SOS Help - Digital Clipping

[sus4chord]

My son produces and runs sound for a podcast. He had to record off the feed from the board at a live event (their equipment, not much setup time, etc), which was actually from a secondary board chained to the main house board. The house audio sounded fine, but it was coming off the main board. The resulting recording from the second board has a lot of digital distortion. Next time, he'll be sure to check the gain settings on both boards, but that doesn't help now. You can really hear it when the speakers' start speaking louder, but not so much when they speak normally. I'm posting a clip here, where you can hear the distortion at the beginning of the clip and it goes back to somewhat normal later on. You can definitely see the flat topped wave forms when you zoom in. I'm really looking for a quick turn around if anyone has any guidance for filters and settings to make it sound better. The original file is 1.6GB, so obviously I'm not posting that (just a snippet), but the snippet is representative of the rest of the recording. I hope that made sense.

I tried using the Forensics De-Clipper filter, but didn't have much luck with it (and it's probably just my own inexperience with it).

Thanks so much in advance for any help.

https://www.dropbox.com/s/cj5bs9ufmk...ippet.wav?dl=0

[Craig Maier]

Try first running the de-clipper after decreasing the gain of the signal (unless you did that automatically). Next, run the 4th order low pass filter set for around 4500 Hz. That is probably as good as you are going to get that.

Craig

[sus4chord]

Thanks, Craig, I will give that a try. I really appreciate the quick feedback.

[Craig Maier]

I reviewed the file and found that de--clipping first and then using the 4th order Low Pass Filter set for 4000 Hz and Butterworth response cleaned up the distortion reasonably well.

Craig

[sus4chord]

Thanks, I'll try that when I get home from work. I'm sure my son will be pleased. The help is much appreciated.

[Craig Maier]

Let me know if your son thinks that process helped to improve the audio track.

Craig

[sus4chord]

We were definitely able to get a cleaner version. Some of the recording was just plain oversaturated, but the bulk of the content was cleanable. Next time, he's going to bring equipment to pull the recording off the main board (since it sounded fine in the house). The problem of course with a live show is you don't get a do over like you can in the studio. I really appreciate the quick response and help. One of the many reasons I appreciate both the Diamond Cut tools and this forum.

[Craig Maier]

One of the most common causes that we have seen in this situation is as follows: PA/Mixing board feeds the house power amplifiers. Videographer takes a Line Level output from said board output buses into the Mic input of the video recorder. The result is something that sounds like that file. I hear several of these types of files a week and it usually has the mentioned cause. The first stage of a Mic input is sensitive down in the mV level and the Line Outputs of the Mixer are in the Volt range; this a large miss-match (and overload) of the first stage amplifier in the video system.

Craig

[sus4chord]

I will pass this along to my son. Thanks.

[Craig Maier]

The solution to this is one of the following:

1. Take a line level output from the mixing board into the line (not Mic) level input of the video recorder (assuming it has one).

2. Construct or purchase a 30 dB or 40 dB PAD (L or T) and insert that between the line out of the mixing board and the Mic level input of the video recorder. You can find a table of standardized (normalized) values in the Diamond Cut Users Guide or Helpfile if you decide to construct your own PAD. I think the table is in the charts and graphs section, as I recall.

Craig

[sus4chord]

Thanks, Craig. It all helps.

[ebrjvd]

I recommend to insert an isolation tranformer between the PA board and the recorder to avoid ground loops and garbage from switching power supplies...
Jos

[Craig Maier]

Jos,

That will fix ground loop hum/buzz induced noise problems. But, it will not fix the clipping that occurs when a line output is fed into a mic level input. That is why I recommend using a PAD to reduce the signal level from line level down to mic level when only mic inputs are available. Of course, if a line level input is available, that is the best way to connect a line level output.

Craig

[ebrjvd]

In that case, a DI box may do the job. It is used to feed high level inputs to mic input (usually with switchable pads and isolation Xformer).
An active DI box needs phantom or battery power (e.g. Behringer DI100, or BSS AR133).
Cheers,
Jos

[Craig Maier]

Another way to decrease the signal level from Line Level to Mic Level is to use something like a Shure A85F Line Matching Transformer backwards. Normally, these are used to couple balanced low Z balanced line level signals to high-Z Inputs. They are passive devices, so no phantom power is needed. To use these to decrease the signal level rather than increase the signal level, just run them in the reverse direction. In other words, the line level output of the mixer board would feed into the high Z side of the transformer and the balanced low Z side of the transformer would feed into the mic level input of the video recorder. One is needed for each channel (two for stereo). This model transformer will decrease the signal level by 24 dB and provide isolation as an added benefit. No batteries required!

Craig