1 (edited by wildthing 2017-01-11 03:00:34)

Topic: 2016 MacBook Pro and Babyface Pro - non-audio related issues

Hi,

I have a new (late 2016) MacBook Pro 15" running macOS 10.12.2 and a Babyface Pro running driver v3.01 and firmware v101.

Everything is fine from an audio perspective, but I've experienced several non-audio related issues described below. Has anyone else experienced these issues?


ISSUE 1: Disconnecting Babyface Pro generates a loud "Thwip!" noise

When I disconnect the BfP from my MBP, the BfP generates a loud "Thwip!" sound through the speakers - sounding vaguely like a laser gun in Star Wars. It sounds somewhat uncomfortable, like it could perhaps damage equipment, speakers, or ears (if wearing headphones). However it doesn't seem to have damaged anything so far. Pressing DIM or decreasing volume on the BfP makes no difference.

This issue occurs in all three of the following situations:

  • when disconnecting the BfP's USB cable (using any USB-C to USB-A adaptor, including Apple's own)

  • when the MBP shuts down, at the end of the shutdown process (when USB bus power stops being delivered to the BfP)

  • when the BfP is connected via a 3rd party USB hub that charges the MBP (i.e. supports "power delivery"), and mains power is switched OFF, causing the MBP to switch from mains power to battery power

All three scenarios generate a "Thwip!" sound.

I can avoid the sound by switching off my amplifier/speakers before doing any of the above. But I'd like to know: Is this a known issue? Can it damage anything? Is it specific to the new Apple MacBook Pro or UBS-C, or a general known issue with the Babyface Pro?


ISSUE 2: Babyface Pro sometimes not recognized by macOS when disconnected and reconnected, causing macOS to hang during shutdown

If I disconnect the BfP's USB cable and then reconnect it while the MBP is running, then the MBP often fails to recognize the BfP (no audio device in System Preferences > Sound, and Fireface USB Settings app does not start).

This happens perhaps half of the time when I reconnect it. I think it's more likely to happen if an audio app is open or has been playing recently, but I can't be sure.

A reboot resolves this issue. HOWEVER when I reboot the MBP in this state, then the MBP always hangs during the shutdown process (black screen with a frozen pointer), forcing me to do a hard reboot by holding the power button.

This happens on a factory-fresh MBP with a very minimal amount of software installed, using Apple's own USB-C to USB-A adaptor, and with no other peripherals connected.

The MBP only hangs in this way if I disconnect the BfP, reconnect it, find it's not recognized, and then reboot. If I reboot the MBP without ever having connected the BfP, it shuts down normally. If I reboot the MBP while the BfP is connected, it shuts down normally. But if I disconnect the BfP by pulling out the USB cable, reconnect it, and find it's not recognized as an audio device - then at that point I can be 100% sure that if I shut down the MBP, the MBP will hang during the shutdown process.


ISSUE 3: Babyface Pro not recognized by macOS on boot up, if already powered by USB hub supporting host power delivery

If my MBP is already shut down and the BfP is connected via a USB adaptor supporting power delivery to host, then if I switch on mains power to the USB hub first, the BfP's lights come on straight away. Then if I boot up the MBP, macOS always fails to recognize the BfP (no audio device in System Preferences > Sound, and Fireface USB Settings app does not start). In this situation, disconnecting and reconnecting the BfP always resolves the issue.

However, if I boot up the MBP first, so that the BfP's lights come on using MBP battery power, then macOS always recognizes the BfP. Following this, if I switch on mains power to the charging hub, the MBP switches from battery power to mains power and the BfP always continues working without any issues.

The hub is made by AUKEY, and the manual states that mains power is only used for charging the host, not the connected peripherals. However it appears that the hub also uses mains power to provide USB bus power to connected peripherals when mains power is on, even when the host is shut down. And if the BfP is already so-powered when the MBP boots up, then macOS fails to recognize it.


Current workarounds

Due to ISSUE 2, it appears there is no "safe" way to disconnect a bus-powered BfP from a running MBP. Therefore, whenever I need to disconnect or reconnect the BfP, I first need to shut down the MBP.

Furthermore, I avoid putting my MBP to sleep overnight with the BfP connected - because then (I assume) the BfP would drain the battery overnight (it certainly stays "on", and has no on/off switch).

So I have to do the following in strict order:

  1. Switch off the speakers (so I don't hear a "Thwip!" noise at the end of shutdown)

  2. Shutdown the MBP (so the BfP isn't draining the battery overnight)

Then, the next next morning I have to do the following in strict order:

  1. Boot up the MBP on battery power only (so that macOS will recognize the BfP when it boots up)

  2. Switch on mains power to the charging USB hub (so that the MBP switches to mains power, and I can close the lid and use external displays)

The requirement to shutdown the MBP when not in use is a bit of a pain, because I have to plan ahead regarding how long I'm likely to not use the MBP, and make a decision whether to sleep (and risk draining the battery) or shutdown.

I can live with these workarounds for now. But I am surprised by the apparent flakiness of the overall setup (MBP + USB-C + BfP) in this regard.

I have no idea if the issues are down to Apple, RME, or USB-C. I never had any of these issues with my old Babyface and 2012 iMac.

2

Re: 2016 MacBook Pro and Babyface Pro - non-audio related issues

Issue 1: normal. But it seems you ignored the level meters and the current fader position. Your monitors are far too sensitive and you need to reduce their sensitivity so that the quietest fart doesn't make your woofers bump.

Issue 2  and 3 have been reported already in this forum. Looks like Apple still has to iron out some basic compatibility with their latest computers.

https://www.forum.rme-audio.de/viewtopic.php?id=24614

https://www.forum.rme-audio.de/viewtopic.php?id=24602

Regards
Matthias Carstens
RME

3 (edited by wildthing 2017-01-12 00:02:55)

Re: 2016 MacBook Pro and Babyface Pro - non-audio related issues

Hi MC,

MC wrote:

Issue 1: normal. But it seems you ignored the level meters and the current fader position.

Changing the main fader level on the Babyface doesn't any difference. The Main fader in TotalMix is assigned to AN 1/2, hence is controlled by both the Babyface rotary dial and macOS system volume (they all change in unison), and the Babyface Pro AN 1/2 outputs are connected directly to my amplifier XLR inputs.

MC wrote:

Your monitors are far too sensitive and you need to reduce their sensitivity so that the quietest fart doesn't make your woofers bump.

My monitors/amplifier do not have any level controls. These are Amphion amp/monitors so we're not talking about some cheap monitoring system. The Amp100 has precisely one control: an on/off switch. The One15 speakers have precisely zero controls. Presumably Amphion expect the level to be controlled by a monitor controller. But the Babyface Pro is my monitor controller. I thought that's why it sits on your desktop and has a rotary dial, no?

As for audio listening levels, it seems fine - the relationship between Babyface level control and audio listening volume seems absolutely fine. But the volume of the "Thwip!" power-down noise is completely fixed, and there is no level I can change which has any affect on its volume - except to switch OFF the amplifier, so that is what I do.

To be clear, this isn't your typical on/off electrical impulse sound. It doesn't make the "woofers bump". It's an extremely high-pitched noise that rapidly sweeps through the frequency spectrum from high to low, and sounds like it's mainly coming from the tweeters. Have you tried reproducing the issue by disconnecting from a Mac USB-C port? Perhaps, we are not be talking about the same thing.

MC wrote:

Issue 2  and 3 have been reported already in this forum.

https://www.forum.rme-audio.de/viewtopic.php?id=24614

This report sounds exactly like my Issue 3 - they are saying, if the RME device is already powered up when the MBP boots up, then it is not recognized as an audio device. But if they disconnect/reconnect, it is recognized. The only difference is they have a Fireface UC which is presumably mains powered and has an on/off switch.

However it does not describe my Issue 2 - that is, after disconnecting/reconnecting the BfP on a running MBP, it's (intermittently) not recognized by macOS on reconnect, requiring a reboot - and furthermore, once in this state, the MBP always hangs during the shutdown process (black screen with a frozen pointer), forcing me to do a hard reboot by holding the power button.

I accept the issue is on Apple's side, and the underlying cause could be the same or closely related. But I don't think you can say definitively it is the same issue - not without a root cause analysis.

Details matter with these kind of things - an issue that requires you to reboot your computer is much more serious than one that requires you to disconnect/reconnect your audio interface. And an issue that makes your computer hang during the shutdown process is much more serious than one that requires you to reboot your computer.

These are different symptoms - and they may have the same or related underlying cause.

He's talking about how it works with some cables, and not others. Surely that can't be a software issue?

MC wrote:

Looks like Apple still has to iron out some basic compatibility with their latest computers.

Has RME made any attempts to inform Apple about this? Or is that considered to be a waste of time? Apple can't fix what they don't know about.

4

Re: 2016 MacBook Pro and Babyface Pro - non-audio related issues

Issue 1: I know that sound and with proper setup it is low volume and no problem at all (even at higher level it is no problem, but annoying, of course).

The BF PRO has XLR outputs with +19 dBu @ 0 dBFS. That is considered professional studio level and clearly stated in manual, website, brochures etc. Amphion fails to give the input sensitivity of their amps, but it looks as if they reach full level far below. So you need a passive voltage divider/attenuator between the BF Pro and the Amp100 input of (most probably) 20 dB. That will also allow you to mix more near a fader position of 0 dB instead of -20 to -40 (where you most probably are now).

Regards
Matthias Carstens
RME

5 (edited by wildthing 2017-01-12 03:03:58)

Re: 2016 MacBook Pro and Babyface Pro - non-audio related issues

OK, i'm currently listening at around -40 dBFS, sometimes even lower.

Do I need to get a passive voltage divider/attenuator, or can I continue working round the issue monitoring at low level and simply switching off my amp before I power down the BfP? Normal audio sounds fine as it is.

I guess there's a danger that if for whatever reason the macOS system volume goes to 0 dBFS then it would be, let's say, painful...

Where can I get a passive voltage divider/attenuator that has no effect on the audio quality?

6

Re: 2016 MacBook Pro and Babyface Pro - non-audio related issues

wildthing wrote:

Do I need to get a passive voltage divider/attenuator, or can I continue working round the issue monitoring at low level and simply switching off my amp before I power down the BfP? Normal audio sounds fine as it is.

I guess there's a danger that if for whatever reason the macOS system volume goes to 0 dBFS then it would be, let's say, painful...

It can kill your speakers, so you definitely need it.

wildthing wrote:

Where can I get a passive voltage divider/attenuator that has no effect on the audio quality?.

The web is full of solutions. Soldering 2 resistors per channel into a cable or buying a simple passive monitor controller like the Nano Patch+ will do the job.

Regards
Matthias Carstens
RME

Re: 2016 MacBook Pro and Babyface Pro - non-audio related issues

OK thanks.

8 (edited by wildthing 2017-01-17 01:10:05)

Re: 2016 MacBook Pro and Babyface Pro - non-audio related issues

I still have concerns about power management.

As you know the 2016 MacBook Pros have 4 USB-C sockets. I am using these as follows:

  • 2 have DisplayPort adaptors for external displays

  • 1 has an "everything else" USB port for keyboard, mouse, printer, MIDI keyboard, etc

  • 1 has both Babyface Pro and is used to charge the laptop.

I thought this arrangement makes sense because then USB audio is not shared with any other USB devices; it's only shared with power delivery (to charge the laptop) so there is no data contention.

I tried this using an AUKEY USB-C / USB 3.0 hub that supports "power delivery" to both charge the MBP and connect the BfP, and it seems to work - as long as I don't switch off mains power to the USB hub while the MBP and BfP are both running, as that can trigger "ISSUE 2" described above.

My concern is that when I put the MBP to sleep, then currently I have to do one of the following:

  • Leave BfP connected, and leave USB charging hub mains power on. That means the BfP continues running off the USB bus using mains power, but it also means the charging hub keeps charging the MBP battery to 100%, which I assume stresses the battery with constant small charges (is this a concern?)

  • Leave the BfP connected, and switch USB charging hub mains power off. But then the BfP draws USB bus power from the MBP while it's sleeping, which drains the battery.

  • Disconnect the BfP. But this triggers ISSUE 2 described above, as I often can't reconnect the BfP without rebooting, and then the MBP hangs during shutdown.

None of these options are ideal.

Therefore I'm thinking, maybe I need to get the RME mains power adaptor to go with the BfP? Then I will be able to put the MBP to sleep, leave BfP mains adaptor switched on, and switch mains power to the charging USB hub off.

I assume this means that while the MBP is sleeping, the BfP will not drain the MBP battery (as it has its own power source), and the USB charging hub will not constantly recharge the MBP battery (as it's switched off).

Furthermore, it should in theory avoid triggering ISSUE 2 - because the MBP never sees the BfP being "disconnected", because it stays on its own power source throughout.

Does this sound reasonable?

9

Re: 2016 MacBook Pro and Babyface Pro - non-audio related issues

There is no problem in constantly re-charging the laptop, Apple has taken care of that. A power supply for the BF Pro can be got in many places, 12V/2A is nothing special, maybe you even have one lying around so you could try it. Indeed current consumption from USB is zero when an external power supply is connected to the BF Pro.

Regards
Matthias Carstens
RME

10 (edited by wildthing 2017-05-14 00:56:22)

Re: 2016 MacBook Pro and Babyface Pro - non-audio related issues

It's been a while, thought I'd give a quick update.

Issue 1 - I resolved this using a pair of balanced in-line attenuators:
  http://www.ekmpowershop16.com/ekmps/sho … rs-3-p.asp

I got the -20 dB version, which still leaves plenty of headroom. And the "Thwip!" sound which the Babyface Pro makes when powered down is now of course 20 dB quieter :-)

The attenuators are expensive (£69 for a pair) - but still cheaper than a passive monitor controller, and I want to be sure they are sonically transparent.

Issue 2 - this issue is still here unfortunately using the latest macOS Sierra 10.12.4. However, today I was only able to reproduce the issue if I disconnected and reconnected the Babyface Pro while audio was playing on the MacBook Pro (e.g. in iTunes). If audio was not playing, I could not reproduce it - but I only tried about 8 times. Before I was sometimes able to reproduce the issue by disconnecting and reconnecting even without audio playing, but it usually took several attempts.

Anyway, I've started an Apple Support Communities thread here about this issue:
  https://discussions.apple.com/thread/7953968

Issue 3. I am now unable to reproduce this - so it may have been fixed. Either that or it's very intermittent.

11 (edited by wildthing 2017-05-14 01:00:10)

Re: 2016 MacBook Pro and Babyface Pro - non-audio related issues

I should add that Issue 2 is a bit of a pain, because there are several real-life scenarios where the Babyface Pro needs to be disconnected from a MacBook Pro:

1. When taking the MacBook Pro away from my desk. Since it's not safe to disconnect the Babyface Pro from a running Mac using USB-C, I have to first shut down the Mac before disconnecting.

2. When switching off mains charging via a USB-C / USB 3.0 hub that is providing both "power delivery" to the Mac and audio for the Babyface Pro. I can also reproduce the issue using Apple's own USB-C to USB adaptor - but due to lack of ports, I normally share the Babyface Pro with a power delivery hub.

I asked Plugable why USB devices drop when a USB-C / USB 3.0 hub is switched from charging to not charging - and they said this is inherent in the current revision of USB-C, adding "this should be fixed in later revisions of USB-C and associated products".

For this reason, using an external power supply with the Babyface Pro doesn't really help. If I put the MacBook Pro to sleep and switch off charging at the mains, the USB hub momentarily drops the Babyface Pro, regardless whether the Babyface Pro has its own power supply or not. So now I just leave mains charging switched on while the mac is asleep, and let the Mac take care of battery management. I only switch off the charging USB hub when the Mac is fully shut down.

3. Finally, USB-C plugs seem physically quite loose and can disconnect easily.

All the above can trigger Issue 2.

12

Re: 2016 MacBook Pro and Babyface Pro - non-audio related issues

wildthing wrote:

Issue 1 - I resolved this using a pair of balanced in-line attenuators:
  http://www.ekmpowershop16.com/ekmps/sho … rs-3-p.asp

I got the -20 dB version, which still leaves plenty of headroom. And the "Thwip!" sound which the Babyface Pro makes when powered down is now of course 20 dB quieter :-)

The attenuators are expensive (£69 for a pair) - but still cheaper than a passive monitor controller, and I want to be sure they are sonically transparent.

A bit overpriced, IMHO. It seems I wrote this useful info after you purchased them:

https://www.forum.rme-audio.de/viewtopic.php?id=25399

Regards
Matthias Carstens
RME

13 (edited by wildthing 2017-12-12 00:11:01)

Re: 2016 MacBook Pro and Babyface Pro - non-audio related issues

Another quick update on this topic. From my original post, ISSUE 2 is the only one outstanding; the other two issues are resolved.

RME have made it clear that this issue is on Apple's side. So I've been waiting and hoping that Apple will eventually fix this issue.

Well, today Apple released macOS 10.13.2 - and the release notes say "Improves compatibility with certain third-party USB audio devices". So, I tried installing the update and testing this issue to see if it's fixed. 

Unfortunately it is not resolved.

The behaviour is exactly as before. I can reproduce the issue 100% reliably as follows:

  • Disconnect all USB devices from the MacBook Pro

  • Connect only the Babyface Pro using Apple's own USB-C to USB-A adaptor

  • Play some audio e.g. in iTunes

  • Disconnect the Babyface Pro while the audio is playing

  • Reconnect the Babyface Pro

Expected behaviour: the Babyface Pro re-appears as a working audio device in System Preferences > Sound > Output.

Actual behaviour: macOS fails to recognize the Babyface Pro (no audio device in System Preferences > Sound, and Fireface USB Settings app does not start). The issue can only be resolved by rebooting - HOWEVER when I reboot the MBP in this state, the MBP ALWAYS hangs during the shutdown process (black screen with a frozen pointer). This forces me to do a hard reboot by holding the power button.

The same issue also happens (albeit less reliably) if the device is disconnected/reconnected when audio is not playing - that means the issue can occur quite easily during normal usage, for example when taking my laptop away from my desk to go travelling.

So it's back to waiting I guess...

Re: 2016 MacBook Pro and Babyface Pro - non-audio related issues

I have the same issue wildthing, but I’m still on 10.12.6. I won’t bother updating to 10.13.2 if the issue is still happening.

Re: 2016 MacBook Pro and Babyface Pro - non-audio related issues

I was reading this post and just had to add - as I think its crazy that anyone would power of a sound device without an attenuator/fader or turning of their speakers first.  For example in a  pro studio you wouldn't power down a desk BEFORE turniong the speaker of as you are likely to blow them.  This is no different - even if the Babyface pro had a power of switch you would still have to power the speakers of first.

16 (edited by rbbrnck 2017-12-15 23:01:49)

Re: 2016 MacBook Pro and Babyface Pro - non-audio related issues

i would not pull cables off from live audio devices of any sort.
or from audio devices of any sort, just simply switched on and not even live.
never. ever.

it's like pulling the key off of the ignition on a car with automatic trans while you're driving.
you just don't do it.

but, if all this sounds weird to you, and you just feel like doing it nonetheless:
then why knocking on the maker's door, asking what's wrong with your car?

Re: 2016 MacBook Pro and Babyface Pro - non-audio related issues

a couple of weeks ago, I picked up a Babyface Pro to use with my 2016 Macbook Pro, and immediately experienced the Issue 2 as reported above. I'm a long time Metric Halo user, and they had a similar issue with a bug in OS X, and suggested a workaround that I'd been using problem-free for a few years until recently (when they wrote a workaround into their driver code). after experiencing Issue 2 on the Babyface, I re-applied this workaround. I've been in heavy sessions with the Babyface Pro all week long in a variety of settings and have not experienced Issue 2 again. I cannot guarantee the problem is gone, but I've probably plugged/unplugged the interface 20-30 times over the course of the past week since applying the workaround and the problem has never shown itself again. (knocking on wood) below, i'll reprint the exact workaround from Metric Halo. BJ from MH has always insisted that this fix will not harm your computer and you can easily revert back if you don't feel comfortable.

best of luck,
jake rodriguez
***************************
Hi Folks,

I know we have posted about this here before, but I think until the problem happens on your system, the issue does not register, and when it does, you may not make the connection to the post that was made months ago.

The issue with playback not starting (usually followed by various apps hanging) is a result of a bug in OS X that affects OS 10.8 and higher; it has been acknowledged by Apple, and they have decided not to fix it.  Unfortunately, it is not something we can work around in the driver, but there is a work-around that you can apply to your system that will prevent the bug from occuring.

You can do the following to work around the problem:

If you are on a Mac running 10.10 or higher, please follow these steps.

reboot and hold cmd+R until the progress bar appears

when you are in recovery mode (macOS Utilities window), choose Terminal from the Utilites Menu

in the terminal, enter the following exactly (followed by the return key):

                 nvram boot-args='dart=0'

Select Restart from the Apple Menu

After that the problem should no longer occur.

To verify that the fix has been applied, after the machine has rebooted, you can open the Terminal command and run the following command (again, copy/paste):

nvram -p | grep boot-args

This should show the output:

boot-args    dart=0

if the fix was applied correctly.

If you reset your PRAM, or update your OS, you will have to run the first command again, as those two things clear the NVRAM that store the boot argument.

Best regards,

B.J. Buchalter
Metric Halo
http://www.mhlabs.com

18 (edited by wildthing 2018-02-07 02:01:23)

Re: 2016 MacBook Pro and Babyface Pro - non-audio related issues

strangedays wrote:

I was reading this post and just had to add - as I think its crazy that anyone would power of a sound device without an attenuator/fader or turning of their speakers first.  For example in a  pro studio you wouldn't power down a desk BEFORE turniong the speaker of as you are likely to blow them.  This is no different - even if the Babyface pro had a power of switch you would still have to power the speakers of first.

Yes but this is a bus-powered USB audio device, and there are a number of actions which have an indirect side-effect of powering up/down bus-powered audio devices, such as rebooting your computer. If you can always remember to switch off your amplifier/active speakers when you reboot your computer that's great, but I don't think it's intuitive nor exactly analogous to the pro studio scenario you describe.

Anyway I've resolved this specific issue by using passive attenuator between the Babyface Pro and the amplifier.

19 (edited by wildthing 2018-02-07 02:02:20)

Re: 2016 MacBook Pro and Babyface Pro - non-audio related issues

rbbrnck wrote:

i would not pull cables off from live audio devices of any sort.
or from audio devices of any sort, just simply switched on and not even live.
never. ever.

it's like pulling the key off of the ignition on a car with automatic trans while you're driving.
you just don't do it.

but, if all this sounds weird to you, and you just feel like doing it nonetheless:
then why knocking on the maker's door, asking what's wrong with your car?

First, this issue occurs when disconnecting USB devices even while no audio is coming out. I only mentioned disconnecting while audio is coming out because it's is easier to reproduce the issue that way - and it's always encouraged to provide vendors with steps that allow them to reliably reproduce an issue.

If you're saying devices should not be disconnected even when NO audio is coming out - well, that would mean that if you're using a laptop with a bus-powered audio device, you need to fully shut down the laptop every time you take it away from your desk.

But before you do that, you need to remember to power down your speakers. Because according to you it's not safe to disconnect your audio device until your laptop is shutdown, and according to the previous poster it's not safe to shutdown your laptop until your speakers are powered down.

So you have to always remember to do things in exactly the right order: 1) power down speakers 2) shutdown laptop, 3) disconnect audio device, and 4) remove laptop.

This is all fine. But at the end of the day this is not how these things are designed to work. It just isn't. Laptops are designed to be portable. Bus-powered audio devices are designed to be plug-and-play. Operating systems are designed NOT to hang!

If an operating system hangs when you disconnect your bus-powered audio device, something is wrong. It is perfectly reasonable to identify this as an issue (and in this case it happens to be an Apple issue).

20 (edited by wildthing 2018-02-07 02:14:37)

Re: 2016 MacBook Pro and Babyface Pro - non-audio related issues

sounjaerk wrote:

a couple of weeks ago, I picked up a Babyface Pro to use with my 2016 Macbook Pro, and immediately experienced the Issue 2 as reported above. I'm a long time Metric Halo user, and they had a similar issue with a bug in OS X, and suggested a workaround that I'd been using problem-free for a few years until recently (when they wrote a workaround into their driver code). after experiencing Issue 2 on the Babyface, I re-applied this workaround. I've been in heavy sessions with the Babyface Pro all week long in a variety of settings and have not experienced Issue 2 again. I cannot guarantee the problem is gone, but I've probably plugged/unplugged the interface 20-30 times over the course of the past week since applying the workaround and the problem has never shown itself again. (knocking on wood) below, i'll reprint the exact workaround from Metric Halo. BJ from MH has always insisted that this fix will not harm your computer and you can easily revert back if you don't feel comfortable.

best of luck,
jake rodriguez

Thanks Jake, I appreciate the help - but personally when it comes to running commands like this I prefer to understand what it does, why it might help, and what the side-effects might be, before I try it.

Re: 2016 MacBook Pro and Babyface Pro - non-audio related issues

understood. i will post another of BJ's early posts (sometime around 2014) when he described in more detail. i have used this fix on my laptops since around that time without any known issues, but i get wanting to be wary. hope this is helpful in your decision. fwiw, i have still not experienced the Issue#2 since applying this fix, and have hot-plugged and unplugged dozens of times.

cheers,
jake
*************************
from MIO-list circa 2014:

Hi Gary,

This is actually a bug in Mac OS X and has been reported to Apple and acknowledged, but not fixed (it appeared in Mac OS 10.8.2).

The Apple bug number is rdar://13093900

I have discussed this further with Apple engineering and while there is no way for Metric Halo to fix this or patch around it, as it is a bug in Mac OS X, there is a user-level work-around that you can apply that will fix the problem for you while we wait for Apple to fix the root-cause.

The problem is occurring because of a change Apple made to the IO layer in Mac OS X to protect against physical access attacks via Firewire and PCIe (Thunderbolt). This is called VT-d (it is an intel technology built into the Processors and Chip Sets) and it virtualizes hardware access; in particular, it denys Memory Mapped peripherals access to all of the HW memory map except for buffers that have had specific access rights granted by the OS. Apple's FireWire driver is supposed to grant access to the FireWire hardware (Apple's -- not MH's) via their VT-d driver. The bug you are running into appears to be a hang or infinite loop in the VT-d driver while trying to allocate the access window.

So, the work around it to turn off VT-d. This obviates the security that VT-d provides, but no Mac had VT-d before 10.8.2, and your problem is related to a bug in Apple's VT-d driver, so you are not really losing anything you ever had.

Best regards,

B.J. Buchalter
Metric Halo

Re: 2016 MacBook Pro and Babyface Pro - non-audio related issues

Thanks jake for the additional info - that's very interesting.

I'd assumed this issue was specific to USB-C because I never experienced it with my 2012 iMac with USB 3.0, only with my 2016 MacBook Pro with USB-C.

However BJ from MH describes an issue affecting "Firewire and PCIe (Thunderbolt)". He doesn't mention USB at all, although the new USB-C ports of course carry Thunderbolt 3 as well.

Just guessing - but maybe this VT-d technology started being applied to USB at the same time Apple moved to USB-C, and that's why we're seeing the issue now?

BJ mentions that the bug appeared in macOS 10.8.2 - that was a Mountain Lion update released back in September 2012, so if it's the same bug it's certainly been around a long time!

It's good to know there might be a workaround. However, setting firmware NVRAM variables sounds a little scary. It might be fine - but on the other hand, what was safe to do in 2014 might not be safe to do in 2018. Apple are constantly making changes to secure their bootchain. Even in 2014 BJ said this change will be reverted "if you reset your PRAM, or update your OS".

On top of that, you have to switch off a security feature (albeit one that only protects against physical access attacks).

I think it's a "try at your own risk" kind of workaround wink

23

Re: 2016 MacBook Pro and Babyface Pro - non-audio related issues

This 'workaround' fixes the USB2 and Touchbar crash on our MacBook Pro 15" that happens when disconnecting a USB 2 interface while audio is running.

We had send a bug report to Apple that also found VT-d involved (several months ago), but were not aware of any method to switch it off. Cudos to Mr. Buchalter and sounjaerk for publishing this useful info!

Regards
Matthias Carstens
RME

Re: 2016 MacBook Pro and Babyface Pro - non-audio related issues

happy to help. it really makes me wonder why Apple isn't actively trying to fix an issue that has been reported by at least two of the best and brightest audio driver authors in the business? why is it not a priority for them? bums me out...

25 (edited by Plsmk 2018-05-03 23:11:44)

Re: 2016 MacBook Pro and Babyface Pro - non-audio related issues

MC wrote:

This 'workaround' fixes the USB2 and Touchbar crash on our MacBook Pro 15" that happens when disconnecting a USB 2 interface while audio is running.

We had send a bug report to Apple that also found VT-d involved (several months ago), but were not aware of any method to switch it off. Cudos to Mr. Buchalter and sounjaerk for publishing this useful info!

I have tried this workaround but it did NOT solve the problem.

My setup: MBP 2016 with High Sierra + Fireface UFX or Digiface USB

Latest drivers installed.

This needs more attention and testing across different hardware configurations from you guys @ RME. I understand it's coming from hardware changes at Apple, but the problem is NOT fixed and should not be treated as such.

Re: 2016 MacBook Pro and Babyface Pro - non-audio related issues

It's Apple that's not testing properly.

I'm on Mavericks. My RME FF400 has been working properly for years. After a recent security update, it suddenly wouldn't approve RME's kernel extension. OSX told me to check system prefs and enable the extension there. No extension showed...

I opened the console. Lots of SMCkey errors. There are no docs for these, AFAIK. So I started trouble shooting, keeping a clean install as last resort. After backing up, resetting pram and SMC it still wasn't working.

I rebooted from the recovery partition and launched Disk utility. DU reported and fixed dozens of ACL permissions on folders in the system. This seems to have fixed the kernel extension problem too.

It's not an RME problem, as OSX succeeded in sandboxing all of Core audio, most of my VST plugins and several unrelated other parts of the system.

When it started working again, I updated from a recent driver to the latest without any hiccups.

What caused the ACL's?

Your guess is as good as mine. The Apple engineer I spoke to, had no clue and told me Mavericks was unsupported and I should upgrade to High Sierra to fix the problem.

Yeah, right. I now have even less confidence in High Sierra, and Apple too, for that matter...

MB Pro - 2 X FireFace 400, FF800 & DigiFace USB
ADAT gear: Korg, Behri, Fostex, Alesis...

Re: 2016 MacBook Pro and Babyface Pro - non-audio related issues

MC wrote:

This 'workaround' fixes the USB2 and Touchbar crash on our MacBook Pro 15" that happens when disconnecting a USB 2 interface while audio is running.

We had send a bug report to Apple that also found VT-d involved (several months ago), but were not aware of any method to switch it off. Cudos to Mr. Buchalter and sounjaerk for publishing this useful info!


Hey,
today I tried again to find again workarounds for these problems I have with my Fireface uc and my MacBook Pro 15 inch 2016.

My biggest issue still is the need of a hard reset (or even sometimes then reinstalling the OS) after I hot deplugged the interface or I switched off the interface while used (for example in Logic or MainStage).

I tried the Workaround by sounjeark and like for plsmk I didn't work. I still have the same Problem. Is there anything I can do?

Best Regards from Germany

Michael

28

Re: 2016 MacBook Pro and Babyface Pro - non-audio related issues

The above workaround does no longer work in Apple's latest macOS. Very unfortunate. That said we are still on the case and do our best to get a solution.

Regards
Matthias Carstens
RME

Re: 2016 MacBook Pro and Babyface Pro - non-audio related issues

Hello, found this thread helpful. Is the bug resolved yet?

I have a MacBook Pro 2017 on high sierra  USB-c only.  Will the Babyface pro function with apple’s USB-b adapter?

Re: 2016 MacBook Pro and Babyface Pro - non-audio related issues

Good evening, I am new here, hope this is the right thread to ask this question. My search here did not find an answer. Is there a way to connect wireless to the ADI 2 Pro from a MacBook? Right now I am using a USB cable but from the couch that's annoying as the cable is a trap to stumble. If I use a USB-B male to USB-A female adapter and a WiFi dongle in the USB-A, will this work? Is there a better way for below 100 EUR?
Thanks in advance, Chris

Re: 2016 MacBook Pro and Babyface Pro - non-audio related issues

c@hk wrote:

Good evening, I am new here, hope this is the right thread to ask this question. My search here did not find an answer. Is there a way to connect wireless to the ADI 2 Pro from a MacBook? Right now I am using a USB cable but from the couch that's annoying as the cable is a trap to stumble. If I use a USB-B male to USB-A female adapter and a WiFi dongle in the USB-A, will this work? Is there a better way for below 100 EUR?
Thanks in advance, Chris

Won't work, as you need a USB connection from the computer to the recording interface so that it can be accessed by its USB drivers. Besides that, the recording interface would have no drivers to handle USB based Wi-Fi devices.

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

Re: 2016 MacBook Pro and Babyface Pro - non-audio related issues

I'm probably late to the party here, but after recently upgrading all my Macs to Apple silicon this issue finally appears to be fixed.

Running the latest DriverKit driver v4.09 with macOS 14 Sonoma on a MacBook Pro (M3 Pro) and Mac mini (M2), I can now disconnect and reconnect RME audio devices from their USB-C ports without then needing to reboot for macOS to recognize the audio device again. And when I do need to reboot (for unrelated reasons), it happens smoothly without hanging during shutdown.

Everything seems stable and working 100% normally. I'm also impressed that both my Babyface Pro (bought in 2016) and even my original Babyface (bought in 2011) are both supported using the latest drivers on my much newer Macs! This makes me happy.