Update: Did various checks on the corrupted files today. DC offset revealed nothing, declipping brought some things of more interest: At the beginning and end of the files it showed some details of the recorded material, so it seems that the signal is just extremly boosted in order to get recorded fully clipped and distorted throughout almost all recorded data.

I then did various tests with the reinstalled driver 2.896:

Using Beta 4 on the first recorded session everything was fine.

On the second recorded session the extreme boost started amidst the recording, since this time there were very low signals present on the 20 analogue inputs, the boost was clearly visible and audible without having to declip in the first place. Time and HD display were red.

On the third recording everything was fine again, except that HD usage was red.

My first conclusion, it is not a driver failure, since 2.896 worked flawlessly before with Beta 3.
My second is, if it is DC Beta 4, what could bring up this behaviour of boosting the recorded signal?

In Global Record's settings I was using, apart from K-System 14db, the default configuration.

A pitty I ran out of time then, so I will have to go on testing tomorrw with Beta 3, this would be the exact same setup, which did flawlessly record before. If the failure won't occur again then I will have to seriously suspect DC 5 Beta 4 to be the at least involved, even if that's not a proof, it might be reason to look into it...

However I will also run  a few sessions with a standard 20 channel recording into single mono files through Sam Pro, to see what happens.

I'll report again tomorrow night.

Best Regards
BV

Hello Matthias,

first thing I did today was going through the list of tweaks and turning of the bootcamp manager, which did a lot as was visible in the DPC latency checker. Also I managed to stop XP System Restore which might have helped too. Write combining was off anyway, powermize shouldn't be present anyway.

Most importend I must admit that I was using a HFS+ volume through MacDrive for external storage what I had thought shouldn't have much of an impact on single disk writes, and that could have been my biggest mistake. Switching to another, NTFS 1Tb via the same route, saved me from getting a red alert.

I then recorded another 25 minutes with DC Beta 4, first it didn't playback properly, distorted sound, a bit like bad driver communication, so I restarted and voila, the saved 20 mono files played back just fine, no boosted up to distortion volume etc.

Hmm... I thought I'd give Samplitude a try, twenty tracks input recorded over 25 minutes, all fine.

This was quite an impressing change and confusing at the same time, I will do further recording tests tomorrow with DC and the changed setup, since one question sill remains: What of these parameters that I have changed would have caused DC to boost the recording level up to full ditortion? Is it possible at all, that a blocked disk write process does force this?

I will again report tomorrow if anything was still fine, doing some heavy recording jobs.

Thanks for your advice!

Best Regards
Boris

Today for some more on this topic:

The recent two recording sessions, using the same setup, writing to NTFS external etc, were still showing those short boosted level spikes - one up to two per recording, in almost every second recording done.

Those were the very same recordings where again the Time and HD Usage indicators went red during recording.

To get into this a little more deeply I saved these recordings as Multiple Channel File as well as Single Mono Files. Those recordings with the red indicators showed the boosted level spikes at the very same position. So it shurely happens during the recording process, i.e. writing the temporary file to disk, not during the saving process.

Another thing I was realizing while going deeper into the waveforms, there are right after those spikes, delayed for about half a second, one or two still noticeable spikes of lesser level, but audible (sounds like a hickup).

I then checked today our studio current while running the latency checker with nothing very obvious coming up. Nevertheless I will check all wiring and grounding in the studio tomorrow.

I already did another 20 input channel recordings last week, via Samplitude, writing to mono files which was shwing none of those spikes at no occasion.

This leaves me again a little clueless. Still I am thinking, that Bootcamp is not an optimum setup when trying two write those big interleaved files to disk. Could that be a reason? I can't tell, since I have no other than DC to record these big ones in one stream.

Still any hints on checking out more to get deeper into this topic are more then welcome.

Best Regards
Boris

Seems like I will crosscheck using the current beta with a different setup, i.e. internal MBP drive, external FW 800 and 400 drive, and last with a Vaio AR-series laptop. After that going on to the next beta ;-) and then someday I will find out and report.

Regards
Boris

Have already downloaded rc 5 :-)

I will do a test run with both harddisk configurations, NTFS and HFS+, during this afternoon.
Additionally I will do a comparision with the Vaio AR-S setup I have been using before I was setting up MPB Bootcamp, using again both external harddisk configurations - no internal, since the internal RAID with the Vaio is plainly unbeatable :-)

You can expect some report on these latest tomorrow. Thanks!

Best Regards
Boris Vogeler

It is not useful to start GR two times, asd the second instance will negatively affect the first GR.

Hard Drive problems are indicated by a red color in the HD Load field. A red color in the Time field shows missed ASIO buffers. This can be caused by system events like maximizing/minimizing windows with animated windows active. And others. I am pretty sure that you have some wrong information in the file which is not so obvious anymore as we fixed the interleave out-of-sync problem.