Where do you set your audio levels on the XL2? How close can you "push it" to zero before overmodulation?

-12dB is a pretty standard target. You'll have a little headroom for unexpected peaks.

I try to have it to where the audio levels don't get much higher than the green dot.

I try to have it to where the audio levels don't get much higher than the green dot.Which is at -12dB :)

What?

Oh. That's the sound guy's problem.

What Jarrod said.

Also, be sure that the sensitivity of the wireless mikes are down a notch, or two to -6db, or -12db. For some stupid reason the Senny E100's are at 0db, or +6, or something. I might even have mine down to -18db.

Where do you set your audio levels on the XL2? How close can you "push it" to zero before overmodulation?

Let me take a crack at this since I have bench-tested and field experimented with my XL2 and have learned a few things that are not stated clearly in the manual. Bear with me. I’ll try not to get too technical, but some specs are necessary.

There are two distinct possibilities of overloading the XL2 and causing distortion (clipping).

First there is analog clipping, or to put it another way, overloading of the microphone input preamp which is not influenced in any way by the side knobs on the outside of the camera.

This will occur when a signal greater than -26.5dBV (0.0473 volts rms) is present at the microphone input. The only way to introduce a higher signal and prevent analog distortion is to turn on the 20dB pad using the switches on the back of the camera. This will attenuate (reduce) the signal by 20dBV enabling a hotter microphone outputting up to a -6.5dBV (0.473) volt signal to be used without distortion.

Check your external microphones spec sheet. High quality mics will tell you what output is at a given SPL (Sound Pressure Level). It might state for instance “-66dBV at 1 Pascal.” 1 Pascal is a common reference and simply means 94 decibels SPL. Another older reference that you may still see is “@1 dyne/cm2”. This means at 74 decibels SPL.

Here is an example of what that means in my world. I am a train buff. I videotape trains which of course are extremely loud things. I use a Crown SASS-P microphone which outputs a -44dBV@1pascal. (94 decibels).

There is only 17.5dBV difference between -44dBV and -26.5dBV so that means the loudest sound that the microphone can pickup and send to the XL2 without clipping is 111.5 decibels SPL.

Trains horns can reach 115-120 dB SPL when the mic is about 25-30 feet from the track on a tripod and the horn blows as it is passing the mic, so without enabling the pads I would get distortion.

Remember, again this has nothing to do with the knobs. I’ll get to them very soon.

I decided to go with external inline pads of 12dB which give me distortion free sound up to 123.5dB SPL, more than the loudest situation I’ll face at trackside, but not as much attenuation as the built in 20dB pads so I don’t have to turn up the gain knobs on the camera to a high level to get a good digital recording level.

(This scenario could apply to a rock band in a club, an airplane taking off or landing or a car race, so it is not limited to my own peculiar hobby.)

If I haven’t lost everyone yet, here’s where those gain knobs come into play. They control the digital audio level which will be recorded on the tape.

Digital audio only works one way. All levels are expressed as minus decibels referenced to Zero Decibels Full Scale (0dBFS or absolute maximum digital level). This means that a -3dBFS signal is 3 decibels below the maximum level that can exist in the digital domain. The green dot in the XL2 indicates -12dBFS or 12 decibels below the maximum, which is referred to as 0dBFS. It is safe to record with the levels falling anywhere along that scale with the exception of the very rightmost tick which will light up red indicating clipping. You ideally want your recording levels between -12dBFS and 0dBFS most of the time.

One more thing, if you haven’t quit on me yet. The maximum analog audio output of the XL2 for 0dBFS is -10dBV which is a rather low audio signal. This may cause you to think you did something wrong and raise the gain knobs the next time you record in a similar situation. Don’t do it.

By comparison your CD player probably outputs somewhat over +5dBV for a 0dBFS digital level and my DirecTV box outputs -12dBV for average dialog on a news channel. So if you want your maximum playback volume to equal your usual CD listening level you’ll have to turn your amp to a higher setting when listening to the XL2’s audio.

In the digital domain this doesn’t affect anything—a digital 0 is a digital 0 so if you import the video with audio into your computer for editing and then burn out a DVD, the 0dBFS level will stay the same from XL2-computer-DVD, so your finished DVD will play back at whatever reference analog level that your player outputs for 0dBFS. (Probably at or close to what your CD player puts out) so the finished DVD will be louder than the XL2 playing the raw tape into your amp/receiver was.

In other words you need not adjust the audio levels up during editing unless the maximum sound was recorded significantly below the -12dBFS mark on the XL2’s scale.

The reason that the -12dBFS mark is highlighted on the scale, and the manual suggests setting your normal level at or close to it is so that you have 12 decibels of “headroom” in case an unexpected loud sound happens, such as a car horn blowing while you’re interviewing someone on the street.
So there is nothing magic about the -12dBfs mark.

I work it the other way. Since the main sound I am after is always louder than the ambient sounds in nature, I have my levels calibrated so that 124 decibels of SPL at the mic will give me 0dBFS on the tape. That way I know that all of my sound should be below clipping.

If I shoot a subject other than trains I recalibrate accordingly, such as removing the pads and and adjusting the gain knobs.

Personally I never use "Auto Level" because I want the sound spectrum to be as natural as possible and really do not like audio compression/limiting which is how "Auto Level" works.

I don't know if it is just that my mike is hot (it's a AT 4073a), but we almost always shoot with attenutation ON for the shotgun. If attenuation is not on then in many instances when people are speaking loudly, the audio will distort and no amount of turning the gain down on the camera can help it.

Our rule of thumb has become that when the subject is speaking normally and the levels look good in the EFV (around -12 db aka, the green dot) then the gain knob on the side of the XL2 should be set around 12 o'clock or 1 o'clock. If the gain knob has to be turned down to 11 o'clock to get the levels to behave then you're going to get distortion. So we turn on attentuation.

It's more about keeping the levels at around -12db, than about the position of the knob itself. The levels are the levels, adjust accordingly.