Mbox CoreAudio and 6.4: fixed?

[Arvid TomaykoPeters]

Hello

I am playing in a crazy electronic music concert in a couple weeks, and I will be using my Alum powerbook 15" 1.25GHz (OS 10.3.3) and probably my Mbox as an interface to MaxMSP. It would be preferable to have nothing crash during my performance! So I would like to know what people's experience with the new CoreAudio driver in Panther is, since the one for v.6.2.x crashed most software I tried to use with it. I have upgraded to 6.4 and I am experimenting. I would like to hear what everyone else has to say - what apps work, what apps crash? On what systems? Cool. Thanks.

Here are my results:

Reason Adapted for Digidesign (standalone [NOT rewired, using coreaudio]): Crash every 30 sec. Boooo!!! (this is the same as what happened before in panther and 6.2.3, except the following) Runs seems to run fine the first time it is launched after Digi Coreaudio prefs are trashed, but that claims "not enough memory" when quit and launched again (I have never seen this before). Once this gave me a kernel panic! - something that, to Apple's credit, I have not seen since upgrading to Panther 2 1/2 months ago.

iTunes: Good. I played about 10 min of music, both from my HD and over the LAN. no crashes.

Quicktime Player: seems OK, played 3 min music.

MaxMSP (the most important for me, followed by Reason): No problems - ran for about 10 min. Was ok when quit and relaunched. But I'm not sure I trust it yet for a performance because it used to crash with PT 6.2.3 and OS 10.3.2/10.3.3.

Spark ME (I know it is no longer supported, but it still works fine for me using built-in audio or Jack): CRASH! after about 20 sec use:

Command: Spark ME
Path: /Applications/SPARK ME 2.8/Spark ME.app/Contents/MacOS/Spark ME
Version: Spark ME 2.81.269 (2.81.269)
PID: 507
Thread: 0

Exception: EXC_BAD_ACCESS (0x0001)
Codes: KERN_INVALID_ADDRESS (0x0001) at 0x1b830010

etc. etc. etc.

Jack Tools: Yes. It still works - thank goodness, as this was the only way I could run many apps through the Mbox with 6.2.3, and still may be the only way for some! Using this I can run Reason and Spark through the Mbox successfully (and at the same time!), but it is a bit of a pain.

Plogue Bidule: Yes, after several minutes of softsynth playing. Works as it always did. I think apps that use the Portaudio library pretty much always work with the Mbox coreaudio drivers.

And the other day the Mbox was making short horrendously loud buzzing noises when the classic environment was on (only because Makemusic or whatever they are called this week has not yet upgraded printmusic to OS X. Grrr.)

---
I am disappointed that I cannot count on this Mbox to work. It frustrates me that I pay $$ for something and I can't be reasonably (no pun intended) confident that it will work when necessary. Anyway, I would appreciate any feedback. Why does some software work and other software not work? Thanks.

[O.G. Killa]

Just because Digidesign releases new drivers doesn't mean old programs will work. Reason crashes on everyone's computer in stand alone mode when using Digi CoreAudio drivers...but yet iTunes works fine...and almost every other program works fine...so why do you think it's a Digi problem?

Maybe you should jump over to the forums for the programs that actually crash, and tell them to update their programs to work with the latest core audio drivers! When was the last time Reason was even updated??? I tell you it was well before panther was ever released! How do you know the crashing isn't because Reason has a bug when running in Panther?

Just because a third party program doesn't work on Digi's core audio driver doesn't mean it's a problem with the driver. Sometimes, yes it totally is a problem with the driver...sometimes its a problem with the software that is trying to interface with the driver.

And...this all leading up to your question, "Why does some software work and other software not work?" Because it's most likely a problem with the software, not the driver!

[Arvid TomaykoPeters]

ok, ok - I understand that. And as far as SPark goes, I can see that being a software problem. But my point is that doesn't Reason work with OTHER coreaudio interfaces, like those from MOTU, M-audio and Edirol? Or doesn't it? Things like Reason and MaxMSP are software that people depend on everyday - aren't they? Maybe I am overestimating the amount that other hardware/software works as compared to DIGI stuff. It just seems everything in so touchy with the Mbox and there should be a better way...

I DON'T know whether the crashing is from a bug in Reason when in Panther. I would like to though. But, for instance MaxMSP used to consistently crash using the Mbox just a few days ago in the same manner as Reason, but now it seems fine (without a MaxMSP update), which leads me to believe that Digi has done something to the driver, so I'm trying to get a sense of what implications that has - that's all. Later.

[Muxlow]

why do you think it's a Digi problem?

Maybe because Digi's been screwing up left and right ever since they made the jump OS X? Maybe because the first 3 or 4 versions of CoreAudio were practically useless? Maybe because key elements of CoreAudio STILL don't work as they should?

I'm not trying to start a pissing match, but in this case the poster's skepticism if fully justified by Digi's recent track record. You may be right and it could very well be a Reason bug, but when you have two companies, and one company puts out a rock solid application that almost never fails, and the other company puts out an incomplete, buggy application that breaks key elements of your system (deep sleep), and crashes regularly and repeatably, who do you think is going to get blamed when a problem arises?

If Digi wants to stop being blamed, Digi needs to get their act together and prove themselves again. The poster's skepticism is a reflection of the frustration Digi's put us all through over the last 12 months.

[antnyzman]

Arvid,
I think the whole idea of Digidesign Core Audio is for audio output through "Digidesign" interfaces. As for me, I have no problems with audio output from Reason 2.5 and DP 4.12 through my Digi 001 using Digidesign Core Audio. In addition, have you repaired permissions and rebuilt your system drive's directory using DiskWarrior? I've seen the repair of permissions followed by DiskWarrior fix a lot of problems. Finally, did you try removing the (questionable) selected apps under Supported Applications in Digidesign Core Audio and then adding them again?
Hope this helps. Probably won't but thought I'd give it a shot.

[Slim Shady]

My Mbox runs great with Logic 6.4.1 - seems more stable than the last version, but I upgraded Logic at the same time, so it's hard to say. Rock solid performance, love it.

[Arvid TomaykoPeters]

Thanks for the suggestions.

I think it is specifically the Mbox that has these problems. The Digi 001 and 002s I have heard work fine with Coreaudio.

Anyway, I have repaired permissions a number of times and use Onyx optimization features regularly, but have never rebuilt this disk (my diskwarrior is for OS 9 and i don't have my old iMac with me for firewire target disk use). Maybe I will try that though over the summer. Yes, my list of "supported apps" is very clean, now that I have identified problem apps.

And about logic working - that's great. I am considering getting logic express soon for composing and for better integration with MaxMSP and Bidule. Cool.

[JFreak]

coreaudio is a moving target at the moment - apple changed coreaudio between 10.3.2 and 10.3.3 and according to rumour site reports they are going to change it to 10.3.4 again. so if the coreaudio driver works now it will likely not work in the next panther upgrade.

at this time i would not do anything that depends on coreaudio, but instead hope that 1) apple gets the coreaudio right in 10.3.4 and 2) digidesign is fast in releasing 6.4.1 soon after apple os update.

coreaudio and audiounits are good thing and i really hope digidesign will some day fully support them. audiounits in protools, please, DIGI are you listening?

[gugu]

I just updated to X.3.4 an digicoreaudio is just crashing Cubase SX2, Peak.....as usual.

May be a question to ask.....I'm travelling at the moment so I bought a PB G4 1,25 + mbox with LE 6.4. PTLE 6.4 just works fine. All other apps are crashing the digicoreaudio (crash logs are always the same).

But when I'm using the core audio driver (6.2.3) at home on my mix+ hardware (not officially suported) everything is running fine.........

Please Digi, is the mbox really reliable with coreaudio ?????????

gugu

[birdFEEDER]

at this time i would not do anything that depends on coreaudio, but instead hope that 1) apple gets the coreaudio right in 10.3.4

I think you might be a bit confused about CoreAudio. CoreAudio is VERY RELIABLE at this point and has been since 10.3. There are STANDARD ways of implementing a device driver to be compliant with Apple's CoreAudio. Apple even supplies a generic template sample code to start from. Digidesign is the rogue lone ranger in the audio-world electing to snub these standard practices to enforce their own restrictions (single client-ness, having to add an app to the "approved software list" with their utility, etc).

SO, the fact that Digidesign's driver breaks (or becomes unstable) everytime CoreAudio is updated is NOT because Apple is significantly changing the CoreAudio API, it's because of Digi's hairpin driver implementation.

I've had the same unupdated CoreAudio driver for my Delta 1010 since Panther's initial release. If M-Audio can do it, Digi should be able to do it.