24-bit feedback problem

[The_Raven]

Howyadoin,

Picked up the new Sound Forge and was checking out the 24-bit goodness, but I ran into a bad feedback problem. Levels were OK until a signal came into a mic, then the level went up and stayed at wherever it peaked. If another, stronger signal came in, it would boost the peak and again stay there... If I reset the record monitor there would be no effect, but if I turned the checkbox for monitor off then back on, the levels would return to normal, but would start freaking out again upon an input to the mic. Note that this does not happen in 16-bit sampling, only 24-bit... any ideas?

Thanks!


-Mark Pags
Sherbet Studios

 

[The_Raven]

Oh yeah... almost forgot

The problem only occurs at the 96KHz sample rate... 48K and 44.1K seem to work fine...


-Rav

 

[LooneyTunez]

Have you tried recording at 24/96? Is this feedback present in the wav file, or just as your monitoring?

I'll venture a guess that your computer might just be freaking out at 24/96, as it takes a lot of processor power and bandwidth to be able to record at that resolution.

 

[drstawl]

What OS are you running Sound Forge 6.0 on?

 

[The_Raven]

Howyadoin,

Here's my specs...

Hardware-

Case w/400W ATX Power supply

Motherboard:
MSI K7N420 PRO, nVidia nForce chipset, BIOS 2.0

CPU:
Athlon XP 1800+ (no OC, box CPU fan)

RAM:
(2) 512 MB DDR2700 (Slots 1&3, dual 64-bit memory controllers)

Drives:
Western Digital 60 GB UDMA 100 7200 RPM
Asus 32X IDE CD-RW
Vintech 52X CD-ROM

Audio:
On-board MCP (DD5.1 nVidia Crush)
Delta 1010

Video:
On-board MCP (GeForce2MX using 32MB shared DDR RAM)
PNY Verto nVidia TNT2/PCI 32MB (TwinView)

Software:
Win2KPro w/SP2
Cakewalk Pro Audio 9
Sound Forge 4.5
Sound Forge 6.0

All Sandra benchmarks are at or near best-of-breed...



-Mark Pags
Sherbet Studios

 

[drstawl]

This sounds like an issue of the settings on your Delta.

What are your settings in the Delta HW panel?

I was recording at 24/96 tonight via a S/PDIF in and the recording went perfectly up until ~ 55 seconds into it- then the sample rate went crazy and the recording was mangled.

The problem turned out to be not using the S/PDIF source as the master clock. Upon resetting this from the internal clock to the S/PDIF source clock the recording went off without a hitch.