Did you test the Macbook plus interface only ? Without any other equipment hub bluetooth etc. ?

If you want stability go Windows. Seriously.

@musicandy76:

Some Apple devices have issues with USB 3. The reasons are different; some are based on the particular Apple model, and some seem to be memory related. 8 GB DRAM seems too little. Also, the OS version can be an issue as well as kernel driver vs. driver kit. Normally the kernel driver should give more stability, as kernel mode runs with higher preferences compared to user mode processes.

But there are also plenty of customers who seem to have no problems...

Whatever the reasons are in particular....

According to what you can read on the RME forum, using Thunderbolt, you should be able to create the most stable setup.

You have the UFX+ so you can use Thunderbolt.

I would try this as the first thing before trying a lot of things with USB ...
One additional remark: with M1 processors with USB 3, customers had many problems.
The solution there was to get certain Thunderbolt docks and to use the USB 3 ports there on the dock.

But please try Thunderbolt first.

I would first try DriverKit drivers because this is Apple's newest driver standard.
If this should still cause issues, then try the kernel mode driver, which gives audio-related processes a much higher precedence compared to user mode processes.

I am a Windows user, but having read this forum for a long time, this is what I would try in your situation.

PLease also advice if you think this is an Apple M4 specific issue, that I would need to contact Apple about.

You can't compare USB2 and USB3 1:1.
USB2 has fewer channels and uses another USB transfer mode for transferring audio.

Nobody can give you a guarantee that Thunderbolt resolves your issues but it is very very likely!

If you would read the forum regularly, then you would know that there were different kind of issues around USB, but never with Thunderbolt. Even if the cable is expensive I would invest in an ORIGINAL Apple thunderbolt cable.

Thunderbolt is "external PCIe" .. its really worth a try.

Thanks its defo worth a consideration given the amount of people who have reported USB issues on Mac.

I am not on Mac, but as far as I know, if there is a known problem, the problem is on side of Apple and only Apple can solve it....

RME has done as much as it could do. They even made third driver that utilises CC mode of Apple and so bypasses "features" of OS that Apple is not eager to fix....

Isnt this a Thunderbolt 3 to 2 adapter ?  https://www.argos.co.uk/product/7428441 … 16EALw_wcB

Yes, it is, it is exactly the one you need and it is made by Apple.

And sorry I did mess up a little, bc I misread you, I thought you wrote TB 3/4 adapter. But yes This is the adapter you need BUT it works only one way. So you need also Apple TB 2 cable.

This is bc this adapter will not work with new TB 3/4/5 cables with USB-C type connectors on both ends. It will only work with old TB2 cable, to connect UFX+( display port type socket) to USB-C type TB 3/4/5 socket in Mac.

I hope I cleared it out.

And if the fault is on Apple is there a way of raising a ticket with them?  My trusty Maschine Mk3 has no such issues when I use that as an audio interface and thats on USB.

SO what do you actually advise?  Is there an issue with the Driver Kit drivers on USB with M4 Pro Macbooks?  Is this a known issue that RME are trying to fix?

I would actually like some decent advise please.

I have turned off Lossless audio in my Apple music settings and so far its not reporting packet losses (i,e in the diagnostics 0/0/0/0/0 list, it was displaying an increasing number in the 2nd and 3rd digits)  Amazingly 3 tracks have played with no glitches, the first time today.

Try secondhand. I got all of those secondhand. Got meself two long (2m) and two short (25cm), bc they come only in those two lenghts, and got four adapters. Didn't cost me much.

Use Group Policy Editor for delaying patches and functional updates to get more stability in.
Especially useful for Windows 11, where Microsoft introduces many changes and not everything seems to be well tested.

More information here in this posting:

https://forum.rme-audio.de/viewtopic.php?id=43086

This would also block important security updates.
I wouldn't recommend it because this impacts the security of your system.

I think the majority of users have the DAW system in use for other tasks as well and are permanently connected to the internet.

I rarely see any postings here from people indicating they have such a setup, where the computer is not connected to the internet.

Even Major DAWs offer features to be able to upload the mixed or mastered stuff directly to the cloud, some even for collaboration.

And even if you do not connect systems to the internet, security updates and updates of Microsoft Defender are not unimportant, because you can also get issues from installers or when connecting USB sticks. Or the problem might even come through the LAN.

I think it's also not easy to decide whether it is a good moment to update or not. It requires you to check several tech forums on the internet, but this is only a snapshot of what is known. Some information needs some time until something cooks up.

Therefore, I think it is a much better approach to delay updates by 2 or 3 weeks, because usually Microsoft is relatively quick to have the biggest issues resolved in one or two weeks.
And even if there should be an issue, back out a patch or restore the system backup.

In the past 10+ years I never ever had to back out a Microsoft patch.
What I do is to regularly restore backups, because I often try new software. And very often you do not get a full installer, but a

Also a trend is, that companies like Steinberg offer a Download Assistant, which saves you time downloading and update your application software. Some other companies offers you a download tool, which automatically fetches the latest version from Internet.

It's getting harder to get full installers for applications; internet connection is more and more important.

Only big studios might go the extra mile, but I think this approach is not the standard anymore. Too many enthusiasts and project studios are nowadays there, for whom this would be a financial burden and waste of time having to maintain several systems manually this way.

Just to be on the safe side. I certainly didn't mean to criticize you.
You simply wrote something about my thoughts, that's perfectly fine and I did the same.

Everyone is free to contribute and then take away something suitable for themselves from what has been said.
It's all good, Vinark. Thank you for your contribution; they are always welcome from my side.

@musicandy76, this has been discussed endlessly on the forum. Basically Apple and RME have both dug their heels - Apple with no desire to move from their security-related choice to kick audio drivers out of the kernel layer, RME with their strategy of offering 3 different imperfect paths to solving the problem: legacy kernel extension, driverkit client driver, and class-compliant mode. To this some users add their own "legacy firewire" (donglelife) solution.

The legacy kext driver (USB) requires you to lower security settings and to put up with frequent nags by the OS about obsolete drivers that will stop working in a next major update. In return, when it works, you get decent stability, and the shortest possible buffers, lowest possible latency.

The DriverKit client driver lives in "user space" and thus hasn't got enough CPU scheduling priority, so is frequently choked by fast data going to another part of the system (for example quick-viewing a sequence of RAW photos from a TB-connected SSD while playing a song in Apple Music). It's great when it works but buffer underruns are most frequently reported/associated with this method, all over the forum.

Class-compliant mode may work for you, depending on how you use your interface. It uses Apple's own driver (which conveniently for them resides in the kernel), and the minimum size buffer is a few samples longer (I think additional 32), so the latency is always a bit more, I think similarly to Firewire's buffer size. I think TotalMix supports CC-mode these days. It previously didn't.

Thunderbolt 3 → Thunderbolt 2 → Firewire → RME seems to work for the first few days, unless you never restart. I tried it, and ultimately reverted back to USB. I sleep my Mac, and never turn off my interface. After about a week of using Firewire it started glitching, until one day it completely crashed, and I had to unplug all cables and quit my browser (no idea), before things were good again.

Lastly, RME are not the only vendor having these issues on Apple devices. I think Thunderbolt is a better choice for stability and I wish RME would consider simply shipping TB interfaces exclusively. Alas..

Hope you find this useful. Absolutely do experiment with USB cables - seems to be another weak link in the system.

Sorry, but it's not really RME offering 'imperfect' paths; it’s Apple forcing them and other pro audio vendors into a corner. The problem is that Apple has kicked third-party drivers out of the kernel while keeping their own native audio driver there.

In the kernel, a driver essentially 'owns' its CPU core and can’t be interrupted by the OS scheduler until it’s done. By moving everyone else to User Space (DriverKit), Apple has made professional drivers subject to system interruptions—hence the buffer underruns when you're doing something as simple as viewing photos or other tasks.

RME is just trying to make the best of this situation by providing workarounds for a design choice by Apple that prioritizes general system security over the specific, uninterrupted CPU access that professional audio requires.

EDIT:

What I don’t understand is why Apple customers and the trade press aren’t picking up on this issue. A few years ago, Apple had serious problems with its USB drivers. At first, they completely ignored the matter, telling customers they weren’t responsible for issues with third-party USB devices—or something along those lines.

It wasn't until the trade press covered the situation with enough visibility that Apple finally took notice and, after nine months, fixed the problems.

I wonder where the 'outcry' is now? It should be obvious that, given the underlying mechanisms (kernel vs. user mode), DriverKit for Audio cannot offer the same degree of prioritization for real-time audio processing. This is simply not a viable solution for professionals (and other impacted customers) who require stability and reliability for their daily audio work and their premium hardware, which costs a lot of money.

Currently, it seems to me that Windows—with its direct access to hardware via ASIO—is actually the superior platform for low-latency audio. It’s rather strange that Apple is giving up its edge in the creative industry so easily, simply by prioritizing general security over professional performance.

Luckily, some of the mitigations offered by RME do work for some people, but this is more a workaround than an appropriate or satisfactory solution

Understood, got your point; let's simply treat it fairly.

BTW, I added something to my posting; see above.

Had one weird thing when playing a backgammon match online that requested use of my microphone, that seemed to crash the  the system.  And oh the irony!! The bloody thing has just disconnected whilst playing from Apple music.  Time to get shut of the UFX+

Had enough