Compatible Audio Interface?

28 replies [Last post]
suddentwigs
User offline. Last seen 4 weeks 16 hours ago. Offline
Joined: 2010-04-28
Posts:

Hi there. Is there a definitive list of audio interfaces that will run under Linux? I am looking to put together a mobile set up and I am having trouble finding an audio interface that suits my needs and is compatible. Also I am wondering what hardware is required to run 24 bit recording at 192 khz, recording 8 tracks simultaneously. Planning to record on location and edit/mix on my existing home setup, so it shouldn't be too taxing for the laptop. Thanks!

LeatusPenguin
User offline. Last seen 4 days 12 hours ago. Offline
Joined: 2010-12-16
Posts:

The ALSA and FFADO project websites have a list of known working devices -

http://www.alsa-project.org/main/index.php/Matrix:Main
http://www.ffado.org/

If you're looking for a firewire device, FFADO is the place to look. For anything else, check out the ALSA project website.

linuxdsp
linuxdsp's picture
User offline. Last seen 1 week 2 days ago. Offline
Joined: 2009-02-04
Posts:

Despite what anyone might tell you there is almost no audible benefit to recording at 192 kHz. You can completely reproduce any audible frequency by sampling it at 44.1kHz
This might help in your choice of hardware.

LeatusPenguin
User offline. Last seen 4 days 12 hours ago. Offline
Joined: 2010-12-16
Posts:

This might also give you some food for thought -

http://people.xiph.org/~xiphmont/demo/neil-young.html

suddentwigs
User offline. Last seen 4 weeks 16 hours ago. Offline
Joined: 2010-04-28
Posts:

I have heard this about 192khz, but I haven't found anything conclusive about it. I have heard that it gives recordings a "pure" sound - is this just wishful thinking? Also, I heard ruled out firewire as I heard it was on the way out - is it worth rethinking this?

suddentwigs
User offline. Last seen 4 weeks 16 hours ago. Offline
Joined: 2010-04-28
Posts:

I have heard this about 192khz, but I haven't found anything conclusive about it. I have heard that it gives recordings a "pure" sound - is this just wishful thinking? Also, I had ruled out firewire as I heard it was on the way out - is it worth rethinking this?

linuxdsp
linuxdsp's picture
User offline. Last seen 1 week 2 days ago. Offline
Joined: 2009-02-04
Posts:

I have heard this about 192khz, but I haven't found anything conclusive about it...

I think the cited material is about as conclusive as you can get.

All signals with content entirely below the Nyquist frequency (half the sampling rate) are captured perfectly and completely by sampling

That's all you need to know, It's a mathematical certainty, and the limit of human hearing is 20kHz at best and for most not even close. Forget analogies with "stair steps" when sampling audio waveforms, that's not how the process works. The justification for 192kHz sounding discernibly better, is attempted mostly by people who want to sell you things, and by influential (non-technical) people they have persuaded to support their argument.

If you want to use 192kHz that's fine but, I just think its important to be able to make an informed decision as to whether its worth doing, or whether your money / time / processing power could be better spent.

linuxdsp
linuxdsp's picture
User offline. Last seen 1 week 2 days ago. Offline
Joined: 2009-02-04
Posts:

*and most of the examples of detectable differences in high bandwidth audio cited by those with "golden ears" can most likely be explained by interactions between the ultrasonic components of the signal, and other (electronic or electromechanical) components of the signal chain, which are not designed to reproduce those frequencies. Ironically, if they were designed to do so, you probably wouldn't be able to hear it.

LeatusPenguin
User offline. Last seen 4 days 12 hours ago. Offline
Joined: 2010-12-16
Posts:

I have heard that it gives recordings a "pure" sound

It's exactly these kind of vague descriptive terms that you should be wary about. If it can't be backed up with meaningful data, it most likely marketing talk.

seablade
User offline. Last seen 18 hours 58 min ago. Offline
Joined: 2007-01-22
Posts:

*and most of the examples of detectable differences in high bandwidth audio cited by those with "golden ears" can most likely be explained by interactions between the ultrasonic components of the signal

Heh I was going to post just this, glad you beat me to it:)

Though I will say I have heard of actually differences not when it comes to recording, but when it comes to processing that are a bit more reputable. Namely synths and reverbs and the like. Either way though the difference is so small that probably at least 99% of people will never actually hear it, and likely quite a bit more.

Seablade

audiodef
User offline. Last seen 2 days 12 hours ago. Offline
Joined: 2009-08-29
Posts:

Actually, digital signal processing is best done at higher sampling rates. I don't know that 192 yields better results than 96, but all of my Ardour projects are done at 96. You can't hear the Nyquist frequency, of course, but when you mixdown from sessions done at higher rates and compare them to sessions done at lower rates, I believe the difference is audible. So these sampling rates are more for DSP than for audibility while in your session. Check out Katz and Ownsinski's publications for more info. ;-)

linuxdsp
linuxdsp's picture
User offline. Last seen 1 week 2 days ago. Offline
Joined: 2009-02-04
Posts:

Actually, digital signal processing is best done at higher sampling rates.

As a general statement that's far too simplistic, and not always true. (without going into a lot of the technicalities, which most people won't read anyway..)

veda_sticks
User offline. Last seen 2 weeks 6 days ago. Offline
Joined: 2011-03-11
Posts:

there are some old digital reverb units that are fantastic that run at 16 bit with lower sample rates and they certainly dont sound great on there own,

howver when blended with a dry instrument just seem to work great.

James Brierley
User offline. Last seen 1 week 3 days ago. Offline
Joined: 2012-06-01
Posts:

I have heard that it gives recordings a "pure" sound

It's exactly these kind of vague descriptive terms that you should be wary about. If it can't be backed up with meaningful data, it most likely marketing talk.

See also ‘warmth’, ‘presence’, ‘punch’, ‘weight’, and every other infinitely malleable term beloved of the VintageSynth.com boys...

http://www.soundonsound.com/sos/dec13/articles/sounding-off-1213.htm remains one of my favourite articles about the confirmation bias of ‘audiophiles’. The ‘spectrophiles’ thing from the Xiph link was hilarious though; I’ll use that the next time I run into ‘one of those people’.

To the best of my knowledge, the majority of commercial recordings from the late 80s onward have been distributed and consumed as 16-bit 44100Hz digital audio. Rebecca Black too, but she didn’t sound as good somehow. Weird, that. Maybe I need my ears checked.

James Brierley
User offline. Last seen 1 week 3 days ago. Offline
Joined: 2012-06-01
Posts:

To actually reply to the OP, I recently bought a used Edirol FA-66 interface (FireWire) of which you can find many going cheap, and have been very happy with it, but it can only directly do 6 tracks in and 6 out. It can do 24/192 if you really want to. The important thing is that it is fully supported by FFADO, and mine hasn’t missed a beat since I got it.

seablade
User offline. Last seen 18 hours 58 min ago. Offline
Joined: 2007-01-22
Posts:

As a general statement that's far too simplistic, and not always true. (without going into a lot of the technicalities, which most people won't read anyway..)

Heh don't let that stop you, I am curious for your take on it. :)

Seablade

CraigPid
User offline. Last seen 2 weeks 1 day ago. Offline
Joined: 2013-02-15
Posts:

My take on the high sampling rates is the if they do really sound better then your playing and intonation better be perfect because it will also pick up the ugly stuff better along with the good. Since I don't have so much time nowadays to keep myself in playing shape,when I do get to work on music I concentrate more on playing what I want to play cleanly and with feeling.Just my opinion but I think it's easy nowadays to get lost in technical details and forget that first and foremost we're musicians (unless of course you're an engineer)

James Brierley
User offline. Last seen 1 week 3 days ago. Offline
Joined: 2012-06-01
Posts:

Quite, and it works in reverse too, as when a couple of years ago I read that Paul McCartney was complaining about ‘compression’, except that he appeared to think that an effect which reduces the dynamic range of an audio signal and the algorithms that make MP3/Vorbis/FLAC files small enough to be usable were the same thing. It might have been bad reporting, as you’d hope for a bit more technical knowledge from a musician of his vintage.

(Note for the OP: this has turned into a bit of a thread hijack, but nobody is calling you gullible/stupid/whatever for asking about sample rates. It’s the marketing folks who confuse the musician in the street who need to apologise, no one else.)

James Brierley
User offline. Last seen 1 week 3 days ago. Offline
Joined: 2012-06-01
Posts:

Oh, and to address the ‘FireWire on the way out’ thing: as long as PCI-E ports exist, I can see users being able to add a FW port to a computer that lacks one. In fact, users should be better able to pick controllers that actually work in that case (http://www.ffado.org/?q=node/251).

Additionally, one of the benefits of working with Linux is that open source drivers are not subject to being deleted to make way for some developer’s porn collection and thus lost forever. Once code is out there, it tends to stay out there, and peripherals tend to stay supported now that the Linux kernel is relatively mature. Yes, I oversimplify, but I have benefited in terms of the flood of ‘obsolete’ computer audio hardware hitting eBay following new Windows releases. :) Planned obsolescence is not the force it is in the proprietary software world.

Apparently some people have used FireWire to Thunderbolt adapters on Macs to run audio interfaces without problem (Thunderbolt sharing many design decisions with FireWire, such as DMA) but I can’t remember where I read this.

So, I wouldn’t worry too much about the FireWire factor.

nedzad
User offline. Last seen 1 week 4 days ago. Offline
Joined: 2011-09-06
Posts:

It is obvius that most of people ( 90% ) are forced to consume all that "products".
Technician ( audio engineer ) can really understand those parameter of sampling rates, oversampling and why, fourier transformation...... Neverendland :)
Yes, really understand and then "consume" ( i dont eat anny A/D converters)

So the idea of internet forum and "googleomania" is based upon supply and demand but not upon " i want to learn and understand". I believe that we are not able to share our knowledge because of "closed source pattern principles" made in our mind.
Industry "says" : If people take and consume, give them recipe ( this cable, that Software this Mic, those Monitors), not heritage and knowledge.
Merchant " says" : So lets make consumers as much as possible.
And now we can "see" how manny of us are listening that diferences in audio spectrum that are under 20 Hz and over 20.000 Hz and so on.
-----------------------------------
Gospel of DJ
Book by DJ "Analog-ist"

20:21
About Future.
And you go to buy a samples for your Kurzweil Sampler, you go into the Tempel of Professionals ( laaaarge music bazar)
and then you must wait for guru, mostly an old man, because only guru knows what is kurzweil sampler, SCSI HD, even CD ROM and 5 pin midi cable.

:)

nedzad
User offline. Last seen 1 week 4 days ago. Offline
Joined: 2011-09-06
Posts:

So, suddentwigs:
ECHO audiofire 12 would be nice
( but inputs are feemale TRS Jack connectors ).
This
Saffire Pro 10/I/O
-8 x XLR (mic/rear)
is much better for locations ( if you want, plug your 8 XLR cable direct to interface)
And Onyx Mixer FireWire Option is perfect if you like to "touch" that gains, EQ´s, faders......

Cheers!

suddentwigs
User offline. Last seen 4 weeks 16 hours ago. Offline
Joined: 2010-04-28
Posts:

Thank you all very much for your input! All extremely interesting reading. And I had never heard the genius of Rebecca Black before, so thanks for that. Nedzad - is that the model number of the Saffire Pro? I have been having trouble finding that specific one. Will it run under Ubuntu?

seablade
User offline. Last seen 18 hours 58 min ago. Offline
Joined: 2007-01-22
Posts:

I am not sure I have ever heard someone describe Rebecca Black as 'genius' before....

I feel sad for the world I think.

Seablade

suddentwigs
User offline. Last seen 4 weeks 16 hours ago. Offline
Joined: 2010-04-28
Posts:

Did I say genius? Sorry, I meant "genius". And by thanks for that, I meant "thanks" for THAT.

seablade
User offline. Last seen 18 hours 58 min ago. Offline
Joined: 2007-01-22
Posts:

Ahh ok, that makes much more sense and I feel much better.

Seablade

allank
allank's picture
User offline. Last seen 1 week 3 days ago. Offline
Joined: 2006-12-07
Posts:

I have the older Focusrite Saffire Pro 10io.. and just upgraded to an A&H Zed R16, not worried about the firewire thing as James pointed out above.. I have a T.I based PCIe controller and a VIA one.. To my surprise the VIA one was actually more stable (contrary to popular opinion).. I picked them both up on eBay for next to nothing. As I already had ffado installed (with all devices enabled as per default), I didn't have to do any kind of driver installation when I changed the devices over...
I'd highly recommend a firewire device that is supported by ffado.. (although check on the mailing list as the website is not up to date with supported devices.. e.g. the A&H Zed R16 is not listed, but is definitely working well)..

suddentwigs
User offline. Last seen 4 weeks 16 hours ago. Offline
Joined: 2010-04-28
Posts:

Things are starting to become a little clearer ... however I am having quite a time finding a laptop/interface combo that will work. Firewire seems to offer a lot more options for compatible interfaces, but I can't find a sufficiently powerful ubuntu compatible laptop that has a firewire socket. I am basing "sufficiently powerful" on the assumptions that I will need a 7200 rpm hard drive and at least a 2 GHz processor ... if this is overkill, I would be very pleased to hear it as it would open my options somewhat. As far as USB goes, the Presonus 1818VSL seems to get mentioned a lot with regard to Ubuntu, but ALSA themselves don't list it at all. Does anyone know about this?

jpnielzen
User offline. Last seen 9 weeks 1 day ago. Offline
Joined: 2014-08-18
Posts:

I'm grateful for this opensource DAW and other great software running on linux.

My setup:
M-Audio Delta 1010:
04:02.0 Multimedia audio controller: VIA Technologies Inc. ICE1712 [Envy24] PCI Multi-Channel I/O Controller (rev 02)

Miditech Midiface 4x4 USB midi interface:
Bus 001 Device 005: ID 1acc:1a0b

Running on a intel quadcore gigabyte-mainboard PC on Debian GNU/Linux "jessie"with rt patched kernel.
Linux vision-linux 3.14.12.04-rt9 #4 SMP PREEMPT RT Mon Aug 18 21:29:33 CEST 2014 x86_64 GNU/Linux

I use Korg Wavestation EX, Korg N364, M-Audio Venom (midi ports, no usb driver available on linux), Korg Wavedrum - and a Yamaha 5 string fretless bass.
I have used zynaddsubfx, rakarrack and just installed Ardour3 ... great!

I hope to learn, and give back what i can. Thanks to everybody who made this possible.

kind regards
Jens Peter Nielsen
Denmark

GMaq
GMaq's picture
User offline. Last seen 16 hours 14 min ago. Offline
Joined: 2007-12-11
Posts:

@suddentwigs

I copied my post from Linuxmusicians about this..

*Note these are my own experiences with the Presonus 1818VSL on a few different machines

- The latency performance is quite poor (ie 1024fpp+) unless you use an RT Kernel
- These units will occasionally just stop streaming audio with JACK, not often... but enough to be annoying
- The headphone output doesn't work unless you manually route the Audio to outputs 7/8 with Qjackctl or Patchage
- Many new computers with USB-3 have very poor audio performance with USB-2 Audio devices in general including the VSL
- The VSL1818 seems to be quite sensitive to USB chipsets in general, for example the VSL works OK for me on an older laptop, however my son has a slightly newer laptop (not with USB-3) and the VSL is too flaky to even use at all and we are using the same kernel/OS, I would say he has a USB hardware problem but he uses a Scarlett 2i2 with absolutely no issues at all...
- There is no hardware direct monitoring so you have to manually connect your inputs to the outputs in a JACK patchbay to fake hardware monitoring

I can't say it works flawlessly but it works decently... most of the time... occasionally :)