I do have the RIAA amplification set to high gain, so that is not the issue. It's enough to connect it an Apple device to restore it's proper volume. In a previous thread, RME has informed me that this low volume is set by the USB connected device. That means the Linux is lowering the volume of the ADI 2/4.

Does anyone have any knowledge on how I could fix that? I've tried ChatGPT, followed many instruction on how to setup volume at 100% in the Linux OS (Dietpi), but nothing works.

At the moment, Fireface devices work quite well on Linux in class-compliant mode, although there are some limitations. For example, the TotalMix FX application is not available for Linux.

However, the Oscmix project (especially the Qt version), which is an independent and unofficial open source project not affiliated with RME, already offers very promising support as an alternative to TotalMix FX.

At the time of writing, the project is still in an alpha stage. That said, for example, the Fireface 802 already works quite well with Oscmix.

It is definitely worth checking out and supporting the project. You can do this by testing it, sharing your experiences with the developer, and reporting any bugs you encounter.

Oscmix also runs on macOS, so feedback from Mac users can help improve the project and benefit users across both operating systems.

Please note that RME is not involved in the Oscmix project and therefore cannot be held responsible for its functionality or for any potential issues or damage that may result from using it.

Of course, the availability of official support is ultimately up to RME.

https://github.com/huddx01/oscmix

I have since moved to Linux. Quite a while back, when I tried the ARC in CC mode (on my UFX+, with an older firmware version I believe), I was only able to control the volume of outputs AN 1/2, and could not seem to find a way to have the ARC primarily control the AES output instead of AN 1/2.

At the time, only the rotary encoder worked as volume control, and no other functions worked, but from what I gather, with the Madiface XT II, as well as in new firmware versions, the ARC support in CC mode has been expanded, correct?

I am now at the point where I have to choose whether to wire up my patchbay so that my Madiface XT II outputs into my monitors, or my ADI 2 outputs. My main question is, in CC mode, with the ARC plugged into the interface directly, am I able to set it so that the rotary dial adjusts AES output instead of AN 1/2? I imagine either one should work okay, since both ADI2 and Madiface XTII DA converters seem quite good quality.

In addition to that, according to the manual, it looks like buttons 1-6 also can recall saved mix snapshots in CC mode, which is a huge blessing. Thanks to the RME developers for implementing this. In addition to this, are there some other ARC features which were also added to CC mode which I should be aware of? (Does the talkback functionality work?)

Thanks for the information!

My biggest concern prior to fully migrating from Microslop Windows 10 to a Linux host was whether I would be able to get my old RME Audio Multiface II to work in Linux, since RME Audio does not officially support Linux.  Thankfully, I found a post in this Linux RME Forum from 2020 providing valuable information.  (As somewhat expected, running Linux VMs under a Windows host is very different than running a Linux host, from an audio perspective.)

I "do" have audio working somewhat... i.e., I can play audio in a Linux browser through my Multiface, I can record and play audio in Audacity... I have not yet installed Reaper (that I used in Windows), Ardour or other audio applications.

My BEGINNER QUESTIONS/ISSUES:
Although my audio 'works' now, in Audacity, for example, I must select HOST = JACK, RECORDING Device = pipewire, PLAYBACK Device = system.
Each time I reboot the pc, I must manually restart the JACK server using the command jack_control start, even if I already started JACK using Cadence.

(1) Is it normal that I cannot select the same Device for recording and playback?

(2) Apparently starting JACK in Cadence is NOT the same as starting jack_control via CLI?

(3) How to auto startup JACK each time I power up Linux?

(4) Long-term... would I be better to eliminate JACK and fully use Pipewire and Wire Plumber?

(5) The Linux version of HDSPMixer (Hammerfall DSP) is not as 'full featured' as the Windows version.  In Windows, I could select an output mix and set channel 'sends' to that output mix, but in this Linux version, this doesn't seem to work.  Am I missing something?

(6) Should I stop fighting this battle, trying to get my Multiface II (Firewire) to work in Linux and just buy a new USB Class Compliant interace?

Thanks in advance for any guidance.

As AI is  integrated in Mac OS and Windows it is very possible that new productions get used as training data for AI audio generators.
I am running a professional recording studio and I would love to switch to Linux to provide a solution for our pro users, and i have absolutely no understanding for the ignorance of the audio industry. At the moment we are running Win10 without an internet connection, which simply sucks.

Come on Mathias, i am working with your hardware since the mid 90ies and i have always been proud of your companyś products and support. Especially the guys in Mitweida have been doing a great job.

PLEASE consider releasing Totalmix for Linux. There is no other way i can see for the audio industry to survive exept with an open source operating system. There is a discussion about forming a group financing a lawsuit at an EU court against audio companies that do not support open source OS.
To me that is simply not the way to go, but because i am in the same boat i can understand the frustration my colleagues are going through.

I have just written an article for Sound & Recording magazine if someone is interested in more details about this topic, it should be released in the march edition.

Arno Jordan
Castle Studios

First time a screenshot of a Linux application takes place in the user manual of the Digiface USB on page 30

Upon plugging in my UFX +, I saw that Pipewire was only able to 'see' 12 inputs and 12 outputs, regardless of the settings I dialed into the UFX +. So, seeing that the UFX 3 supports far more input and output channels in CC mode, I have considered selling my UFX + and upgrading to the UFX 3. However, since it is a large purchase, before spending the money on the upgrade, I was hoping to verify a few things about using the UFX series in Linux in CC mode:

1. The UFX 3 would be plugged into a Ferrofish A32 via MADI, and an ADI 2 Pro FS R Black via AES. This gives me an additional 34 outputs, on top of the built-in 8 outputs. I mix mostly out-of-the-box, and my console is 36 channel plus 4 aux (and 2 for monitors), so I need at least 42 outputs, preferably at 96k. Needs for input channels is less- I only need about 24 inputs at most. At the moment, the UFX + is able to do this in Windows 10 without any trouble, but since I am not interested in staying with that OS, is the UFX 3 capable of this channel count in CC mode in Linux? In CC mode, I assume the MADI and AES-connected converters should still function properly, correct?

2. Seeing that "Headphone Mode" copies the output 1-2 to 9-10 and 11-12, this assumes outputs 1-2 are used as the primary master outs if one does an in-the-box mix. Is it possible to set a different pair of channels as the primary monitor out in CC mode?

3. I assume push-to-talk talkback functionality would not work in CC mode, but this could likely be 'faked' by simply creating another configuration which monitors the talkback mic to the headphone outputs, and swapping in and out of that configuration, correct?

4. Does anyone know if the converters in the UFX 3 are the same as the ADI 2 Pro FS R Black? Currently, my monitors are plugged directly into my ADI 2, that way I can use my best converters for monitoring, but if the UFX 3's converters are so close in sound to the ADI 2 Pro FS R Black, then it may be easier for me to set channels 1-2 as monitors, channels 3-38 as the console, and so on. Reaper allows for rearranging and renaming of outputs and inputs, so I'm not too particular about what numbers are assigned to what task.

5. Lastly, I am using the "Cables" flatpak app for routing signal around. It works well, but is a tad rudimentary. I have seen that one exists for the UCX II (https://github.com/michaelforney/oscmix), but does anyone know of such a thing which can also work with the UFX 3?

Thanks for the help everyone.

I’ve bought a second NVME and intend to install Nobara  https://nobaraproject.org/ for everything else, Gaming , Surfing, YouTube, Office etc…

Kernel version is 6.14.6

I have a RME USB Digiface…Will this work for Desktop audio on Linux? YouTube, movies , Flac playback etc?

As I understand it Digiface works in class compliant mode…?

I’ll wait patiently for Linux drivers and USE windows cause I have to for real time audio and multitracking in my studio and Dual boot by using the abios Boot menu.

Thank you.

On Win10Pro, Reaper, UFX, USB connection, I am able to use 64sample buffers, which is very useful for live keyboard playing into Reaper, with both h/w synth and VST layer elements to the sound/MIDI recorded.

Somewhat unimpressed by Microsoft's TPM/Win11 proposition, and before spending £1000's on new hardware just to suit TPM, I thought I'd install a new 2Tb SSD, and make a serious attempt to move to Linux. 

I chose Ubuntu25.10, without any motive aside from it seems to be mainstream, and lacking ANY prior experience with Linux.  My logic is that it is the Linux Kernel on test, rather than any specific distro.  Provided it has RT potential, that should be basis enough.

To summarise a couple of days of this test, Ubuntu25.10 easily and competently replaces Win10 for office, email, browser, network, printer.  I have access to all my old files, once mounted.  VSCode, Reaper, Signal, GnuCash all load easily enough, either from App Center, or running .sh scripts.  Audio less so.

The more challenging things, like installing a newer version of Wine, are a struggle, if simply following the instructions at WineHQ - they are out of date, or perhaps incomplete.  I got errors about keys, out of date commands, and so on.  So here I figure is where someone more experienced in Linux would have done better.  I do not mind the learning curve, but it would be helpful if the how-to notes were correct!  I will likely come back to Wine, and Yabridge.

Now, I would have thought, if the original Win10 SSD is mounted, reinstalling VSTs would be pointless, and Wine mght be able to simply point at the files on the original Windows SSD?  Jury is out.  Every tutorial I have seen, naturally, expects you to install the VST again.  Small price to pay for not losing the whole VST estate I suppose, but we have over 650 of them to move, if so.  If someone has done this - or thinks it's possible - that would be useful info.  Maybe it needs reg keys recreated as part of the install.  Who knows.

Crucially to this forum, the UFX (a mk1) was set to Class Compliant mode (CC mode), and provided 22x22 channels, although Ubuntu does not let you choose the correct pair of said 22 for audio out under ALSA at least, so, YouTube video audio (for "how to" resources) were sent to a VDU speaker output.  It worked for that. 

In Reaper, things were more hopeful.  All 22 ins and outs were listed under ALSA.  It is very similar to the Win10 version in most regards, although I seriously miss the convenience of ASIO and all 30x30channels.  Half my connected audio kit is missing - I am using both ADAT ports for line in/out to rackmounted gear.  One ADAT unit's I/O is simply inaccessible in Reaper.

The low point for me is that the audio buffers have to be a huge 256samples, compared with Win10's 64samples.  There is something called "periods" in there as well, and I have little notion of what is best at this stage of the game.  It produce high pitched digital tones when it messed up the audio. 

v7.52 of Reaper seems to have some nasty UI bug with menus freezing - the forum there shows others experiencing this.  Bugs acknowledged.  One version to avoid.

So, my conclusion, is that in the absence of an RME kernel audio driver, Ubuntu underperforms compared with Win10.  I miss the control panel (Total Mix UI) because I use the submixes and matrix a lot, and run OSC commands to it, using OSCii bot. 

However, I am not persuaded by Microsoft and Win11 at all, so appear to be at an impass.  Never has something so offensive provided more motivation to jump ship.  Hence, I shall likely move all office/text/code to Linux, and only remain do audio on Win10, for the lack of an ASIO-esque driver suited to the main RT Kernel.  How ironic, that it is for the want of a high performance audio driver, I cannot completely make the jump to Linux - and that is the ONLY reason, at least for my PC usage, with the studio/live/business.  Not a rant, just piqued!

I would be most interested to know if other people can get better buffer settings working.  That would help.

I would also like to know what potential the mainline RT Kernel has for future drivers - could Linux achieve the same performance as Win10 theoretically?  Or is ALSA just not performant / yet?

Perhaps it would help to vote up one RT Kerrnel to play with as a community, irrespective of personal distro preference, as a test bed in such things.  But, then I am new to Linux, and this is for more capable users to chime in.  Here to learn.  What's your thought?

We have exceeded the 5% Desktop Share in U.S..

Furthermore, a important driver of that, DHH, is is confident this will be 14% or more in the foreseeable future..

Excited about the future of RME devices!

-------------------------------------

More thoughts if you want to read more:

I myself am now - for the first time in my life - on a macOS device: A Macbook Air M4.
What held me back from Apple was the inflated pricing. Now in 2025, I am convinced the Macbook Air is the best bang for the buck *hardware* that is available - without question, imo.
Now that I have an RME compatible device, I got me an RME babyface. What a wonderful device! Thank you guys!

However, as a lifetime Linux user it is obvious that

a) power users / techies / nerds will be disappointed by the software. I hope that Apple will do an in-house fully customizable tiling window manager. I have been an avid user of i3, am now using sway on Fedora). UX - in comparison - is bad. Still ok on an absolute level. Still better than Windows - in most, but not all (!) aspects. (I never thought I could say something like that ^^).

b) the quality of Apple Software and it's ecosystem seems troubled, from my POV: I have seen more bugs using my Macbook for a few months now than I have in the last 5 years or so of using Fedora. (Still not as ridiculous as Windows, though )

Maybe Apple will get back on track. But that won't happen with Tim Cook, I think.

Anyway.
[s]I read a long while ago somewhere here that RME will not support Linux as long as the 5% mark is not exceeded.[/s]
[Rephrasing that, because of the replies and because of .. it was badly phrased / wrong:]
I read a long while ago somewhere here that RME will work on full Linux functionality (imo that includes a Linux version of Totalmix!) when that 5% mark is exceeded.
I have seen a few thread titles coming here that indicate: RME seems to be on track - some devices seem to work in an unmediated / complete fashion as on the other OS choices or at there seems to be an official effort to have it that way.

Nice.
Excited about the future of RME devices

I greatly appreciate the sound quality, stability, and converters that RME hardware offers. However, one thing many Linux users (myself included) really miss is the native Linux support for TotalMix FX / advanced DSP (routing, on-board EQ/compression, low-latency effects, routing matrix, etc.).

Here are a few points I hope you’ll consider:

1. Market demand: The RME forums already host many posts over the years requesting Linux support (e.g. “Official Linux drivers/tools a possibility?” thread)

2. Value to users: With a native Linux TotalMix FX, many more professional and pro-audio users would confidently choose RME hardware for Linux-based setups.

3. Technical feasibility: Because the devices already support class-compliant modes, this request is about exposing and controlling DSP features via a Linux GUI or API, not changing fundamental hardware.

4. Willingness to help: I, and many others in the community, would be happy to test, provide feedback, or even help with documentation if RME provides partial SDKs or communication specs.

What I hope for:

A Linux-compatible version of TotalMix FX (native app, Flatpak, AppImage, etc.)

Or at least a documented control API / protocol so the community can build compatible tools

Activation of advanced DSP (routing, EQ, effects) under Linux, not just class-compliant I/O

I believe this would significantly strengthen RME’s standing in the pro audio community across all platforms.

Thanks for reading. I hope RME will consider adding (or at least exploring) native Linux DSP support.

Best regards,

I have a RME Hammerfall PCMCIA card (CardBus) with a Multiface I on a current Windows 11 PC. It works happily through Windows drivers, also ASIO (Cakewalk BandLab / Sonar).

However, I cannot get the "jack" audio server (e.g. used by Ardour) to work with the intended "portaudio" ASIO driver.
There is distorted output to the card without any application running that loosely follows audio input but gets chopped up (ring modulator sound). This looks to me like there is a "disagreement" in the buffer memory layout e.g. interleaved vs. block.

Settings are aligned between "jack" and the driver's control panel e.g. 48 kHz, 4096 samples and I think I have exhausted the obvious knobs like non-realtime mode (but please prove me wrong...)

Is this, by any chance, a known issue with a workaround?