audio distortion on mic input contaminating other channel at DVinfo.net

Jeremy Davidson · September 2nd, 2004, 12:40 PM

I taped a choir concert as a personal favor two days after getting my brand new GL2 (lack of experience -- my first mistake). I used a homebuilt adapter cable to break out the stereo mic jack into two XLR cables. One went to the mixer for the house signal, and the other ran to a Shure SM57 for ambient pickup. I didn't have headphones (my second mistake), so I just switched into manual mode, flipped on the built-in attenuator (due to the loud mixer feed), and visually adjusted the levels using the onscreen meters. I was surprised to see how "constant" the mixer feed was ("oh, they must be using a compressor").

I played the tape back and found out exactly why it was so level: it was distorted so badly that it rarely ever left the top of the preamp's headroom (even though by adjusting the level controls it always stayed below 0db). Oops.

The strange part was that on particularly hard hits the distortion would cross over and contaminate the audio coming in through the other channel (the SM57). I've since realized that I was probably overdriving the camera by about 20-30db (ouch), so this may be an extreme case, but has anyone else ever encountered something like this?

As for the rest of the story, I had my little Optura20 also running, so I was able to salvage the audio from its on-board mic and align it to the GL2 video. I now never go anywhere without headphones (Sennheisers, though if I did it again I'd probably get the Sony 7506's... but that's been debated plenty on another thread) and some ProCo 20db inline attenuators (B&H).

Don Palomaki · September 2nd, 2004, 07:25 PM

With 30 dB of ovedrive it is likely that you will get cross coupling to the other channel. The audio circuits probably have on the order of 60 dB channel separation (after all, they are in two channels on one stereo audio integrated circuit). To simplify a bit, if one signal is driven +30 dB over the top, allowing for channel separation, it nets to about -30 dB in the other channel which will be audible.

Hank Freeman · September 2nd, 2004, 07:27 PM

You need to understand the difference between 'line' level audio and 'mic' level signals. The GL2 microphone input is just that. It looks for a mic level signal. Most all house or mixers will provide line level. You took a line level signal and interjected it into your mic level input. compression did take place, but it was the camera attempting to compress the line level. In a nut shell, line level signals are 40-50DB greater than mic level. Line is 100 times greater than mic from a voltage standpoint.

I think everyone learns this earily in their videography experience so welcome to the club.

Jeremy Davidson · September 2nd, 2004, 09:44 PM

I should have known better with all the work I've done with pro audio equipment, but since it looked OK on the meters, I thought I had it below clipping. I must have still been overdriving something in the mic preamp (upstream from the level controls).

I have since done another shoot with the same setup, except this time I put 40db of attenuation on the signal coming from the mixer (with an ambient SM57 on the other channel) and no camera attenuator, with much better results.

I always do my best to get the highest quality recording I can, but if I was going to have this problem, I'm glad it happened on that project (like I said, it was just a personal favor a.k.a. home video).

Don, thanks for the heads-up on the 60db channel separation -- I'll hide that little tip away along with all those just-in-case things I sometimes need to know. I recall reading a post on here about sending the same signal to both channels but at different levels (so that you could select the best one in post), but this would seem to indicate that if the more-sensitive one clipped too hard it could still take out the other channel. 'Just me thinking out loud.

Anyway, thanks for the help!

David Ennis · September 6th, 2004, 04:22 PM

Jeremy, another thing I learned here (after suffering the consequences in the field, of course) is worth filing away.

The recording level indicators do not tell you what's happening at the mic input. If you overload the input and cause clipping you won't know until you hear the sound. Even a hot condenser mic can overload the GL2's mic preamp if you don't turn the attenuator on when needed.

The recording level indicators will just tell you that you're not adding distortion at that stage, which, according to my educated guess, is the input of the analog to digital converter.

Hank Freeman · September 6th, 2004, 06:50 PM

...and always, always monitor your recording with headphones! I use a noise-cancelling set from radio shack that works just fine.

September 2nd, 2004, 12:40 PM	#1
Jeremy Davidson Regular Crew Join Date: Aug 2004 Location: Eau Claire, WI Posts: 167	audio distortion on mic input contaminating other channel I taped a choir concert as a personal favor two days after getting my brand new GL2 (lack of experience -- my first mistake). I used a homebuilt adapter cable to break out the stereo mic jack into two XLR cables. One went to the mixer for the house signal, and the other ran to a Shure SM57 for ambient pickup. I didn't have headphones (my second mistake), so I just switched into manual mode, flipped on the built-in attenuator (due to the loud mixer feed), and visually adjusted the levels using the onscreen meters. I was surprised to see how "constant" the mixer feed was ("oh, they must be using a compressor"). I played the tape back and found out exactly why it was so level: it was distorted so badly that it rarely ever left the top of the preamp's headroom (even though by adjusting the level controls it always stayed below 0db). Oops. The strange part was that on particularly hard hits the distortion would cross over and contaminate the audio coming in through the other channel (the SM57). I've since realized that I was probably overdriving the camera by about 20-30db (ouch), so this may be an extreme case, but has anyone else ever encountered something like this? As for the rest of the story, I had my little Optura20 also running, so I was able to salvage the audio from its on-board mic and align it to the GL2 video. I now never go anywhere without headphones (Sennheisers, though if I did it again I'd probably get the Sony 7506's... but that's been debated plenty on another thread) and some ProCo 20db inline attenuators (B&H). __________________ There's no way for you to know if what I'm saying is true unless you know what the truth is, and there's no way for you to know what the truth is unless there is a truth that you can know. -- Frank Peretti

September 2nd, 2004, 07:25 PM	#2
Don Palomaki Wrangler Join Date: Sep 2001 Location: Northern VA Posts: 4,489	With 30 dB of ovedrive it is likely that you will get cross coupling to the other channel. The audio circuits probably have on the order of 60 dB channel separation (after all, they are in two channels on one stereo audio integrated circuit). To simplify a bit, if one signal is driven +30 dB over the top, allowing for channel separation, it nets to about -30 dB in the other channel which will be audible. __________________ dpalomaki@dspalomaki.com

September 2nd, 2004, 09:44 PM	#4
Jeremy Davidson Regular Crew Join Date: Aug 2004 Location: Eau Claire, WI Posts: 167	I should have known better with all the work I've done with pro audio equipment, but since it looked OK on the meters, I thought I had it below clipping. I must have still been overdriving something in the mic preamp (upstream from the level controls). I have since done another shoot with the same setup, except this time I put 40db of attenuation on the signal coming from the mixer (with an ambient SM57 on the other channel) and no camera attenuator, with much better results. I always do my best to get the highest quality recording I can, but if I was going to have this problem, I'm glad it happened on that project (like I said, it was just a personal favor a.k.a. home video). Don, thanks for the heads-up on the 60db channel separation -- I'll hide that little tip away along with all those just-in-case things I sometimes need to know. I recall reading a post on here about sending the same signal to both channels but at different levels (so that you could select the best one in post), but this would seem to indicate that if the more-sensitive one clipped too hard it could still take out the other channel. 'Just me thinking out loud. Anyway, thanks for the help! __________________ There's no way for you to know if what I'm saying is true unless you know what the truth is, and there's no way for you to know what the truth is unless there is a truth that you can know. -- Frank Peretti

September 6th, 2004, 04:22 PM	#5
David Ennis Fred Retread Join Date: Jul 2004 Location: Hartford, CT Posts: 1,227	Recording Level does not tell all Jeremy, another thing I learned here (after suffering the consequences in the field, of course) is worth filing away. The recording level indicators do not tell you what's happening at the mic input. If you overload the input and cause clipping you won't know until you hear the sound. Even a hot condenser mic can overload the GL2's mic preamp if you don't turn the attenuator on when needed. The recording level indicators will just tell you that you're not adding distortion at that stage, which, according to my educated guess, is the input of the analog to digital converter. __________________ "Nothing in the world can take the place of persistence..." - Calvin Coolidge "My brain is wired to want to know how other things are wired." - Me

September 2nd, 2004, 07:27 PM	#3
Hank Freeman Regular Crew Join Date: Nov 2003 Location: Indiana Posts: 37	You need to understand the difference between 'line' level audio and 'mic' level signals. The GL2 microphone input is just that. It looks for a mic level signal. Most all house or mixers will provide line level. You took a line level signal and interjected it into your mic level input. compression did take place, but it was the camera attempting to compress the line level. In a nut shell, line level signals are 40-50DB greater than mic level. Line is 100 times greater than mic from a voltage standpoint. I think everyone learns this earily in their videography experience so welcome to the club.

September 6th, 2004, 06:50 PM	#6
Hank Freeman Regular Crew Join Date: Nov 2003 Location: Indiana Posts: 37	...and always, always monitor your recording with headphones! I use a noise-cancelling set from radio shack that works just fine.