic guitar
and vocals (both singing and rap vocals). Is there any reason for me to
run
> 24bit, maybe I could get away with 16bit? I do not really want to reduce
quality, even by a little, so if there's any reason not to go 16bit I
probably won't.
If you already have a 24bit device, get it working. There are not very
many PCIe sound cards around and those that are seem to be (overly)
pricey. Even a PCI 24 bit card will probably cost you more than a new USB
card. I think 24bit is worth having, I just figured you were running
USB1.1 at 16 bits right now. I was mistaken.
--
Len Ovens
www.OvenWerks.net
On Tue, March 5, 2013 8:45 am, Gabriel Beddingfield wrote:
> On Mon, Mar 4, 2013 at 3:26 AM, Gabbe Nord <gabbe.nord(a)gmail.com> wrote:
>> Hmm, alarming! I do use both USB keyboard and USB mouse. My soundcard is
>> also USB. I do have alot of other devices plugged in aswell to my USB.
>> Is
>> there any way I can fix this, other then getting "real" plugs for the
>> USB
>> keyboard/mouse? Like I said, there's still a couple of devices that uses
>> USB
>> even if I get rid of the mouse/keyboard.
All that on two ports? On the same irq?
> I'm not saying this is your problem... but use something like 'sudo
> lsusb -t' and inspect which devices are connected to which hubs.
> Unplug as much as you can. Try different ports.
He has only two USB ports for audio/midi, kb, rodent... and two more things.
Depending on price/needs, my first option would be to get an internal
card, PCI or PCIe. If you are using a USB1.1 device at 16bit/48k anyway, I
would suggest the original ensoniq es1370 cards (should be able to find
free) as they are reasonably quiet, have midi and just work with just
about anything. For 24bit, look for the ice1712 based cards. You will
probably have to pay more for them though. If you want to make use of the
USB device you have, get a PCI(e) USB port. It should have it's own irq if
it goes in the right slot. After an es1370 this is probably the cheapest
solution... you may even find one of these for free too from back when USB
was just coming out.
--
Len Ovens
www.OvenWerks.net
On Tue, 2013-03-05 at 08:45 -0800, Gabriel Beddingfield wrote:
> On Mon, Mar 4, 2013 at 3:26 AM, Gabbe Nord <gabbe.nord(a)gmail.com> wrote:
> > Hmm, alarming! I do use both USB keyboard and USB mouse. My soundcard is
> > also USB. I do have alot of other devices plugged in aswell to my USB. Is
> > there any way I can fix this, other then getting "real" plugs for the USB
> > keyboard/mouse? Like I said, there's still a couple of devices that uses USB
> > even if I get rid of the mouse/keyboard.
>
> USB devices are allowed to reserve bandwidth... and that affects every
> device attached to the same USB hub. Now, you might say, "I don't use
> a USB hub" -- but that's not true. Your PC has one or two USB hubs
> that it uses internally. In my case, I had a USB audio device on the
> same hub as a USB Webcam. Even when I wasn't using the Webcam, it was
> reserving bandwidth. Thing would be running fine, then all the sudden
> audio comes crashing down with an xrun.
>
> For me the PC had two internal hubs. I used a different USB port
> (attached to the non-webcam-hub) and the problem was solved.
>
> I'm not saying this is your problem... but use something like 'sudo
> lsusb -t' and inspect which devices are connected to which hubs.
> Unplug as much as you can. Try different ports.
Slot plates for using the USB ports provided by the mobo do cost around
EUR 2 only.
FWIW it's possible to unbind USB devices, assumed an unneeded port
should share the IRQ with something that is important.
