Calibration help
 

I would like to ask for some help to see if I am calibrating correctly, as I have picked my method up from various posts and some books I have read, but none specific to my situation with Canon GL2. It seems to have worked ok, but I'm trying to improve my audio and thought it best to ask for help, suggestions, etc.

I am using a GL2 and a MixPre, and have switchable pads (-10,-20,-30).

I have the pads coming out of the MixPre to the GL2 to adjust for the lack of Line/Mic switches on the GL2 and go into the GL using a dual xlr connector into a 1/8" jack. I am typically set on -20 on the pads (no specific reason, just liked it), turn the gain on both MixPre channels down to approx 1/4 of their max, engage the tone oscillator and adjust the camera volume controls to just exceed the 12 point on the on-camera scale. While shooting I monitor the sound from the camera jack, not the mixer.

Your comments and suggestions are sincerely appreciated. I have a great deal to learn about audio, for sure. Thanks!

I own a MixPre and have recently been looking into preparing for situations like this, but my knowledge isn't tested...

The MixPre's limiters are (by default, screws completely counter-clockwise) set to kick in at +18dBu. The calibration tone is 0dBu.

So the primary concern is that you MixPre's 1kHz tone is read at -20dBFS on the camera (so the MixPre's limiters activate when you're at -2dBFS at the camera and near clipping). Hopefully the GL2 has bars indicating the -12 and -20 points to make this calibration easy, but not always. Something that can work is connecting the camera to FCP, opening log and capture which DOES show -12 and -20dBFS points, and then adjusting levels using that screen.

Then during the take, since your MixPre is a peak indicator, try to have the dialog peaks going to ~+8 to 12dBu on the MixPre (which translates to -12 and -8 peaks at the camera).

For pads, easiest way is a Beachtek adapter, granted you have pads (I'm not sure if the Beachtek units have a transformer in them to help deal with the impedance difference). Granted your current pads are probably going to work just as well.
I believe you generally want to pad pro line level signal (pro line level is +4dBu, consumer line level like in RCA jacks is -10dBV) 50dB as per Line signal to microphone input.

Although as long as you have the calibration levels set correctly with proper headroom, that's more important than how much your pad the signal.

Thanks Alex, great info.

Are there any other suggestions on this?

I can't speak to your exact setup but what I CAN offer is that on all pro audio mixers, 70 - 75% of "all the way up" is NORMALLY "unity gain" (the point where the signal is being neither amplified or attenuated): on your camera I would strive for 3/4 of the way up in regards to gain and matching everything backward to get your reference. This is to maximize signal to noise. Every time you add gain you add noise. When you subtract gain, you lower your signal in reference to your "noise floor" or the amount of self generated noise in the audio components.

Alex's advice looks right to me. To pad down from pro line level to mic level you should be attenuating between 54 dB to 35 dB. Many people go for the hot mic level signal (at -35 dB), and don't sweat the impedance mismatch because it lets them dial down the noisy gain on the recording device. This can be tricky for very loud sound sources, as there is potential for distortion, so decent headphones are a must.

Does your XLR adapter not have a line attenuation switch? Personally, I'd invest in one that did. If you shoot much at all with the GL2 it is indispensable. I believe the GL2 does have a marking for -20 dBFS, but you can cross check it by getting a tone plug, cranking the gain on the mixer until it reads +20 (on the mixer) with the limiter switched off. Adjust your gain and attenuation at the camera until it just reads 0 dBFS at the camera. Now you should be calibrated. Turn off the tone plug and throw the tone switch on the mixer. If it reads -20 dBFS on the camera you can be doubly sure that you are calibrated properly. Do some tests to be sure that the limiter is indeed kicking in at +18 on the mixer. In fact, if it's adjustable I'd take it down to +16 and give yourself generous headroom just to be safe.

What XLR adapter are you using? That's not a place to cheap out. You need at least Sign Video or Beachtek quality, that is all metal construction with passive fader knobs that don't introduce any clicks or hissing when you turn them.

XLR-PRO XLR adapter

Marco, I have a Beachtek, but have not been using it, replacing it with an adapter made up by Front End audio to get the XLR input into the 1/8" camera jack. The thinking was that fewer ciurcuits in the sound chain were better, so I was trying to come from the mixer into the camera as directly as possible, going through the AT pads rather than the Beachtek. I am way over my head here, and sincerely appreciate all the help I can get.

A passive box like that is about as transparent as you get. It doesn't add any noise itself, for one thing. That's one of my problems with active boxes like the Juicedlink. I would also expect it to be more resistant to RF interference than any custom cable you could have made.

The thing with the GL2 is you need to set the camera gain at about 1/3 of the way up and adjust your attenuation accordingly. That's going to give you the least hiss and distortion. If it were me, I'd go into the box with 35 dB of attenuation, set the camera to 1/3 as mentioned, and then dial it in with the passive faders until you are calibrated. Or you could just flip the switch and be done with it, but that might force the camera gain higher than is ideal.

Another advantage of the box is it gives you more control over feeding the camera a dual mono signal, or managing right and left channels separately.

Having tested both, give me active any day.

The key is to have the first item in the chain provide the cleanest gain available, as is generally true with the juicedLink, and definitely true when my 5D Mark II is the recorder.

It's especially important to add gain upstream of any unbalanced line. Unbalanced lines are susceptible to interference. Give me 6dB more signal into an unbalanced line and I'll give you 6dB less interference noise. Yes, an active device adds noise, but it increases the signal too. As long as the signal to noise ratio isn't significantly degraded, it's a good bargain.

And it's not quite true that a passive device adds no noise. Every resistor, capacitor, inductor and potentiomenter adds thermal noise. Carbon resistors add more than wire wound resistors. If you have enough signal, the thermal noise is swamped out; if not, thermal noise can play a role.

Here is my noise evaluation: 5. Canon 5D Mark II Audio Exposed - Noise on Vimeo

I agree you want to add the gain upstream, but that is already the case, since we're talking about a line level signal. It's just a matter of how much to attenuate it, and how. It doesn't make any sense to attenuate the signal, and then boost it again with an active box at the camera. This makes the assumption that you are using a decent mixer (to feed the line level signal) that has cleaner preamps than the Juicedlink.