Understanding mixing

[Alex H.]

Do any of you use the technique of shutting off your limiters on your mixer then setting tone on your mixer (using an outside source as CV2 suggested) so that it shows just below or at 0dBFS (+20dBVU) on the mixer and the same on the camera/recorder using a blend of levels that maintain some unity on both. Then, when that has been done, turning the limiters back on the mixer and running the dialogue peaks in the -6 to -8 range on the mixer scale. I have been doing this on cameras that I am not confident of the -12 or -20 dBFS reference mark. I haven't had any complaints with the results but I sometimes worry I might be setting my self up for a fall. Any comments?? Thanks.

With analog mixers, which most portable production mixers are (save for the Sony), tone should be set to 0dB. That is equal to -20dBFS, though with the DVX the reference mark given is for -12dB.

If you set tone to hit +20dB on an analog mixer, you are setting your reference at the top of the analog device's tolerance. Though some mixers may have that headroom, many don't meter past +7 or +10, and may start to distort earlier than others. This whole process is about calibrating the ideal signal-to-noise ratio on your mixer and in-camera as well as ensuring level continuity from mixer to camera.

And external tone sources are only necessary when your mixer doesn't have one built in. If you are using an SD mixer, you have an internal tone generator.

 

[Baker Park]

Two cameras have given me some concern. A couple months ago I was working with two PD170s. Both were set to the same level with standard tone. (I taped the audio setup button). Result was one of them was +- 5 db lower than the other. The second camera, which I have asked about before is the JVC 250. I use the hot tone check more as a safety but I have ended up trusting that more that standard tone when dealing with a sometimes questionalble indicator on a prosumer camera. The 250 is interesting to me as it has two settings -12 and -20 in its menu. I have never been clear if this is just a reference change or involves other aspects of the recorded signal. Oleg has a lot of time with this camera and suggested this some time ago.

So my question is what is the downside of this approach??

CV2 you just posted while I was typing. Thanks. I am using a SD mixer and I tend to press the limiters when working with fairly predictable dialogue. I run my average peaks in the -10 to -14 dBFS range. Am I making a mistake thinking that the best levels for digital camera recording is as hot as possible with out hitting the audio blue screen of death at 0dBFS?

 

[Alex H.]

Am I making a mistake thinking that the best levels for digital camera recording is as hot as possible with out hitting the audio blue screen of death at 0dBFS?

Pretty much.

In the analog world, we have lots of headroom past 0dB and in certain cases (especially music) a light touch of analog distortion is even desired. That mark, 0dB, is also the best spot for signal-to-noise ratio. Keep in mind also that analog has self-noise no matter where it goes, so recording 20dB too low and bringing it up later in post in an analog system will add noise floor.

In the digital world, there is nowhere to go after 0dB. The signal starts clipping into square waves and sounds terrible... and there's no coming back from that. Digital also has no self-noise. The only noise floor you have to worry about is what came from pre-amps and the acoustic environment when the signal was recorded. So the signal can be brought up in post without the worry of added noise.

The marks of -12dB and -20dB in digital recording both give you headroom to work with, but that headroom is a safety margin for sudden peaks. Think of it like the last 25% of the runway... the plane, most often, will never hit that patch of tarmac, but it's there just in case. If you record as close to 0dBFS as possible, you've left yourself nowhere to go... and for what? You'll be mixing the dialog to -12dB or so in post anyway.

 

[Baker Park]

It is hard for me to sort out. I tend, with a mostly forum education, to think of the 0dB orientation as a product of the analog world. You aim for it because it is the sweet spot. If you go over, that is what the head room is for. When it comes to the digital world of modern cameras combined with the quality of limiters (both for their exactness and soft push down) it seems to me that the ceiling is now much more important than the sweet spot. Especially in the 16bit world and pro-sumer cameras with their own brand of floor noise. I work with a couple editors who want to see -6dB fairly often in the stuff they work on from lower grade cameras.

Would you compare the demands of analog vs digital recording and where you place your bets? By the way, to me, locking in where the ceiling is on the recorder is all about avoiding decapitation.

 

[Alex H.]

Would you compare the demands of analog vs digital recording and where you place your bets?

Not sure I understand your question.

Long story short, I use an analog field mixer and shoot with a DVX. It's the same mixer I've used for years with both BetaCam SP and DigiBeta.

Mixer tone = 0dB.
Mixer at 0dB = BetaSP (analog) at 0dB = DVCAM/MiniDV at -12dB = DigiBeta at -20dB.

In the DigiBeta world, I calibrated the reference tone at -20dB but my dialog peaked at -10dB. (Also the DBeta delivery specs that networks from Scripps, Discovery, and MTV wanted... reference at -20dB and peak at -10dB.)

With my DVX, the tone is calibrated to the -12dB mark though my dialog may peak to -6 or so on occasion. As long as your average with miniDV/HDV cameras is at -12, you are safe to peak at -6 and you aren't losing anything. But to set your dialog at -6 and peak as close to 0dB as possible is not a good practice for the reasons I mentioned before.