Out of all the interfaces mentioned recently as working well under linux,
all I have looked up seem to still use potentiometers to set gain. That
is fine for recording mono instruments or vocals, but for stereo or
surround acoustic recording I would really like an interface which allows
for very close gain matching between channels, or at the very least very
repeatable gain settings so that I can correct in post production if
needed.
Are there any interfaces which are commonly available and have either
accurate electronic gain setting, or stepped/switched resistors for gain
setting?
The new MOTU devices which are class compliant seem to, starting at around
US$ 1200-1500.
Is that just the price range I have to move to before that feature is
included? I thought Cymatic Audio was going to have a 4 mic amp/2 line
out device for around US$ 500, but it was cancelled before release. That
is closer to the style of device I am looking for.
The MOTU devices without network audio (USB only) seem to have the feature
set and lower price, but has anyone checked whether they can be used
without a proprietary control application? The nice feature of the AVB
interfaces for linux use is that they can be controlled from a web
browser, so no proprietary application needed to set gain and routing from
a linux machine.
The webpage for one of the USB interfaces mentions a template for TouchOSC
which can run on an iPad, but I don't know how iPad software sends OSC
commands. If there is a standard for sending OSC control messages to USB
audio interfaces perhaps it could be adapted to linux. Any of the OSC
users familiar with whether that would be some kind of proprietary
interface, or whether it should be possible to interface one of the linux
OSC applications to the MOTU interface?
--
Chris Caudle
Hello,
Everytime I start vlc to play an audio file there's a short glitch over
perhaps 500ms then afterwards everything's fine. This happens every
single time vlc is launched to play an audio file.
Technically speaking, what would be the reason for such a recurrent
glitch ?
Cheers.
Hi,
I have to apologise, this mail is going to be long, so I'll try to
structure it a bit.
Overview
I have been live streaming on twitch for the past year, mostly games
and purely as a hobby. The possibilities of PA and OBS for routing and
audio processing are rather limited, I'd like to try and improve my
setup by adding JACK to the mix.
Current Setup
2-3 Audio interfaces. My main one is the UA-25 usb interface. There is
another in the GPU connected via HDMI to the monitor and a usually
deactivated one on the motherboard.
I am mostly streaming gameplay, most if not all games rely on ALSA or
PulseAudio, so I have PulseAudio running. In OBS I simply capture the
UA-25 "desktop output".
I also have it capture the UA-25 inputs, of which only one is connected
to the microphone. There is an option within OBS to have that signal on
both channels.
I use the compressor plugin built into OBS with pretty much "default"
voice/voiceover settings.
Afterwards "desktop" and voice get mixed to a stereo signal and
streamed.
I monitor the "desktop" signal as it is immediate and doesn't contain
my voice. To check whether the mix is fine I need to check either the
outgoing or the incoming signal (which is delayed by a few seconds).
Currently I route the incoming signal via PA to the GPU audio
interface. To actually hear it I need to physically switch the
headphones for the in-ears connected to the monitor.
I do not use a cam and intend to keep it that way.
Current Problems
A) Monitoring is cumbersome because I need to switch between the proper
headphones and the in-ears. While the in-ears aren't bad by themselves,
the sound through GPU -> HDMI -> Monitor is pretty bad. It is barely
enough to judge whether or not the levels are totally off.
I'd much prefer to be able to monitor using my proper headphones at the
push of a button.
B) OBS comes with rather few audio plugins by default
(https://github.com/obsproject/obs-studio/wiki/Filters-Guide) and I'm
not convinced they are more than serviceable. The compressor is hard to
set up, I wish it had at least some visual indication for the threshold.
Possible solution
While there are a few effects rack for Pulse Audio, like
https://github.com/wwmm/pulseeffects, and Pulse Audio can do routing I
don't understand PA's routing and rather use JACK. Since I haven't
followed Linux Audio land quite as closely the last few years I have a
few questions.
1) Can someone tell me whether the OBS jack plugin is good enough or
does something stupid? The code can be found here:
https://github.com/obsproject/obs-studio/tree/master/plugins/linux-jack
2) Is there a reliable and if possible hassle free way to get ALSA and
PA into JACK? ALSA loopback would probably not work for programs using
PA directly.
3) I think switching around monitoring should be doable with a simple
script calling jack_connect, provided the ports are predictably named.
Even I should be able to do that. Is there another solution?
4) Even though it's a small setup I'd need some sort of session
management to bring up at least the jack side of things.
5) Finally what do you recommend as plugin host and compressor plugin
for voice?
Thank you in advance for your help. I can at least document the setup
somewhere, for the next one who tries to do something like that. I
could do that on my blog (that no one ever reads) or if you know some
good wiki or something I can put it there.
Best Regards,
Philipp
