Topic: Introduction of Static during Post-Processing?

[3-Fan]

I recorded a fairly quiet show a couple weeks ago (R-09HR (24/48) > CA-9100 > AT853).  After I went through all of my normal workflow (SD > HD > Sony Soundforge (wave hammer, pencil tool, fades, resample, dither) > CD WAVE > TLH).  After I distributed the show I had one of the guys who got the show ask me about some static in an area.  I went through my FLAC 16 and FLAC 24 edited copies and there is static in the area he pointed out.  I pull up my master and listen to that spot and there is no static.  So I ask myself, what in the world could have caused the introduction of static?

I spent about 30 minutes really looking closely at the edited waveform in the static area.  Nothing really raised a red flag to me.  There were a couple areas where the waveform was cut level with some small fluctuations of maybe 0.5 db up and down in maybe 0.1 seconds.  Totally different looking than anything around them.  I took the pencil tool and tried to copy the adjacent waveforms and then relistened to the problem area.  Still static.  It is not in an area where I did any compressing or limiting, just normalize.

Anyone have any input on what might have caused this?  I will more than likely go back to square 1 and start over with the master.

Thanks,
Greg

[Patrick]

I recorded a fairly quiet show a couple weeks ago (R-09HR (24/48) > CA-9100 > AT853).  After I went through all of my normal workflow (SD > HD > Sony Soundforge (wave hammer, pencil tool, fades, resample, dither) > CD WAVE > TLH).  After I distributed the show I had one of the guys who got the show ask me about some static in an area.  I went through my FLAC 16 and FLAC 24 edited copies and there is static in the area he pointed out.  I pull up my master and listen to that spot and there is no static.  So I ask myself, what in the world could have caused the introduction of static?
I spent about 30 minutes really looking closely at the edited waveform in the static area.  Nothing really raised a red flag to me.  There were a couple areas where the waveform was cut level with some small fluctuations of maybe 0.5 db up and down in maybe 0.1 seconds.  Totally different looking than anything around them.  I took the pencil tool and tried to copy the adjacent waveforms and then relistened to the problem area.  Still static.  It is not in an area where I did any compressing or limiting, just normalize.
Anyone have any input on what might have caused this?  I will more than likely go back to square 1 and start over with the master.

Thanks,
Greg

I've had this problem with lots of DAW's before, from the freewares up to Pro Tools HD.  Hard drive errors or audio buffer underruns are an issue with digital audio, and these issues are often heard as pops, clicks, diginoise, etc.  Also, I've had certain programs put these kinds of noises in the file when using any sort of plug in or real time edits (crossfades, normalizing, etc).  So, as unhelpful as this post was, I'm still just going to chalk it up to it being a drawback of digital audio.

[boojum]

I get this in SAM 10.  I may not have enough horsepower for the DAW.  My computer is seven years old.  But I have a new one with a quad AMD and W7 and 4GB of memory I have to unpack.  I hope that will help.

FWIW I do not get the Rice Crispies (snap, crackle and pop) in the finished product when I play it on a stereo.  .

[Scooter123]

Diginoise. 

Generally an indicator of not enough memory when processing files.  I get it all the time on processing files, burning disks, and ripping disks. 

Easy way to fix it is do not run any other program when doing these things, or go to the computer store and get a couple more sticks of memory.  If XP and earlier, 4g is what you need.  If you are going Win764b, then you can run almost unlimited memory--I happen to run 8g on my hp z800. 

[kcmule]

If XP and earlier, 4g is what you need.

XP can't access more than 3gb.

[unclelouie]

That happened to me once when I forgot to dither to 16-bit before burning to CD.

[it-goes-to-eleven]

Diginoise. 

Generally an indicator of not enough memory when processing files.  I get it all the time on processing files, burning disks, and ripping disks. 

That's weird.  I *never* do.   And I'd stop using any tool (or OS) which silently corrupted my audio.  If there is a shortage of memory, that should trigger a warning, not corruption.

I think a major distinction should be made between processing audio vs. burning disks.  Those are very different activities.  Of course the original poster made no mention of disk burning being involved, so I think that is off-topic in this discussion.  I do all of my burning and ripping from Linux, never windows.

[unclelouie]

the original poster made no mention of disk burning being involved, so I think that is off-topic in this discussion.

After I went through all of my normal workflow (SD > HD > Sony Soundforge (wave hammer, pencil tool, fades, resample, dither) > CD WAVE > TLH). 

True, there was no mention of burning CDs, but it looks like the OP intended the final package to be for CD and mastered as such. That's probably what prompted the responses. I agree though - I've never had a memory shortage cause static - failed burns yes, but not static.  Again, the only time I heard static was when I burned 24/44.1 on CD by accident.

Like the OP, I do most of my processing in SF9.0 but I don't use wave hammer which, as far as effects go, could be the culprit: It is both a compressor and volume maximizer designed for 16-bit mastering and is being used early in the workflow. 

I'd skip wave hammer and try:
normalize to -0.1db > region/track > fades> resample > dither with noise shaping.

[3-Fan]

Thanks for the tips.

Additional information:

The show was never burned to a CD.  It has always been in "digital" format.  So the burning is not the culprit.

The reason for resampling and dithering is that "most" folks aren't equiped to listen to 24 bit.

The show was processed on a 1 year old laptop running Windows Vista with IIRC 3Gb of ran.  I was doing other things at the same time while I was processing the show on the laptop.

On second thought, I did use Wave Hammer initially, but didn't like the results so I switched over to using :  Effects > Dynamics > Graphic > hard limit to whatever I needed.

I did normalize a couple times.

I'm gonna start back over with the master and only run Soundforge for the entire time I am processing and see if I can get a clean finished product.  I've never had this issue before but don't recall if I was ever doing other things at the same time.

Greg

[anr]

When I used Soundforge 6.0 in XP with 4GB (sees 3.2GB) this never happened, no matter how many programmes were running in the background.

Now I use Windows 7, which sees all 4GB RAM, it happens if I try to run anything else. 

I'm thinking of using this PC as audio only and get a cheapo for office/internet.  I'm sure many here would say the same.

[ghellquist]

There are a lot of ways you can do this unintentionally.

One of them is to first normalize to 0dB and then dither. Dithering can in a way be descrided as adding a slight amount of noise, which means the signal may go above 0dB = clipping. Solution = normalize to -1dB to leave some headroom. Also set any final limiter to -1dB instead of 0dB.

A limiter always modifies the sound once it kicks in (above the threshold ) . So you need to learn how to use it. You might not hear the modification which might be because your listening chain might be too "forgiving" while others with a different chain may be disturbed by the sound (by the way, this is not necessarily a question of a "better" chain, just different) . A too slow attack time on the limiter will let the transient pass through for most limiters, and this might give a signal above 0dB.

Doing only mixing / mastering uses very little RAM in most programs. A buffer overrun then is not related to too little RAM but instead to too slow audio card (perhaps you are using the built-in card, some are good, some less good). Unless you run ASIO to the audio card you cannot really be sure what happens -- windows can do a lot of background fixing of the sound such as not telling you that it is beeing resampled. Still, you should not have trouble going from the program out to a file regardless of the hardware.

Disk hardware errors simply should not show themselves this way. Hard discs have error detection and will give hard errors (program halts) instead of pops or clicks.

// Gunnar

[rastasean]

If XP and earlier, 4g is what you need.

XP can't access more than 3gb.

what about xp 64 bit?

[kcmule]

If XP and earlier, 4g is what you need.

XP can't access more than 3gb.

what about xp 64 bit?

Good point.  I was assuming 32 bit OS.