Are audio overload errors grounds to claim a refund?

[glygly]

I certainly think so, the software isn't working as advertised. Sure it turns on, but once you work with it doesn't function well enough to do what's advertised, i.e. MAKE MUSIC. ...Discuss.

 

And please, any "I've never had a problem with logic, ever ever ever" users, don't bother posting here. This is an honest question.

[inmazevo]

To answer your question:

Yes

 

However, for the most part this is extremely difficult to do here in the states. Apple isn't likely to grant a refund for a problem they haven't fixed in 3 years. It's a non-issue, historically, complete with a company line on the matter that more-or-less puts the burden of the problem on you, rather than the software itself.

 

Again, however, this is because it has been (up until recently) a fringe problem. There are a variety of reasons for this fringe-ness of it, but it appears to be changing recently.

The web is getting full of people, particularly recent converts who just bought a Mac and Logic because it "just works," and they're truly astounded. It's becoming rather common to see it 3 or 4 times a week, by different people around various forums... maybe more. And they're p**sed.

 

It cannot remain receiving the fringe-bug treatment now.

 

So, as much as I agree that people receiving this should be able to get a refund, I believe the most constructive thing to do all around (albeit annoying to many) is to keep the heat on.

 

I don't mean flood David's forum with continuous threads about it. That would be disrespectful and counterproductive, and everybody will just end up mad at each other, which is not what I want this forum to become.

I mean: file reports to the Logic Pro feedback site and Apple bug reporting site with detailed specifics on your machine, your project, and repro steps required to reliably produce the error... once per week until it gets fixed. Everyone should do this (no denial of service attacks, please; just respectful reminders through the channels they've provided for exactly this type of thing).

 

On a (slightly) more positive note regarding the actual issue itself:

In a strange twist of my last 3 months of trying figure this thing out, I stumbled on an accidental "workaround," that actually worked quite well on my system. In fact, purely by accident, I found a system configuration that made my Overloads on 100% of my projects go away 100% of the time (note that this does not mean I found something wrong with my system, or think that this isn't a bug).

 

I just went through it last night, so I'll have to compile my notes of everything I've tested so far. But, I think I know what the problem is, and now actaully have a workaround that eliminated the symptom of the problem 100% of the time (the Core Audio Overload, IMHO, is actually the symptom of an underlying problem with Logic). I also think I now know why there is so much apparent variability... you know: "I frequently get Overloads" vs "I virtually never get Overloads."

 

Once I've gotten all my testing notes and observations/theories together, I'll post a thread with all the info. It might actually help some people find a way to work with Logic now while waiting for the bug fix that I finally think will happen (soon... ??... maybe, but certainly eventually).

 

Take care,

- zevo

[el-bo]

i have these problems AND i make music......does that make me bi ????

 

 

 

 

seriously, though

 

 

if this is really stopping you in your tracks (no pun intended)

 

1) try posting your system specs

 

2) tell us the project sizes/types in which these problems mostly occur

 

 

this has been posted, re-posted and then again posted

 

we are all trying to find workable solutions.....although some folk just aint interested

 

if you really can't work, take it up with apple and maybe you will be successful...but to ask us to just "discuss"....sometimes it feels like this is the only thing we discuss

 

search the boards and narrow your problem down

 

as i said, we are all interested in solving this, but with the knowledge its existence, this problem didn't keep me from upgrading to l8 (less problems FOR ME, but not non-existant).......and even with this annoying problem the advantages of the software outweigh the disadvantages, and i am grateful (daily) for having such a powerful tool at my disposal

[el-bo]

right on, zevo

 

 

i would be really interested to read and experiment with your findings, once ready

 

 

good day to you both

[inmazevo]

right on, zevo

i would be really interested to read and experiment with your findings, once ready

 

good day to you both

 

I should be able to post everything in the next couple of days.

I've done extensive testing on it, and I THINK its fairly holistic for my systems. At least, I can reproduce everything reliably.

 

I have one more test case for my "workaround," maybe two. Then, I'll get all my notes from across the past 3 months, post them here (in an organized fashion, which isn't my strong suit) and file those reports to Apple.

 

My hope is that we can point at the smoking guns (yes, plural... I think there are two, and maybe a related third) and get a fix, as well as (hopefully) make future Apple hardware/software better for everybody.

 

It would be great if we can get a thread of testable systems (well-defined) with reproduce-able data, that I can reference when filing feedback/bug reports.

 

Have a good day as well. I'm off to bed,

- zevo

[Fred B]

My hope is that we can point at the smoking guns (yes, plural... I think there are two, and maybe a related third) and get a fix, as well as (hopefully) make future Apple hardware/software better for everybody.

 

Thanks for your efforts. I'm very curious about the results.

 

But unless proven otherwise I still think the overload problem is due to a basic design issue. The only cure would be that Logic analyzes the complete song and allocates all needed resources *before* starting playback/recording.

[inmazevo]

My hope is that we can point at the smoking guns (yes, plural... I think there are two, and maybe a related third) and get a fix, as well as (hopefully) make future Apple hardware/software better for everybody.

 

Thanks for your efforts. I'm very curious about the results.

 

But unless proven otherwise I still think the overload problem is due to a basic design issue. The only cure would be that Logic analyzes the complete song and allocates all needed resources *before* starting playback/recording.

 

That's pretty much my opinion as well.

It seems to be a basic problem with how and when (and where) Logic allocates system resources during playback. Core Audio Overload messages are merely a symptom of that issue.

 

Like you said, Logic either needs to either allocate resources upon song load, or they need to write a VERY good read-ahead buffer. I think the architecture and marrying of Logic Pro to Tiger/Leopard on Intel Mac is less than optimized.

 

Take care,

- zevo

[Musicianista]

I'm not sure I understand the problem.

 

How can there be users running Logic Pro 8 without any problems? Are you saying that there is a bug that affects certain Mac models and configurations? Therefore the minimum spec requirement as listed is wrong?

 

Just curious.

[edubz]

I'm not sure I understand the problem.

 

How can there be users running Logic Pro 8 without any problems? Are you saying that there is a bug that affects certain Mac models and configurations? Therefore the minimum spec requirement as listed is wrong?

 

Just curious.

 

keep in mind apple changed CPU chips 3 times already?

 

PPC motorola

PPC IBM

Intel

 

now, understand that apple will be developing for intel ONLY. PPC for now on is just being "supported" and not developed

 

There is no test to see if a universal binary really works as advertised. What I mean is, logic 8 MIGHT run better on Intel then on say a PPC.

 

There are to many factors out there

 

I found this read which does state logic 8 sucks, at least in its 8.0

 

http://knobsnfaders.com/2007/10/13/logic-studio--cant-be-pt-killer-with-system-overload.aspx

 

"I found that i cant even load 2 instrument tracks without latency and overload messages. But when i run Logic 7.0 or even Ableton live, i can run 10 times more tracks and plugins as in logic 8. No wrong setting can even remotely invalidate a software applications performance like that."

 

And don't get it twisted, this is not a "Macbook" issue. One user (anthro) notes :

 

"It's just quite frustrating having spend that much money on that MacPro 8-core Music-Workstation and suffering from that old Overload-Message again!"

 

Now, you have an 8-core machine receiving overload messages? That's just not cool. What part of the game is that? This further makes me, and others, believe that Logic is more than likely the culprit.

 

And we may be right, there could be something wrong with Logic 8's architecture. It's unacceptable to call yourself a professional application with system overload messages popping up every so often. The funny thing is that the messages will pop up even though (according to activity monitor) the cpu is not being whipped and beaten.

 

I love Logic, and as stated I am a devoted Logic user, but truth is truth. I'm tired of seeing system overload messages. We all are. I'm not buying a new machine (I've already done that, and will not do it again). A MBPro with 3gb ram is just fine. No other DAW is tossing monkey wrenches into my productivity like this one. I advise that everyone gives Apple FEEDBACK. I know I have.

 

There are other tiny little problems with Logic 8 but the System Overload is by far the most frustrating, and in my opinion should be a priority to resolve. I'm sure some of you Logic users out there are just like me, anxiously awaiting the next software update. I'm sorry Apple but if this isn't taken care of there is no way I can use L8 for professional applications anymore. I'll make the move to Cubase. Part of me believes this is some clever conspiracy to force me into purchasing a new Mac Pro

[inmazevo]

I'm not sure I understand the problem.

Are you saying that there is a bug that affects certain Mac models and configurations?

 

Yes. Well, sort of.

It's complicated.

 

This "issue" has always been here, since before it was an issue. Something that used to work doesn't work so well anymore.

 

Picture it this way:

Logic Pro and Core Audio load their assets into memory/virtual memory in a particular way.

When you hit a midi region that triggers a virtual instrument (for example), Logic Pro's initialization logic (in concert with OS X's Core Audio system) reacts by reading some things from memory and virtual memory (ie - disk), and writing some things to memory and virtual memory (ie - disk). Pretty normal stuff.

 

But, Logic 6, Logic Pro 6.4 and 7.* were primarily written and optimized over time for a completely different memory/bussing subsystem than is currently the Apple line of machines, with a completely different architecture for writing bytes up and down motherboard busses (ie - G4/G5).

 

This is where it gets a little bit complicated:

When you go through this initialization routine on the current Intel architecture, you are reading AND writing to your hard drive quickly, and in relatively larger chunks (in bytes) when compared to the G4/G5 method

 

Without some type of gate on the host end (which used to come in the form of smaller chunks of data being transferred across busses - 4k pages I think - as opposed to the Intel bursting method - 64k pages I think), you are literally giving and asking too much too fast. The connection to your hard drive, and the amount/size of the data you're running across this connection, exceeds the ability of most current hard drives to keep up, particularly laptop hard drives, but certainly not limited to them.

 

When you exceed this limitation:

Core Audio Overload (couldn't process the data in time) - this is the default, "I don't know quite what's going on but I can't process everything so I'm quitting" error

Disk Too Slow - this, IMO, is actually the truthful error, and the system also quits

Error Synchronizing Audio/MIDI - this is the "nearly failed" case... Logic keeps going, but it's pretty sure something bad happened, or nearly happened

 

(NOTE that I'm not saying that every instance of the above errors is related to this particular issue)

 

You can simulate a "gate" by running your system (OS and all) across a bus that (theoretically) is going to have some side-effect gating when compared to the SATA interfaces most of us are using.

 

What I did was to run an exact copy of my system from an external Firewire drive, daisy chained off another external Firewire drive, daisy chained off my audio interface. (NOTE that I don't know whether it's the FW gating it, or the daisy chaining, or both)

I booted, and ran through my repro steps for the projects that always throw the error (which is always accompanied by a CPU spike and disk IO spike at the moment of the error).

 

Guess what:

While the CPU still spiked a bit where I expected it to, and the disk I/O still spiked a bit where I expected it to, I did not receive the error... NO CORE AUDIO OVERLOADS. On larger projects, where there are multiple plugins initializing @ the same time, I did get some audio glitching, and one audio/midi sync error, but the projects played significantly better.

 

I didn't make the problem go away. I just circumvented one of the symptoms a bit.

 

To be perfectly honest, I'm getting a little out of my eliment here, unfortunately. I'm not a guru down here in these bowels.

 

So, I'm enlisting some workmates who know far more about such things than I do (particularly with Mac hardware), and we'll be continuing the investigation until we have exhausted our abilities to do so.

 

It SEEMS to make sense, so far, and certainly would explain why some get it and some don't... G4/G5 = less likely, and depending on your current model of Mac the bussing setup might vary a bit, and of course hard drive capabilities vary quite a bit, and then of course everyone's using different audio interfaces (it is, after all, likely to contain a bit of the Core Audio layer of things).

 

To make a long story longer:

It's not going to be easy to fix. I agree that its Logic, but it might not be that simple. I think it's Logic + Core Audio + Intel + bussing architecture...

 

In effect: Logic Pro, and possibly Core Audio, need to be optimized for the Intel architecture.

One possible short term fix would be to add a read-ahead buffer to Logic Pro so that it initializes ahead of time, or gates itself in some way. The problem with this is that it would be difficult to do this without touching the performance on systems where the issue doesn't exist. This is, however, probably how the applications that don't have this issue get around it (ie - Cubase/Ableton Live).

 

My honest opinion is that Core Audio, at the OS level, needs to find out if you're on Intel and adjust its bandwidth accordingly (ie - it gates for you). While you could do this at runtime, perhaps a better solution would be to profile your system once, and then run off that profile until you (or the system) updates it.

 

Either way, it's expensive.

 

And, before anyone rips my theory to bits:

This is the best I can do to explain what I mean while at work today... no notes, and not so much time.

But, if anyone has some testing suggestions, I'd be happy to check it out when time permits.

 

- zevo