I am trying to figure out the correct audio levels for broadcast. Researching online, it says peaks can be -12db or even -20 db. i have received HD footage, and imported it into FCP and it is hitting 0 and even clipping during the loudest parts such as gunshots.
Should the audio peak at -12db during the gunshots? It sounds awfully low for my taste.
[Jason Rodriguez]"I am trying to figure out the correct audio levels for broadcast."
Correct answer? Figure out where you're delivering to and ask them what they want. Too many variables to sort out if you're guessing or researching online.
[Jason Rodriguez]"imported it into FCP and it is hitting 0"
Lots of music and source footage will not be "broadcast level" - they're intended to be mixed, so they're likely taking advantage of the full dynamic range offered by digital audio recording. If you're mixing, it's your job to bring all of your audio elements into balance at the required output level.
[Jason Rodriguez]"It sounds awfully low for my taste."
This one gets me every time. Volume is relative. dBFs is not. -10dBFs is an exact digital value - if this sounds too low to you, turn up your monitors.
The correct level is whatever the broadcaster says it is. Usually the peak is around -6 or -3 but we delivered to one European network that required the peak to be -16 with the entire show normaled at -20. Way too low for my tastes, but that's what they required and that's what we delivered.
I did an article about this a while back here on the Cow. You need the specs from the broadcaster.
For stereo TV broadcasts, the peak level is usually either -10dB, or -12db. It varies from station to station, so it's worth checking. If your audio currently has a very wide dynamic range, you might want to consider applying some dynamic range compression using something like the Dynamics Processor, which ships with Final Cut as an Audio Units plug-in. This will allow you to keep the peak level at -12, whilst bringing the average program level up.
One other thing to consider is that there are also standards for dialogue levels. To add to the confusion, this is typically measure on another scale, called the PPM scale. Normally, a broadcaster will require dialogue to fall between PPM 3 and PPM 5 (UKPPM), although it can peak at 6 briefly. I'd recommend looking into how the PPM scale relates to the standard digital (dBfs) scale. Here's a useful reference chart.
If all of the above seems too much, just follow these simple rules (I'm assuming that the station has a -12dB peak level:
1: Make sure no dialogue falls below -20dB on the final cut pro meters.
2: Make sure the overall peak level of the whole mix doesn't exceed -12dB on the same meters.
Don't get confused with the PPM metering. Broadcast specs are ultimately measured by the digital dbfs metering. I have never owned a PPM meter and entirely base my mixes on digital scale, a fixed monitoring level and my ears.
With over 800 broadcast docos mixed to date, I have never needed any other metering scale. They are all approximates and only phons metering is a reasonable apparent loudness measure. PPMs are like an almost accurate digital meter so use digital metering as it is actually accurate and is the basis for broadcast deliverable spec. (yes I deliver all the time to BBC and others in Europe so lets keep it simple and honest for the poor fellow).
You absolutely must have a good compressor limiter which will stop all peak excursions over -10dbfs. FCP is a poor mixing and audio editing tool and my best advice is to go to an audio pro to mix properly. Frankly you won't stand a chance of getting a mix that has the apparent loudness and dynamics control that a pro with proper tools can achieve.