Weird DPC readings after DAW application crash, what could still be wrong, and the case of the PC that speaks with forked tongue ...............

I had cause to investigate DPC meausrements after an application crash, and afterwards the audio application would not run with the soundcard's ASIO driver without dropouts. I measured the DPC peaks as over 45,000 (!) ie 45ms. Hmmmm. (well actually that was not exactly I said, but anyway ....)

So I then went to the other partition on the same machine and measured again, and found all green readings at around 250us average, and whilst playing a soundfile using Directsound with WM player, the DPC readings were doubling, i.e. around 500us.

Now this second partition was in fact my original hard drive partition, and I always liked the sound I got off this using WMPlayer (with all the EQ/SRS etc switched off).

So then began the long process of trying to find the driver that had obviously been taken down by the application crash. Anyway, after reinstalling the chipset drivers, NIC card driver, and graphics card driver, I found using RATTV3 that the main culprit was atapi.sys - so I booted into the other (E:) partition and from there replaced the disk drivers in the (C:) system32 folder and in the dll cache. I now had the DPC readings down to around 300us with a regular spike appr every 5.5 seconds of 4000 - 5000us. That is except when there was any hard disk activity whatsoever, when the whole system went absolutely ballistic with a constant DPC reading level of around 10,000us.

Nothing I did made any difference, until I found one of my original installation CDs that came with the computer - which contained the manufacturer's system software. Aha !!

So I reinstalled this software, and hey presto, the DPC spikes vanished. The measurements were all in the green, at around 300us, but on this partition (C:) no change in readings when playing an audio file. So far so good. I then reinstalled the soundcard, and played an audio file with WM Player, just as a test.

Huuuuh ??? It sounded different to how it had before the crash!! All the drivers were reading fine, DPC was fine, everything OK in Device Manager, but it sounded different.

So just in case I was imagining things, I went back to the E: drive and listed to the same file again using WM Player, and no I was not imagining things. It did sound different. The drum track I was using as a reference was now much soggier (but only in the C: drive) - mpre bass, less top. So next I went back to the DAW software and EQ'd the file to how I remeber it sounding before the crash, and it needed about 6db at 1kHz and 12 db at 20khz rsing up from about 10kHz. Now that's a lot of EQ just to get to the point where a drum track sounds like it did before a DAW app crash.

And what is bothering me now is that I have no way of knowing which is the correct playback - that from the E: drive - punchy as anything - or that from the C: drive - lots more whooomph in the bass but quite soggy overall.

But what is even more concerning about this little episode is that I could be mixing away ITB, blissfully unaware of these problems, adding my bucketfuls of EQ to get the drum track to cook again, and then if it transpires that my C: drive system is indeed still messed up (even though it now measures well from a DPC perspective), then if I render files straight from the DAW software to a WAV or MP3, then the rendering will be done AFTER I have added the EQ !!!

So really, unless you get the sound out of the little beasts in the first place, and recaptuing it, you really do have no way of knowing whether you are getting what you think you are getting. Sorry to all the forum members that see this as completely obvious, but as far as I am concerned this i something I have just discovered - and I have a major problem in my system that is affecting the sound quality coming out of it.

It is tempting to offer the standard solution which is to reformat the C: drive (after backing up the data files !!), and then I'll know it's correct. Perhaps that is true. And it is also true that there are still some tell-tale signs of the C: drive system misbehaving, which is that opening and closing windows sometimes leads to audio droputs even though the DPC checker shows no variation whatsoever. Go figure !!

I have posted this to try and shed some light on what on earth could be the problem in the system (it seems to be a system issue as the file played first with the Directsound driver and the ASIO driver does not sound right), but also to raise some awareness on this type of problem, so maybe we can all share some intelligence on this type of issue, whcih is why I believe Martin posted his request for DPC readings initially.

Before this application crash around 1 week ago, everything was relatively ok and stable.

Any comments Martin ??!!

Post Extras