Need USB audio interface

[ThatJackElliott]

I'm new here. I'm the studio technician for a community radio station in Bend, Oregon, USA. I've been asked to set up our broadcast audio processing, which consists of two instances of Stereo Tool running on the same computer. One instance will provide aggressive processing for the FM broadcast, the second will provide gentler processing for the live stream. I need a USB audio interface with the following:

• Class-compliant (easy, no drivers)

• Balanced analog inputs and outputs (easy)

• At least one analog-to-digital converter input which will be shared by both instances of Stereo Tool (easy)

• At least two independent digital-to-analog converters, so each instance of Stereo Tool has its own output (not so easy?)

• A common clock running the adc and the two dacs so the two output devices don't drift out of sync with each other and we run into buffer issues (not so easy?)

My PC is a BeeLink SER running Ubuntu Studio. Budget is tight - community radio stations don't receive government money and are not allowed to do advertising, they are donation and underwriter-funded only.

Thank you!

[MattKingUSA]

I would recommend products that are made by Focusrite. I'm sure there are many others that would fit the bill. But in my personal experience Focusrite continually provides out of the box supported hardware.

EDIT: Also shout-out to BeeLink. They make such great hardware. I have several of their machines and they're really impressive little machines.

[ThatJackElliott]

Thank you.

Yes, I have used Focusrite in the past, and they seem to make a solid product. The Scarlett 4i4 looks like a candidate, but Focusrite aren't telling me whether the device needs special drivers, or whether it will "just work" under Linux and expose all four outputs to my JACK patchbay. Does anyone know?

[ThatJackElliott]

Thank you. I don't get the impression from their product pages that the devices are class-compliant and can be used under Linux/JACK without drivers, and they don't show any for Linux. Unclear whether the device will expose all its inputs and outputs to Linux.

[Babarosa]

Hi there!

I am using a MOTU M4 which works perfectly well with ALSA under Debian 11/12 when using kernel > 6, otherwise (kernel < 6) you'll need jack.
There is a MOTU M6 which "should" work perfectly well too. The technical specs of these units are at least very good.

See the specs here "https://motu.com/en-us/products/m-series/".

I am also using an ESI Maya22 USB which also works perfectly well with ALSA/Jack under Debian 11/12, but this unit only has 2 input/2 outputs. Technical specs are very good, the price is very attractive.

[Linuxmusician01]

Most, if not all, of the the Behringer U-Phoria's work without drivers. Beware: the latest revision of the 404 will only work properly with a very recent Linux kernel.

I don't think they're balanced in- and out though. And I can understand why that would be important for you...

About the DAC ... I don't know how that works in USB audio devices. Some of the U-Phoria's have Midas pre-amps which seems to be a big deal... But what DAC they use... ?

Good luck in choosing and let us know what you bought in the end.

[icaria36]

I'm getting a bit lost with some of your requirements but the Evo 16 has an affordable price and has plenty of room for future expansions in case you need them. I have been using it for a couple of months with Ubuntu + low-latency kernel and I'm very happy. If you have any questions specific to this model, please ask on this thread: viewtopic.php?p=157354#p157354

[ThatJackElliott]

Hey, thanks guys for the info.

"99% of all usb audio interfaces will work with linux.
Maybe you should do a lil research yourself at the alsa web page for example."

A reasonable suggestion! But it doesn't seem to be very well maintained, and doesn't address some of my questions. Like, balanced analog ins and outs? More than one output DAC, and do the ADCs and DACs share one clock? Perhaps all of these are a given and I'm being too fussy, but I don't want to buy a thing and find it's not gonna work. OTOH, I don't want to be that guy who pesters the forum with questions that have obvious answers. A thin line to walk!

[merlyn]

...

• At least two independent digital-to-analog converters, so each instance of Stereo Tool has its own output (not so easy?)

• A common clock running the adc and the two dacs so the two output devices don't drift out of sync with each other and we run into buffer issues (not so easy?)

...

An audio interface with multiple outputs would do this. Each output channel has its own DAC. If you think about it it has to.

I'm not sure about the signal flow here. If you're streaming to the internet you don't need to turn the stream back into analogue, do you?

You're right about the ALSA website -- it's out of date and not of a huge amount of use for recent hardware.

[ThatJackElliott]

Hi, thank you.

This is not primarily a web streaming setup. This is for FM radio broadcasting -- KPOV, a community radio station in Bend, Oregon, USA. One dac is to provide aggressively-processed analog audio for the "air chain" -- the signal that goes to the transmitter for broadcast; the second dac is to provide more gently-processed analog audio that goes to a distribution amplifier that feeds various devices: yes, a web streamer, but also a recorder for archiving purposes. Two instances of Stereo Tool (https://www.thimeo.com/stereo-tool/ are running on the PC I'm trying to configure: one provides the aggressively-processed FM signal, the other provides the less-aggressive one. So each one needs its own Playback device.

[merlyn]

... Two instances of Stereo Tool (https://www.thimeo.com/stereo-tool/ are running on the PC I'm trying to configure: one provides the aggressively-processed FM signal, the other provides the less-aggressive one. So each one needs its own Playback device. ...

Each one needs a stereo output, and you could do that with one interface that has four outs. Looking at the Stereo Tool website, it appears to support JACK, so you could use JACK to route one Stereo Tool instance to outputs 1 + 2, and the other instance to outputs 3 + 4.

The least expensive interfaces are Behringer, but, as Linuxmusician01 pointed out, a recent kernel is required for some Behringer interfaces.

[j_e_f_f_g]

Scarlett 4i4

This has 2 stereo output busses. So it technically meets your requirements. Just note that these are meant to feed professional grade equipment such as a crown or qsc amplifier. If you run these into consumer grade equipment such as a sony or panasonic amp you got from wallmart or target, it will probably overload (distort) the amp.

[erlkönig]

If so, why do you need to run it through the audio interface?

Usually you have to treat the streams through the air and on the internet differently, in Europe (don't know in the US) "via air" you have to conform to EBU128, while for a webstream you have all freedom concerning squeezing, making it scratchy and simply making the stream <lotsofechoandreverb>better and louder</lotsofechoandreverb>...

[j_e_f_f_g]

Yes, I just saw that in your earlier post. So you do need the 2 stereo busses.

conform to EBU128

Uh oh. If you want to make sure everything is going to work ok, the best thing to do is to see if you can get the manuals for your "distribution amplifier", your "recorder", and your "transmitter". Look in the back of each manual for the "hardware specifications". Scan these 3 pages so you can attach pics of them to your post here.

Then meryln will look at the hardware specs for a particular audio interface you want to buy, and tell you whether things (you don't need to know all about), such as impedance matching and output levels, are suitable. He can also tell you which outputs on the audio interface should be connected to your "distribution amplifier", your "recorder", and your "transmitter".

[merlyn]

EBU R 128 is about loudness. It might look like some digital interconnect format, like AES3, but it's not. Basically LUFS -23, dBTP -1. Fortunately x42 wrote us an EBU R 128 meter.

Whatever the standard is in the US, it will simply be different values of the LUFS and true peak.