audio glitches random noise revisited

[Steve Moore]

A while back I complained of 1 second audio glitches that were appearing in my track files at random. These glitches were NOT at the beginning or end of a song (an anomoly covered elsewhere on the DUC) but spread at random throughout. Sometimes the glitches were numerous, sometimes only 1 every 4 or five minutes... Occasionaly, none at all.

While defragging seemed to help, it did not solve the problem. I also double checked several clock issues which, in spite of a 42.5 reading instead of a 44.1 at one point, was not the cause. No matter which device was master and which were slaves, the problem persisted.

I reinstalled/updated drivers for the motherboard, etc. No change. Cabling was also verified, to no end. This problem was occuring on a machine that had run fine for almost a year and a half, with virtually no crashes or other problems.

At the point of the glitch, the waveform would appear as a solid clip. This would NOT show up until zoomed in significantly, almost to sample level. What's more, the clip never appeared or was heard during record, only during playback.

So I took the clipped portion of the waveform, separated the region and saved it to a disk file. Upon opening the file in a hex editor, the raw data showed that the file contained all 00's at the point of the glitch. The exact length was varied. It seemed that portions of the sample data being written to disk were being lost. (BTW, the glitches never appeared in more than one track at a time. Even on a stereo track, if there was a glitch it would appear on one channel or the other, but not both at the same time.)

Bottom line: a bad cache on the audio drive. Replacing the drive solved the problem. (I can't beleive it took me that long to go the trouble of replacing the drive!)

[Steve Moore]

!@*&!!****$%^&@!

argh!

After two days the same problem appeared! What the fuh?

I'm down to three variables1) bad power supply lead (going to swap power connectors with somwthing else), (2) failing motherboard (wouldn't his affect other drives as well?) and (3) the 001 itself.

Anyone know a way I can check the 001 for this type of marginal error?

[alleecat]

Bump! What gives? No one has else had this problem? No responses or opinions??!! Alleecat

[alleecat]

Bump 2!!!

[Chris Cavell]

Sorry Alleecat, I've never had this problem on my PC LE system. I've had something similar occur on a much older mac TDM system with scsi drives. I found that it occurred almost exclusively after defragmenting the drives. No need anymore for that. So much storage via firewire that I never even touch the scsi's anymore or have to defrag. Even upgraded computers since then and really haven't seen the problem in years. All i really can remember is that it seemed to be linked to defragging in my case.

[Steve Moore]

For the record, the problem turned out to be a faulty Promise controller chip on the motherboard. Bypassing the Promise controller (integrated - not a card) and simply using the secondary IDE channel solved the problem. My CD is now on the Promise chip, however, since it is read only I have not seen any problems there, yet. Nevertheless, I am going to try and disable the onboard promise controller and install a promise card.