Hey all!
I've been using Sonar X3e pretty smoothly for a while now, probably a little too smoothly because I've become a lot more lax about keeping a "pure" DAW computer...
Today I had a VO record session and twice (out of about 60 record start/stops) after hitting stop on the record, the audio file shrank a little in size on the timeline (about 5%) and was just a solid block of 0db white noise.  The actual file on disk is just noise as well, the recording lost.
This has happened once before a couple weeks ago, but I chalked it up to a freak glitch... now I'm getting worried!
So, I'm going to be doing some trouble shooting, but wondered if you guys have ever experienced this, or have any tips on common culprits/where to start looking?
 
Here's the specs/info that might be relevant - and thanks!
 
- New project created in Sonar X3e
- No plugins running (bypassed)
- Bad recorded clips about 30 seconds each
- RME Fireface 800, Texas Instrument FW card, FW400
- ASIO, 512
- Windows 7 64, Intel i7-3930K @3.20
- 32 GB RAM
- Recording to 1TB 7200 RPM SATA HD with 175GB free
- OS drive separate Solid State
- NVIDIA GTX 660 Ti Video Card
- Matrox MX02 Mini installed
- Canopus ADVC110 installed
- Adobe CC installed
- AVG installed
- USB Dongles plugged in (Steinberg, ilok)
- UAD Quad PCI installed
- Connected to internet via Ethernet cable
 
 

Were you monitoring the audio through Sonar during the recording (i.e. not direct through your interface)?

My guess would be that it's probably not a software problem, but more likely an issue with the microphone preamp, or conceivably the microphone itself.
 
If you're going straight into the RME, try switching to another channel for awhile and see if the problem reoccurs. If it does, it's possible there's an issue with either the microphone itself or the phantom power supply - assuming it's a condenser, of course. A dynamic mic won't exhibit that kind of problem.

This can happen for a variety of reasons, like a loss of clock sync or misset sample rates. The "character" of the noise is a clue - mic preamp hiss is "gentler"-sounding than the "spiky" noise from clock issues.

Thanks for the responses guys!
To answer your questions, I was monitoring through Sonar (echo on track) with no plugins activated, and everything sounded fine.  Sonar's meters were normal and the waveform looked normal as it was drawing.
I was wondering if the fault could be connected to the Fireface as it's getting on in years (8 now!), the puzzling bit for me was that Sonar was drawing the waveform fine as it was recording, it was only after I hit stop that the length of the clip shortened, and the waveform showed solid noise.
The character of the noise is super harsh, a static white noise at a solid 0db - very jarring!
 

Biscuits
The character of the noise is super harsh, a static white noise at a solid 0db - very jarring!

 
Hmmm...that sounds like a digital issue to me.

Anderton
 
Hmmm...that sounds like a digital issue to me.

Would it be possible for a clock/sync issue communicating with the Fireface to affect the clip after it's been recorded (just as you hit stop)?  I'm not sure exactly how Sonar records its files, whether the audio information is stored in RAM as it records and then committed to HD as a wav file only when the recording is complete?  It's something during this process that is causing the trouble it seems...

As I understand it, Sonar writes audio directly to the disk as it's recorded and streams direct from disk on playback, unless Windows disk caching has been switched on in preferences. Not so many years ago recording a few CD quality stereo tracks a few minutes long would have exceeded the available RAM and direct to disk streaming avoided that problem. It's also claimed to protect recorded audio from being lost even if the software or PC crashes.

As for the original issue, I've been scratching my head about this because I'm sure something very much like it happened to me a few years ago, maybe in Sonar 6 or early in the v.8 life cycle. Same kind of thing at any rate. Waveform drawn OK during recording then turning into a different length of what looked and sounded like digital clipping. I can't be certain but I think the cause was either sample rate mismatch or a clocking problem.

Thanks tlw!
Did you ever figure out the root of the cause?  I'll be sad to see the day when the the old FF800 dies, but I have my suspi cions...

I don't have a solution for you, but I wouldn't give up on the hardware yet...RME stuff is built like a tank, and has a reputation of working solidly with Sonar (that's been my experience too when reviewing RME gear). Try swapping out the Firewire cable and other obvious things first. Maybe borrow a Firewire card for your computer and connect to that if you're using the motherboard's Firewire port.
 
My usual troubleshooting procedure is to try the simplest and most obvious things first, even if it seems unlikely they're the cause of the problem. That won't take much time, and you won't kick yourself if you spend time on esoteric possibilities only to find out a cable wasn't in all the way or whatever.