Any 5.12 problems (Allegedly)

[Phil O]

I'm usually a hold-out. But for some reason I jumped in with both feet this time. I'm using 5.12 with 10.4.9 and having no problems.

Phil

[monkey man]

`
Glad to hear it, Phil. You bucked your own Phil Osophy; that takes guts.
Funny, it seems we've swapped roles; I usually jump in blindfolded.

I don't get it.
There seems to be no rhyme or reason to it all.

The thicket plottens.

[TheHopiWay]

FWIW after a 6 month hiatus from DP (working on a PT project) I zeroed the data on a drive partition, installed OS 10.4.9, DP 5.12, The latest versions of:
UAD and TC software (for 4 uad cards and one powercore), Waves Gold bundle (MAS)
AutoTune
ApTrigga

So far smooth as silk.

[michael77]

Does anybody out there have crash problems with Auto tune 4 in DP 5.12?

I have to do vocal projects in DP 5.01

[TheHopiWay]

Autotune 5 is recommended

[Rick Averill]

Just today, I started noticing audio bursts with VIs

Funny you should mention it. I also am getting audio bursts. They are sporadic and seem to come at random times in my sequences. I think you are correct about them having to do with event chasing; I recently turned it on, and it seems that maybe they didn't happen before then.

Event chasing is certainly convenient while I'm working on a sequence, but I suppose you're right that it could be turned off during the final stages of freezing tracks, since then you would be starting play at the beginning of the sequence anyway. But it's so annoying trying to get your mix right when the sound keeps popping out at you.

Can't tell if it's the VI's or what, but I think the sound increase is pretty much across the board, mix-wise. They are so unexpected and so brief, it's hard to tell. If anyone knows any more about this or discovers anything, I'd sure be beholden for any insight into this problem.

[Frodo]

Hey Rick:

As best as I can tell, the audio bursts could be some sort of glitch in the way virtual memory caches info. It seems to be double caching some blocks of samples, depending how, when and where the user stops and starts or resets.

So there are several things at play concurrently which may be contributing to this anomaly:

1. For samples to stream, a portion of the data is loaded and the rest if referenced.

2. Virtual Memory remains a virtual mystery with how it really works. Not even the VSL people could explain why OSX was able to funnel more data in and out of RAM, but it seems that extra file references are being cached on the fly. Lots of room for an 'oops' or two there.

3. DP's buffers *may be* loading copies of this same data for its own purposes as a redundancy, but the result is that DPs buffers and VM are overcompensating at times where "missing" data is called for, resulting in both batches of the same samples sounding at the same time (ie: bursts).

4. Add Event Chasing on top of all of this, and DP is suddenly looking ahead and look back to preload buffers with samples. I suspect that there is no duplicate sample reconciler where one copy of the same sample or groups of samples are used and the others are set aside in a sort of "smart file selection" process.

More to the point, the conflict may be one of definition: caching and referencing files don't always intelligently dump unneeded samples-- or there are times when they 'can't yet' dump unneeded samples before they get played back. Instead, they are fairly thoroughly cached and are loath to depart. New references appear to be written before old references can be dumped. If this is the case, then I think it might be something handled (or overlooked) at the host level.

Here's why I hesitate to install 5.12: it seems to me that MOTU, Apple, and third-party virtual instrument developers are all covering their butts to make sure that samples get loaded and don't drop out. That's a good thing-- at least in principle. Everyone is filling in the gaps, but the result of covering the bases is sample overlapping at times. For this reason, no single developer (MOTU, Apple, or plugin X) may feel that it's a deficiency on their part if inspected individually. Collectively, I am certain that reproducing the problem is part of the problem with MOTU, OSX, and third-part plugs working in combo.

Analogy-- three entities working together to do laundry. Each one makes allowances for adding plenty of soap unbeknownst to the other. As a result, the washer is loaded with three times the soap and the user has a mess on their hands.

I'm not convinced yet that the issue is any different in 5.12 than it is in 5.11.

These are only my theories and specs!!

But the audio leakage is quite real. One thing I haven't noticed is this occuring with outboard MIDI gear-- only with virtual instruments.

I've got event chasing off for the moment, but it will take at least a day of working with DP before I can determine if the bursts happen again. The odd thing is that I'd finished one project already without this happening at all-- with event chasing enabled. Same template, same setup, etc.

Still testing.

[Rick Averill]

But the audio leakage is quite real.

Thanks for your reply. Glad to know someone more knowlegable than me is on the case.

I've noticed audio leakage also. There are puddles on the floor under my DAW. I've been blaming my cat, but...

One other thing--does it happen more when you put the ring on? There could be evil forces at work here. Oh, I'm sorry--did I say ring? What ring? I didn't see any ring.

[waxman]

Please explain audio leakage?

[Phil Jeffers]

I have replicated the OMF import problem on my system:

Copied a 500MB OMF 2 file (from Avid Adrenaline) to local media drive.

Launched DP 5.12

Opened OMF file

The Project was opening for approx 20 secs (converted around 200MB of the audio files) then popped up an error window reading:

An error occurred in OMF import: OMFI_ERR:
Read past end of data (OMF error code = 122)

This DID NOT occur on my system with DP 5.11. I have not changed anything else on this system except for update DP to 5.12 from 5.11.

Cheers, Phil J

[monkey man]

Do these audio "bursts" sound at all phased?

The symptoms sound like the "MIDI volume" bug that's been there since November/5.11 was released.
This was the issue I (and many others) were emailed about by tech support.
It seemed they were working hard at nailing the final glitches in 5.11.
Unfortunately, they were experiencing difficulty replicating the problem.
This wasn't surprising, considering its "ramdom", seemingly unrepeatable-on-demand nature.

Could it be that this "new" glitch is simply 5.12's manifestation of this as-yet-unquashed bug?

[Frodo]

Please explain audio leakage?

I'm not sure you really want to know. In fact, I rather envy the notion that you've not experienced this. Consider yourself blessed!!

Monkey Man has asked again if this were not the same issue many reported as a MIDI burst-- a retriggering of the same sample via MIDI, which in turn causes a sudden explosion of MIDI notes.

I thought at first what I'm now experience was indeed MIDI related, but as I said, I think it's more closely related to the way audio samples are cached, parsed, and delivered to the audio card. It just appears that it's happening in duplicate at times. It's not quite the same as the MIDI bursts of yore.

In the case of MIDI only, there would be some obvious 1:1 cancellation or phasing, especially where the exact same velocities are being delivered simultaneously. The effect is akin to having double notes on a MIDI track

In the case of audio, it's trickier because while there is some phasing and some cancellation, there is also an aggregate increase in volume output. This is what I'm experiencing now, whereas before I'd get nothing playing then suddenly an entire measure of notes comes crashing through on a single beat concurrently. There's just enough timing displacement to mask some of the cancellation, but phasing itself is easier to recognize when it happens. The samples are not being triggered at the exact same time, though close enough together that it appears to be the case.

This is what I meant by audio leakage-- more audio is being delivered than it ought. Multiple references to audio files are being fed through to the outputs instead of being filtered out before the leakage resulting in the bursts can take place.

The only other thing I can come up with is that some ghost in the machine is sending a sudden bump to the master fader. But I'm not running any continuous data at the moment -- and volume data specifically.

Because nothing appears anywhere in DP when the bursts or leakages occur, it's impossible for me nail this down.

[monkey man]

Hey, Fwoggy!

Thanks for those thoughts.
The "volume increase and phasing" bursts I experienced appeared to emanate from within DP's mixing engine, as I no longer mix OTB, running the ROMplers directly into aux tracks.
This, coupled with the uncanny circumstantial similarities, is why I asked again.

Thank you again for allowing the monkey to distract you from your mission.
Nicky

[Frodo]

Hey, Fwoggy!

Thanks for those thoughts.
The "volume increase and phasing" bursts I experienced appeared to emanate from within DP's mixing engine, as I no longer mix OTB, running the ROMplers directly into aux tracks.
This, coupled with the uncanny circumstantial similarities, is why I asked again.

Thank you again for allowing the monkey to distract you from your mission.
Nicky

No distraction at all, Nicky. For all my previous verbage, I'm rather stumped, actually. Gladly, it's not become a crippling issue by any means. I'm currently just preparing a series of orchestral scores for porting into Finale, but projects slated for summer will be much more critical where audio mixing and presentation will be mission-critical. I'll have to start in on running QT movies as well.

DP's mixing engine, eh? Funny- I've not put in any mixing automations at all, so if this is indeed the case then the solution is fairly beyond reach. If master volume levels are doing an internal trampoline act not represented in the GUI, I'm not quite sure quite what to do-- and the bursts appear to be occurring (shire-side) with *all* instruments as opposed to individual ones. Is this similar to your experience?

If it is indeed DP's mixing engine I am actually considering installing Logic on my MacPro just for the sake of comparison. If it happens with Logic as well, then that means it's an AU/OSX issue.

A better comparison would be to use another non-Apple DAW, but this is more trouble than its worth, me-thinks.

[monkey man]

Hey, Fwoggy!
Thanks for those thoughts.
The "volume increase and phasing" bursts I experienced appeared to emanate from within DP's mixing engine, as I no longer mix OTB, running the ROMplers directly into aux tracks.
This, coupled with the uncanny circumstantial similarities, is why I asked again.
Thank you again for allowing the monkey to distract you from your mission.
Nicky

DP's mixing engine, eh? Funny- I've not put in any mixing automations at all, so if this is indeed the case then the solution is fairly beyond reach. If master volume levels are doing an internal trampoline act not represented in the GUI, I'm not quite sure quite what to do-- and the bursts appear to be occurring (shire-side) with *all* instruments as opposed to individual ones. Is this similar to your experience?

Yup, it does seem to be the full mix that rises in volume for a second or so.

If it is indeed DP's mixing engine I am actually considering installing Logic on my MacPro just for the sake of comparison. If it happens with Logic as well, then that means it's an AU/OSX issue.

A surf of the Logic boards (!) could yield less pain.
Besides, if it was an OSX issue, we'd surely have heard something from SOS and various other commentators.

A better comparison would be to use another non-Apple DAW, but this is more trouble than its worth, me-thinks.

Agreed.

Considering the fact that you really only require 1 "clean" pass per bounce or mixdown, and taking into account the frequency of this anomoly, I can't help but wonder whether you'll be OK.
There's also the possibility that MOTU's tearing their hair out over this, and will either post another incremental, predominantly bug-fixing update soon or jump ahead to 5.5 or whatever if they've deemed it somehow immune to or lacking the bug.

Look at it this way, if it rears its ugly head during a critical pass, the chances of its happening immediately thereafter are practically zero, meaning you'd then have a window in which to do the bounce, recording or mix.

Monkey-thoughts for the Fwogster.
Nicky