Re: Official Linux drivers/tools a possibility?

fierylungs wrote:

I would even financially support Linux support. That is, if RME were to fund raise or even charge for a driver, I'd throw down in a ♥ beat.

Me too. If a payment wall is the only option to get Linux support, then it is. As long as the price is fair. I’ll be happy to pay for it if I know things will work out.

Re: Official Linux drivers/tools a possibility?

RME Support wrote:

Mind that the UCX does work perfectly fine with Linux in CC mode - sans Totalmix, but otherwise it can be used as it is....

Regards
Daniel Fuchs
RME

Just trying to understand this statement. So "class compliant" means that interface will work on linux and it's just Total Mix that is the missing link? If this is the case then doesn't that mean that I can get all the routing etc. sorted in UCX or UFX when connected to a Windows computer and then just run in Linux? That might be perfect for my application which is for the UFX to be part of a convolution engine for a speaker system.

Re: Official Linux drivers/tools a possibility?

Quote: “I can get all the routing etc. sorted in UCX or UFX when connected to a Windows computer and then just run in Linux?”


I don’t believe it to be rocket science.

And recall that the Hubble space telescope ran on DOS.

But friends working for Elon Musk inform me that using a number of programming languages for differing facilities, Space X’s Falcon 9, Dragon 2 and Starlink run on Linux, (albeit a stripped-down version).

There is an essential salient point to actively consider here and it’s that there are very substantial, totally practical, reasons, some, safety related, for utilising a Linux Operating System, as opposed to a Microsoft Windows Operating System.


The main one being that in outer space, astronauts are not able to open Windows.

Re: Official Linux drivers/tools a possibility?

I need these drivers for Debian and RME has had nothing but excuses since 2018. For 3 years you have been too busy to compile a set of drivers for one Linux distro? Not one has been released in the last 3 years!

I will NOT be buying a new computer just to install Winblows 11, and Winblows 10 has a deadline to shut off support in a couple years.

Linux support MUST become a priority if you want to retain your customers. Making good audio interfaces only goes so far if you don't care about your customers post sales.

Re: Official Linux drivers/tools a possibility?

Quote: “Making good audio interfaces only goes so far if you don't care about your customers post sales.”


With respect, this statement is just about as totally mendacious as a statement could be.

RME’s stellar reputation is built upon a solid history of designing, manufacturing and supporting its products faithfully, decades after other manufacturers would have ceased to.

If you have a good point or argument to make, make it in manner that is truthful and can be respected. In a way that is convicting and convincing. But you won’t achieve that by positing clearly recognisable, blatant untruths.

If I might seek to help you out here as you clearly need assistance, the biggest weakness in your argument would appear to be the fact that when users purchased their RME equipment, they did so, in the crystal clear knowledge that Linux was not supported.

But still went ahead and purchased them none the less, recognising RME’s superlative balance in quality and value.


After they were purchased, a relatively small number users made the unilateral decision to move to an entirely different computing operating system. An operating system that RME has never supported.

Subsequently complaining that the product they knowingly purchased did not support Linux Operating System, does not support Linux.  Such a scenario scripted as a play, would normally be described as a farce.

What’s really at work here, is the simple economics of market forces in a small company with strictly limited resources.

It has to remain focussed upon its core business opportunities.


Its utterly unreasonable that simply because someone decided to change their operating system, they should complain that a product they purchased knowing the operating system they decided to change to was not supported, means that “RME doesn’t care about its customers or support its products”. Its compete and utter tosh!

In Shakespeare play Hamlet, Queen Gertrude responds to the insincere overacting of a character in a play within a play by stating:

“The lady doth protest too much, me thinks”.


It would be more sincere if you admitted, “my product was never designed or intended to be used for Linux, but I changed to that Operating system anyway, so now, I have created some thing of a problem for myself.”


I fully appreciate that there are people who would like Linux drivers.

But with all due respect, if they complain, they do themselves a great disservice.

1: by virtue of their clear fore knowledge and understanding of product specification pre-purchase.

2: The poorly argued case they make, not taking into account the market place economic factors governing such business decisions.

3: Andi n particular by miscreant attempts of transference of problems arising from their personal subsequent decisions to RME, falsely describing the company in a poor light.


The only people they actually show up in a bad light light is themselves

Take personal ownership and responsibility for your decisions.

Don’t blame RME for your personal subsequent choices.

Its unreasonable and unfair to a good company.

Re: Official Linux drivers/tools a possibility?

Linux support cannot be obtained by blaming each other and the manufacturer here. Instead, we could discuss how much work is required to support Linux and how to get there. On the other hand, we might be wondering how to get more users covered (More Linux users). I know the situation is frustrating, but one day it will happen. So the thing is, Am I alive to see it. But surely one of the RME users is. The only thing we can do is hope and ask politely. And hope again. The only thing we can do is hope and ask politely. And hope again.

Re: Official Linux drivers/tools a possibility?

Quote: “The only thing we can do is hope and ask politely.”


My understanding is, that there has been a profound shortage of good, experienced programmers, for some considerable time.

Not simply good programmers, but also programmers whose salary expectations, directly correspond with the smaller budgets that specialist companies can afford.

I know programmers who are at this moment totally inundated with work, their hours elongated as the work continues to pile up. I am told that this situation has been exacerbated as certain large companies have stooped to poaching programmers from smaller companies, with remaining loyal staff, facing backlogs of work.

At times, the remaining programmers are permitted to have holidays.

Inconvenient to business as this is, it seems churlish to deny them.


There are genuine reasons smaller companies need to clearly focus on the absolute priorities for their business.

Almost everyday some bright spark has an idea for an extra feature they would like to be incorporated into a product, that few others, if any at all, require.

Whilst on occasion, some ideas are actually implementable with relative ease, they are likely to be the exception, rather than the rule, because of the reasons described above.

RME products seem to me to have an extraordinary advantage over many other manufacturers products. Its that an immensely profound amount of attention has been focussed on their initial design concept.

The targeted consumer for their product ranges, their expectations and requirements have been analysed with meticulous care and fully provided for.  Such that implementation of their product designs, pack so much into so little space, with such efficiency, that additions requested later, would often necessitate a complete redesign.

Although, it is also true to say that their brilliantly conceived designs often provide for future developments where possible, such is the excellence of their implementation.

The late Rupert Neve intimated “implementation is everything”.


Once series production is under way, to make even a most minor change in the major industries I have an interest in, would cost many, many, millions to implement.

What new product development, necessary existing product update, or customers service needs, do you believe RME should bring to a halt, in order to address the requests in this thread?

How much work, far more important to the sustainability, future development and expansion of their business should they abandon? How much will that really cost in terms of losing competitive edge to other manufacturers and lost timing in product to market schedules?

If they were able to afford to hire more good programmers and the cost distributed to existing Windows and Macintosh users by way of higher product prices, making their products less competitive in the market place, losing that unique niche’ of quality, features and value, would that be a responsible decision for a CEO to make?

And how fraternally equitable would that be for the overwhelming masses, all non-Linux users, the people that faithfully bought their product to use within its stated specification parameters, who would pay such a heavy price in such circumstances?


By the way, blaming people is not my style, pointing out the sophism in fallacious arguments is. So, here’s another I would respectfully request you to consider.

Quote: “we could discuss how much work is required to support Linux and how to get there.”


On the surface, this statement seems sweet reason itself.

However, it contains an implicit, underlying, basic false assumption.

That is, after discussion and work, it is possible to arrive at the sought result.

The foreseeable, unforeseen problem that belies this statement, is of course that having provided Linux Drivers. Daily, weekly, monthly, annually ongoing demands for support, future development, and customer service requests will require the services of highly specialised programmers to address.

We have already established, there is a world-wide shortage of such talented individuals.


Put simply.

With respect, you imagine that developing Linux Drivers is all that is required. You see it as an arrival platform. A final destination.

Whereas, in reality, it is nothing of the sort. Understood correctly, it is in fact merely a departure platform, the beginning of a journey.

And the entire necessary aspects of providing completely adequate resources, on an ongoing basis for an additional Operating System, are involved.


Please appreciate, I am not trying to argue with you.

Rather, to highlight the fact that the simplicity with which this issue appears to be viewed by certain Linux users, is wholly delusional.

The bigger picture of the world in which the commercial manufacturing industry thrives, is far, far more complex altogether. That is the world in which RME has to compete.

Re: Official Linux drivers/tools a possibility?

Well maybe the Linux community ends up doing things themselves and these devices are for them but just shells for their endless creativity.

There is now a huge shortage of components in the world, many things can change very quickly. The same goes for the software side. We are living in interesting and also scary times.

Opensource may show its power in a surprising way. Because product development is spread around the world. To countless projects and communities that interact with each other

59 (edited by strathamer 2021-12-04 08:19:22)

Re: Official Linux drivers/tools a possibility?

I also really need my hardware to be supported on Linux. It's crazy to read that customers have been asking for this support for almost 15 years and that it's being ignored. RME is recognized as high quality material and is not available on the most qualitative platform on the market. This is nonsense.

Everything in my hardware and software is supported on Linux except my Fireface 802 RME ! Please listen to your customers and give us a basic first version of TotalMix!

Re: Official Linux drivers/tools a possibility?

Hi there… just found this discussion. I wanted to add that once there were different times and I still benefit from those. I actually use an RME Digi PCI card since 2006. It does have official Linux support. This was the last card that didn't have TotalMix yet I think. It is remarkable how long-lasting this device is, my PC runs basically for many hours every day and it's the 3rd or 4th PC that this card lives in. Me and myself, I quit using Windows around 1998, and with the RME card in 2006 took the route to music production on Linux, using Ardour and nowadays Mixbus. I use those via ADAT with a Yamaha 01v96i in my small studio room.

Just recently I bought a Fireface 400 used because ffado-mixer on Linux actually does operate its TotalMix functions. Of course it is tough to get Firewire still to work, not due to the software, but due to the hardware – I use a PC Express card in an old laptop currently. It is sad that the current Fireface series can't give me this functionality via USB.

So far I had been using a Focusrite Scarlett 18i8, which in its early generations also has full Linux support for its matrix mixer. However, even though Focusrite makes good products, I wanted to go further in quality. After buying some excellent microphones I didn't want to be bothered by questioning my preamps and converters any more.

I'm not sure if this is possible, but if a Linux version of TotalMix could be open source, this might put part of the maintenance load off to the community. Like compiling packages for several distributions, or adapting to never versions of libc and whatnot. After all, the RME Digi driver, which btw. for Windows is available until Win 7 32bit, lives in the Linux kernel happily forever.

It's a sad thing though, that on Linux, developers usually don't understand the needs of pro audio users. I've had plenty of discussions on the quality of plugins and why I use Windows VSTs on Linux, they said it's all there as native Linux stuff what you need and I keep saying: Yes, but it doesn't sound good. And then it turns out that you'd need 10 years of coding training plus 15 years of pro audio experience to hear what I need… and I guess these are the Linux developers that are impossible to get. Me and myself, I quit any coding some years ago in order to be able to focus on audio (and video).

Re: Official Linux drivers/tools a possibility?

Hi everyone, I would like to add my thoughts to this issue. I'm a software engineer and do both some podcasting for fun and since I'm building several (language) learning Apps for customers I also do audio recordings for those apps to make them more interactive.

I would very much love to see a solution to access the DSP capabilities of RME devices from Linux. The Class Compliant mode is already a great step to be able to use Linux as a platform. Still many applications like podcasting or recording a dialog for an App with several people requires monitoring and mixing directly in the hardware which is not possible at this time when using Linux. This can only be done properly with Totalmix.

Because many RME devices now support CC mode it should not be required to build a low level driver for Linux anyway. CC is sufficient to read and write low latency audio streams to and from the hardware. No need to reinvent the wheel here. (still would be great to get CC mode on some more devices but let's not get greedy)

As software engineer I totally understand that it is not always easy to port a software to another platform depending on the architecture and frameworks used to build the application and thus it might be too much effort to port Totalmix to Linux. On the other hand Totalmix is already available on Windows and Mac OS and I would guess that there are not two isolated implementations but a single implementation using cross platform tooling and only platform specific implementations for the low level driver parts when necessary.

Still I can understand that RME does not invest the time and money to create a Linux version of Totalmix. Linux is still not as big as a target platform as Windows and Mac OS. But as other members wrote in this thread it should be possible for RME to release the specifications of the communication protocol between Totalmix and the device. Making these specs public should not reveal any internals of the hardware or firmware that might be considered company secrets. With the specs available there will surely be people willing to create an open source implementation of the communication protocol as well as a user interface that lets users access the capabilities of the device similar to Totalmix.

I do understand that you RME might be a bit reluctant to share those specs but I believe this would actually increase the number of RME users and would open new opportunities for additional revenue streams. For example the specs could be released under a license that only allows using the protocol in a non-commercial implementation and explicitly denying all official support from RME for this implementation(s). At the same time the license could allow commercial use but only when the implementing party pays a license fee. This might attract other parties that would like to implement hardware control solutions for the devices with more functionality than the Arc.

If RME would be willing to do this at some point I would love to work on an open source implementation and I'm sure there are others that would join such a project as well.

cheers tupel

62

Re: Official Linux drivers/tools a possibility?

Just adding my +1 for this. Would be really nice with official Linux support.

The mention of running Totalmix in a browser for a more system-agnostic use seems like an idea worth exploring. That means you could potentially also control your RME device via a tablet, phone or a laptop, just as long as it's connected to the same network.

Re: Official Linux drivers/tools a possibility?

If I could only get one Christmas present, it would be Totalmix for that.
https://forum.mxlinux.org/viewtopic.php?t=68136

Re: Official Linux drivers/tools a possibility?

+1 !!

I'm musician and sound engineer, I work mostly with sound creation, diffusion (spatialization), live performances, theater live music and sound design, also field and music recording.
I'm now working in a studio, where we plan to build a Linux/GNU system beside our totally-analog system and Pyramix.
I use mostly Ardour, Bitwig (which changed my linux-audio-life!!), Reaper, 1000 plugins, Pure Data, REW and other great tools.

I'm working with Linux, and ONLY Linux since 2006, and when I see where it is going with other OS, I'd prefer to stop digital production rather than change.

I'm happy to see a lot of people use and plan to use Linux professionally!

I need a new sound device since my good old HDSP Multiface has become unusable as my new computer won't provide Expresscard slot...
A bit sad about that: it was the most stable system I experienced with Linux, I used it >10 years, with almost no xruns, incredibly low latencies, never crashed Jack. No USB class compliant material gave me such results, even pro (01V96i...).

And it was a nice old time when RME and ALSA collaborated for a driver and graphic tools which work very good!!!

My first idea was to buy a small only-digital interface (like Digiface), with different converters / preamps following my different needs.
But there are no driver for Linux.
so I'm considering a Fireface, but the lack of control is a pain...

Some people seem to get tense, let's continue to be friendly and cordial.
Just understand we defend a system which was always considered in audio world "only for nerds", "experimental", "with poor tools". And now people admit it could be as valuable as others (no better no worse), but we are still and still missing serious hardware support.

Hope this helps!
an RME fan

Re: Official Linux drivers/tools a possibility?

100% agreed that we should be friendly people. Also we should understand that RME is a tremendously huge company, so I think they can't just do some stuff because they have staff getting bored otherwise. ;-)

The Linux audio thing is a sad chicken&egg story. There are plenty of good tools available these days. In the last couple of days I did a recording with RME gear and the next weeks will feature mixdown sessions in Mixbus and I won't miss a thing. However, I'm mostly mixing "real" recordings, so instruments via microphones etc. - talk software synthesizers or even Melodyne, you're quite alone on Linux still. Even though it's getting better. Linux is not an important market in pro audio, so there are not as many products for Linux, so Linux doesn't become a relevant market in pro audio…

Me & myself, I quit using Windows in 1999 or so and I won't be going back. But where will I got to once my RME Digi 9636 or my Fireface 400 are dead and gone?

Well I guess, it will be DANTE then. I ordered my first DANTE-enabled mixing console some months ago, but digital consoles and electronics crisis... perhaps I will get it somewhen in 2023? ;-)

It would be sad to have to do without RME at some point.

Re: Official Linux drivers/tools a possibility?

Will UAD be the first pro interfaces we can buy with propper linux support?
Apparently they are counting potential users...
https://www.change.org/p/linux-support- … igned=true

67 (edited by ramses 2022-01-18 10:31:13)

Re: Official Linux drivers/tools a possibility?

Thanks for sharing, here my personal view on this, in case that you are interested:
https://forum.rme-audio.de/viewtopic.ph … 49#p181649

BR Ramses - UFX III, 12Mic, XTC, ADI-2 Pro FS R BE, RayDAT, X10SRi-F, E5-1680v4, Win10Pro22H2, Cub13

Re: Official Linux drivers/tools a possibility?

Chiming in to say that I'd love to see first-party Linux support.

Re: Official Linux drivers/tools a possibility?

+1 for Linux Support - for the time being just ad CC Mode to the Interfaces that don't support it yet. Btw the Digiface Dante would be the only USB Dante Solution with more than two Channels on the Market. If that's not a selling point i don't know...

Re: Official Linux drivers/tools a possibility?

Linux is ok, but BSDs would be even better.  Particularly OpenBSD, which has outstanding audio support that is minimal and simple.

Re: Official Linux drivers/tools a possibility?

I too would love to see linux drivers for the Madi FX. Even if it would only be CC.
And i would pay for it too.

Re: Official Linux drivers/tools a possibility?

Another one for linux here.

Look, the official support and reasons have been discussed to death, no need to repeat that. But maybe things don't have to be so black and white. For instance, as far as I know, Focusrite Scarlett driver was upstreamed to linux kernel without the involvement of Focusrite themselves. Then another developer created a 3rd party UI with matching functionality of the official software. Perhaps something like that could be done here too? Maybe RME could throw 3rd party developers a bone in a form of some minimal documentation at least?

There's more than one way to skin a cat.

  • Open source driver could be upstreamed to the kernel. Alternatively if that's not possible - a binary blob kernel module, aka Nvidia approach

  • If the UI was developed in something like Electron - platform compatibility would not be an issue and that's not just for linux

Re: Official Linux drivers/tools a possibility?

At least I imagine that if RME's industrial secrets are not compromised and they don't have to invest financial resources in the project, that open source driver and the alternative "TotalMixFx" would be possible. I don't know why they haven't given, for example, the ALSA project that opportunity.

This is just my opinion and not based on any real information.

Re: Official Linux drivers/tools a possibility?

+1 for requesting Linux support. Hope that this will gain more momentum....

Re: Official Linux drivers/tools a possibility?

As a software engineer I'd even like to contribute to development of drivers and totalmix software for linux. It's a must have thing for such devices. Please make us happy! smile

Re: Official Linux drivers/tools a possibility?

akoppela wrote:

As a software engineer I'd even like to contribute to development of drivers and totalmix software for linux. It's a must have thing for such devices. Please make us happy! smile

As a software engineer - me too.

But the problem here is not so much technical as it is political. It's likely RME will have to start paying attention to Linux if not for other reasons then as a side effect of Windows decline. Windows has lost 17% desktop market share in the last 10 years. It's ceding market share to Apple, Chrome OS and Linux distros. Various software and hardware vendors that 15 years ago would have laughed in your face if asked for Linux support now are lining up to change their mind. Then there are geopolitical reasons for Linux growth - as in Russia, China and the like are less than impressed with dependency on USA's trojan horses M$ and Apple. I believe Linux support will happen with RME too, just a matter of time.

Re: Official Linux drivers/tools a possibility?

Why Linux?   It has a jungle of distributions and is mostly hacked together by college kids with some paid professionals mixed in.

A much more stable and high quality target would be OpenBSD or FreeBSD.    These are a very small market, however.

Re: Official Linux drivers/tools a possibility?

hellrazr wrote:

Why Linux?   It has a jungle of distributions and is mostly hacked together by college kids with some paid professionals mixed in.

A much more stable and high quality target would be OpenBSD or FreeBSD.    These are a very small market, however.

Same lame argument all over again... Forget about the distros. All RME could do is 2 things: the kernel driver, even if not open source then kernel module blob aka Nvidia and some basic documentation. I can guarantee you, weeks/months later someone will have implemented TouchMix for Linux on Electron or whatever that would work on all the distros just like all other similar apps do.

79 (edited by mrjulius 2022-12-14 16:27:18)

Re: Official Linux drivers/tools a possibility?

Exactly! Some people keep repeating the same old clichés about linux. The scene has changed a lot in a couple of years - over half of our company laptops are linux (officially supported by Lenovo), half of my friends and even my parents use linux. Yes, some of those friends have good computer skills - perhaps 10 years ago they would have been considered "nerds", but now they're just normal people.  My DAW runs commercial software (Bitwig and Reaper, to name a few) on linux more stable than it ever did on Windows. And setting this environment up is easier and takes much less time - I much prefer using my time actually making music, than jumping through hoops. My commercial hardware (Native Instruments, Novation, M-Audio, CME, ...) all runs great on linux.
Yes, my RME Babyface pro has CC-mode, but in studio it's just not enough.

Personally I've promised to purchase RME FireFace UFX II if it would have a driver from RME or community. Until then, I'm basically stuck with Behringer UMC1820 as my main interface, which is a shame.

80 (edited by ramses 2022-12-14 18:17:36)

Re: Official Linux drivers/tools a possibility?

mrjulius wrote:

The scene has changed a lot in a couple of years - over half of our company laptops are linux (officially supported by Lenovo), half of my friends and even my parents use linux.

TBH .. Nothing changed significantly.

We are getting more and more not really needed distros, as if there weren't enough.

Linux distributions do not reach the quality of former commercial Unixes and BSD variants.

Not long ago I wrote an AWK script which fails only on Linux because something somewhere is broken. On BSD no issues, code works as expected. Same gawk version, so most likely a library or compiler bug. Not so nice.

In a project I needed to rewrite all Makefiles because one needed Make instruction simply did not work on Ubuntu, on Centos it still worked. In both cases gmake.

All these little errors are unnecessary and simply annoying. Too many distributions and randomly put together tools are very unlikely to result into something called quality.

It would less likely happen, if there would have been something like one well maintained Linux distribution (best from the beginning) containing all the tools that you expect from a Unix system and this well maintained.
Only then everything fits smoothly together.

The so called bazaar concept is a lame excuse for lack of experience, no good vision and at the end missing standardization.

Lenovo will surely also focus on only one or two Linux distribution which they support, but not 5, 10 or hundreds of them.

To argue with friends, family, colleagues and "cats and dogs", everybody using Linux is a weak argument. Most people use only very few applications and only scratch at the surface.

Yes I also believe, that mail reading, browsing the Web and printing is possible with Linux and even more.

But recording / audio processing is different. And people would have higher quality expectations when it comes to RME and expensive hardware.

Instead of yelling at people who do not like this distribution mess for reasons you should better yell at the people who created this mess.

Ask Linus why he didn't deliver a base system such like any standard commercial Unix or FreeBSD or NetBSD.

To complain is easy for you. I could think that everybody would expect a high quality solution from RME also under Linux  which could be very challenging. Because ..  at times even Microsoft and Apple are challenging and this with only different versions / snapshots of the same OS.

How can you ensure that all runs fine on any Linux? To tell people run this only on Debian or Ubuntu or RedHat won't work.

As soon as one Linux is supported everybody will expect that it runs on his beloved distribution. If something would fail and you payed for the HW, then you want to have support for your Linux.

I am not against Unix. I only have doubts how all of this can be supported under Linux.

Maybe doable for free software or where the software is not that expensive. But what if you bought expensive hardware?

The only companies that I know, who sell expensive stuff and enterprise services for Linux are for example Oracle and RedHat. Guess why they built their own Linux distribution.

BR Ramses - UFX III, 12Mic, XTC, ADI-2 Pro FS R BE, RayDAT, X10SRi-F, E5-1680v4, Win10Pro22H2, Cub13

81 (edited by ramses 2022-12-15 17:27:46)

Re: Official Linux drivers/tools a possibility?

But nevertheless, I am also not satisfied with Microsoft and Apple.

Apple because of the price, too few HW  upgrade possibilities and screwing things up from time to time where it takes too long until they recognize, accept and solve issues.

Microsoft because they change too many things to the worse for no reason and the new security model, which sucks too much performance and forces you to trash good hardware for no real reason is a joke.

So at the end I also hope that RME finds a business and support model that would make it possible to support Linux not only by CC mode.

But also Linux should improve in terms of consolidation and standardization

A cool thing would be if Linus would get a vision to create a core team to build a solid Linux standard distribution.
You would need such an important person to initiate this.

But I fear he has zero interest and no vision for such an ambitious project to become the new OS standard on the market. Very ambitious would be the goal to deliver even better quality.

It would also be very lovely for companies, if things in Linux would not change too rapidly to have more continuity and stability for customers as well as vendors.

BR Ramses - UFX III, 12Mic, XTC, ADI-2 Pro FS R BE, RayDAT, X10SRi-F, E5-1680v4, Win10Pro22H2, Cub13

82 (edited by sjzstudio 2022-12-30 18:32:54)

Re: Official Linux drivers/tools a possibility?

If we forget everything else and think of this as a purely technical problem, is there some technical issue that makes it impossible to create official drivers and Totalmix fx for Linux?

I have understood that the drivers are Kernel modules, so they probably work in all distros.

Totalmix Fx works on the user base, so translating it for Linux shouldn't be a difficult thing if you can do it for macOS.

It is only a user interface program and does not process anything other than controller data.

Other tools are a separate matter

Note, I'm no expert, but this is how I rationalized it.

EDIT:
It's kind of comical that Behringer has been able to ship their Xair 18 controller software for linux. I tried it and all you need is to download and click the icon. Has Behringer hired any Linux gurus on the payroll?

https://linuxmusicians.com/viewtopic.php?t=24621

Re: Official Linux drivers/tools a possibility?

ramses wrote:

How can you ensure that all runs fine on any Linux? To tell people run this only on Debian or Ubuntu or RedHat won't work.

Why? Mind to elaborate? Please, enlighten me.

I personally can't see why that approach wouldn't work, Debian or Ubuntu sound great to me (I'm currently using Debian)



Best

Re: Official Linux drivers/tools a possibility?

sjzstudio wrote:

It's kind of comical that Behringer has been able to ship their Xair 18 controller software for linux. I tried it and all you need is to download and click the icon. Has Behringer hired any Linux gurus on the payroll?

https://linuxmusicians.com/viewtopic.php?t=24621

Today !
It works today. But how will the situation be in 2 years ?
Will Ubuntu and Debian still be the same ?
and all the other hundreds of Linux variations users? They should switch to Debian ?
When was Debian invented ?
Do you expect ALSA to be the audio system of Linux in 20 years ?
Soon the Babyface got 20 year birthday and drivers are still in development.

M1-Sequoia, Madiface Pro, Digiface USB, Babyface silver and blue

Re: Official Linux drivers/tools a possibility?

@waedi:

It works today. But how will the situation be in 2 years ?
Works if desired and drivers are maintained. Just as it is done in other systems.

Will Ubuntu and Debian still be the same ?
They aren't. They are constantly evolving, as are Apple and Microsoft products.

and all the other hundreds of Linux variations users? They should switch to Debian ?

No need. But it is not possible to support all possible distributions with 100% certainty. Therefore, one must choose something widespread. Debian is therefore a good choice because it is the basis of many distributions. And distributions may possibly start to support E.g. RME driver/software or lib base.

When was Debian invented ?
1993

Do you expect ALSA to be the audio system of Linux in 20 years ?

It may be, it may not be. Someone can develop a better one at any time.

Soon the Babyface got 20 year birthday and drivers are still in development.

Programs are always being developed until the sun stops rising.

Note, I am not an expert in Linux or anything else. I'm just trying to get the sound on the hard drive and for the audience to hear.

86 (edited by cupakm 2023-06-01 22:54:28)

Re: Official Linux drivers/tools a possibility?

psyndrome wrote:

+1 for Linux Support! (TotalMix and Digicheck)

Btw: would Linux drivers deliver less latency than running CC mode? (This is critical for my live looper setup)

Hi psyndrome,

I guess the latency would be comparable.

I'm running the original Babyface (USB ID 0424:3fb7 ) for around 9+ years in CC mode under Linux (either Debian or Ubuntu) with a low-latency kernel (currently 6.2.0-1003-lowlatency ) through a Renesas uPD720201 controller PCIe expansion card (into which I do not connect any other devices), with its original USB cable (yes, the cable matters), no JACK just ALSA, and am getting 3,85 ms RTL measured in Ardour at 48 kHz, 32 samples / 3 periods, without sound artifacts like cracks, pops, buzzes, etc.

So I'm pretty much calling it a day... and think it may be enough for your needs too.

I imagine BF Pro FS would run just like that.

As I do the monitoring, mixing and analysis in DAW, I don't miss TotalMix that much... what I do are firmware updates.
Hence, what I could suggest to RME and their new interfaces is to allow firmware updates via the DURec ports in stand-alone operation. Or somehow allow Linux-only users to update their firmwares.

87 (edited by ramses 2023-06-03 06:17:20)

Re: Official Linux drivers/tools a possibility?

> I imagine BF Pro FS would run just like that.

Has newer / faster converters
Babyface:            43 / 28
Babyface Pro FS:   5 /   7
samples latency in/out @44.1 kHz (values taken from manual, chapter "Latency and Monitoring").

> what I could suggest to RME and their new interfaces is to allow firmware updates
> via the DURec ports in stand-alone operation.

I think you mean USB port, as not every interface has DURec

But usually a driver is required as a prerequisite for flashing, there will be reasons for that requirement.

BR Ramses - UFX III, 12Mic, XTC, ADI-2 Pro FS R BE, RayDAT, X10SRi-F, E5-1680v4, Win10Pro22H2, Cub13

Re: Official Linux drivers/tools a possibility?

Only that few samples?! Oh... *tempted*

For example SoundDevices MixPre II can have its firmware updated by uploading a file to its SD-card, making sure there is enough battery left or plug into mains, and then just launch the update. Takes about a minute, including device reboot, and done.

Had RME implemented similar firmware update mechanism on their interfaces, we Linux-only people would live that much happier life, instead of shamefully seeking some alternative OSes once upon a time to update the otherwise totally awesome interface(s).

But sure, official support of at least some distro / kernel-line will bring the thing to pure perfection itself.

89 (edited by ramses 2023-06-03 06:21:06)

Re: Official Linux drivers/tools a possibility?

cupakm wrote:

Only that few samples?! Oh... *tempted*

I added the information that the values are taken from the manuals.
In my Excel you can also find this information and compare it between products.

https://forum.rme-audio.de/viewtopic.php?id=35156

There I took only the values @44.1 kHz, which is kind of "worst case" for converter latencies, as the latency decreases with increasing sample rate.

BR Ramses - UFX III, 12Mic, XTC, ADI-2 Pro FS R BE, RayDAT, X10SRi-F, E5-1680v4, Win10Pro22H2, Cub13

Re: Official Linux drivers/tools a possibility?

Thanks for the table, interesting.

It wasn't clear to me how you came to those numbers, but probably, as you say from the manual:
1) "Safety Buffer: An additional small Safety Buffer on the playback side has proven to be very efficient and useful. The Babyface uses a fixed additional buffer of 32 samples, which is added to the current
buffer size." and
2) "Under OS X, every audio interface has to use a so called Safety Offset on record and playback, otherwise Core Audio won't operate click-free. The Babyface uses a Safety Offset of 16 samples." hence 32 + 16 = 48

Well okay, but 1) they say "on the playback side", and 2) doesn't that apply to Mac OS only?

I still don't know where you got the output samples number from.

Re: Official Linux drivers/tools a possibility?

There is a table ...

BR Ramses - UFX III, 12Mic, XTC, ADI-2 Pro FS R BE, RayDAT, X10SRi-F, E5-1680v4, Win10Pro22H2, Cub13

Re: Official Linux drivers/tools a possibility?

Yes, I've seen your Excel sheet and PDF compilation too https://www.tonstudio-forum.de/attachme … 03-25-pdf/
...but I don't see an explanation there how you get values for the "Converter latency in samples @44.1kHz AD/DA" row, and your latest post doesn't comment on my previous two points "on the playback side" and "to Mac OS only".

93 (edited by Muffin 2023-06-12 18:48:18)

Re: Official Linux drivers/tools a possibility?

cupakm wrote:

Yes, I've seen your Excel sheet and PDF compilation too https://www.tonstudio-forum.de/attachme … 03-25-pdf/
...but I don't see an explanation there how you get values for the "Converter latency in samples @44.1kHz AD/DA" row, and your latest post doesn't comment on my previous two points "on the playback side" and "to Mac OS only".

At least for the UCX II the latencies given are internal to the device and is what is written in Section 40.2 in the manual, so this is independent of what the device is connected to. His spreadsheet numbers matches the manual.

https://www.rme-audio.de/downloads/fface_ucx2_e.pdf

"The Fireface UCX II uses AD and DA converters with an innovative digital filter, causing a delay
of only a few samples. With 5 samples AD and 6 samples DA the delay caused by the conversion
is only a fraction of previous generations. "

94 (edited by ramses 2023-06-12 20:28:19)

Re: Official Linux drivers/tools a possibility?

See screenshot from the manual for ADI-8 QS … values for 44.1 kHz marked in red.

https://www.tonstudio-forum.de/attachment/3028-rme-rtl-comparison-v2-jpg/

Taken from my latest blog article:
https://www.tonstudio-forum.de/blog/ent … cts-en-de/

BR Ramses - UFX III, 12Mic, XTC, ADI-2 Pro FS R BE, RayDAT, X10SRi-F, E5-1680v4, Win10Pro22H2, Cub13

95 (edited by cupakm 2023-06-28 23:15:53)

Re: Official Linux drivers/tools a possibility?

Thanks for the clarification. Fair enough.

I notice you use Sonnet USB3-PRO-4PM-E with Fresco Logic chipset for the measurements in your previous post. Have you measured with Renesas uPD720201 or uPD720202? Are there any comparisons?

I still somewhat wonder how can I achieve 3,8-something ms latency with BF1 at 32 samples / 48 khz with uPD720201 and a PREEMPT kernel. Looks to good to be true, but I tried measuring it many times in JACK+ALSA and ALSA-only too, and it still is sub 4 ms... the prev-gen converters.

Re: Official Linux drivers/tools a possibility?

Latency has nothing to do with different USB3 chip types.

BR Ramses - UFX III, 12Mic, XTC, ADI-2 Pro FS R BE, RayDAT, X10SRi-F, E5-1680v4, Win10Pro22H2, Cub13

Re: Official Linux drivers/tools a possibility?

How would you explain different latencies when using different USB cards / controllers, when all the rest remains same? (even plugging them into the same PCIe slot)

98 (edited by ramses 2023-06-28 16:40:34)

Re: Official Linux drivers/tools a possibility?

cupakm wrote:

How would you explain different latencies when using different USB cards / controllers, when all the rest remains same? (even plugging them into the same PCIe slot)

Just notice you talk about Linux, I can't tell, maybe the drivers are different there, you are not using RME drivers, you are using the recording interface in USB class compliant mode with a "non-RME" linux driver. Maybe there you have performance differences.

At least under Windows the latencies between USB and PCIe/TB are so small (really not significant in any way) that I think you can exclude the chipset vendor. Additionally the ASIO driver bypasses the Windows sound system fully.

More important is here, that the chipset is well designed and works according to standards.
Especially critical this was in the past with USB3 chipsets of ASMedia.
But there it was more a matter of whether they run stable at all.

BR Ramses - UFX III, 12Mic, XTC, ADI-2 Pro FS R BE, RayDAT, X10SRi-F, E5-1680v4, Win10Pro22H2, Cub13

Re: Official Linux drivers/tools a possibility?

Would flatpak be the way to make this work for most of the linux distributions?

https://www.flatpak.org/

100

Re: Official Linux drivers/tools a possibility?

Shipping Linux drivers for professional sound devices is a solved problem. Ubuntu LTS and RHEL releases are maintained for 10 years and that's what Digigram and Audioscience use for driver releases. Less of a moving target than any version of Win/Mac in the history of ever.

On the video side you have Blackmagic, Magewell, and AJA shipping Linux drivers using the same formula.

RME cards would (probably) work great in production but good luck convincing work to run 24/7 services on Windows.

I bought a RME AIO Pro last year for personal use, unfortunately the gentlemen developing the driver passed away https://github.com/PhilippeBekaert/snd-hdspe/issues/13.