What Level to Set Gain in NLE?

[Peter Moretti]

I'm working on footage with background music and two tracks of voice over.

I'm not sure approximately what level my meters should be peeking at. I know when recording they should not go above 0. But while editing I can change the gain for the tracks. I like the way their levels sound relative to each other. But I don't know a good rule of thumb for the cumulative peek level for typical voice over on top of music audio.

Thanks much!

[Steve House]

Originally Posted by Peter Moretti

I'm working on footage with background music and two tracks of voice over.

I'm not sure approximately what level my meters should be peeking at. I know when recording they should not go above 0. But while editing I can change the gain for the tracks. I like the way their levels sound relative to each other. But I don't know a good rule of thumb for the cumulative peek level for typical voice over on top of music audio.

Thanks much!

What is the ultimate destination of the mix - SD broadcast, HD broadcast, DVD, theatrical, web? The headroom of your delivery target determines the relative maximum peak to average levels and the "never exceed" peak level. The overall subjective levels should be relatively constant so the listener doesn't have to grab the gain control when the voice enters or leaves. In other words, the average level of voice+music versus music alone should remain about the same. When the voice enters, instead of setting its level higher than the music so it rides on top, the broadcaster's axiom is to dip the music about 3dB, maybe up to 6 dB, from the level it's at when playing by itself. The amount of dip depends in part on the nature of the musical material - if it's lacking a bit in the midrage so there's a hole for the voice frequencies you can get by with less gain reduction than if the midrange is predominant. That strategy implies that the voice+music peaks wouldn't be signifigantly different from the music-only peaks.

[Peter Moretti]

Steve,

The material is a promo for a friend's camp. Its ultimate desitantion will be standard def DVD and probably web. Where should the sound be peeking at for these modes of delivery?

Just out of curiosity, how would the level differ if I were working on a project for theatrical release? (I'm also working on a doc that I have big dreams--errr delusions, LOL--for.)

Thanks much as always.

[Steve House]

Originally Posted by Peter Moretti

Steve,

The material is a promo for a friend's camp. Its ultimate desitantion will be standard def DVD and probably web. Where should the sound be peeking at for these modes of delivery?

Just out of curiosity, how would the level differ if I were working on a project for theatrical release? (I'm also working on a doc that I have big dreams--errr delusions, LOL--for.)

Thanks much as always.

For SD and broadcast, tone at -20dBFS and never exceed -10dBFS on peaks. DVD and theatrical, tone at -20 and never exceed -2dBFS. Web, tone at -12dBFS, ever exceed -2.

[Phil Bambridge]

Originally Posted by Steve House

For SD and broadcast, tone at -20dBFS and never exceed -10dBFS on peaks. DVD and theatrical, tone at -20 and never exceed -2dBFS. Web, tone at -12dBFS, ever exceed -2.

SD? Standard Definition? If so, standard definition what? I think I might have the wrong abbreviation here, as DVDs count as SD too, of course. Oh, wait, was that meant to say "SD and HD Broadcast"?

Is the higher tone volume on web due to anticipated worse signal to noise ratios, what with typical bad speakers, bad sound cards, and all that compression?

Is the broadcast lower peaks due to making sure your television doesn't make any sudden extra-loud noises (due to some regulation)?

So, what would your advice be for someone who has no idea where their product might end up? I'm guessing you can assume best-case scenario (wide dynamic range) then compress and lower gain as appropriate?

Cheers in advance.

[Roshdi Alkadri]

steve is a huge help on this subject and i learned from him. It is best to know your final target, this way you have one master project and can output to different media. One technique i found helpful is to set tone at -20 and mix right up to -2, then say i need a broadcast mix, then all you gotta do is turn down the master 10db so now your peaking at -10 and then put a limiter to keep the lid on it. then for web, turn it down a little more. For DVD and theatrical ( i did not mix theatrical yet :) but generally it can handle the full volume range.

[Steve House]

Originally Posted by Phil Bambridge

SD? Standard Definition? If so, standard definition what? I think I might have the wrong abbreviation here, as DVDs count as SD too, of course. Oh, wait, was that meant to say "SD and HD Broadcast"?

...

My typo ... sholuld have said SD broadcast without the "and." (SD VHS tape would be similar with maybe even less headroom so peaks shouldn't exceed about -12dBFS, but who goes to VHS these days?) HD broadcast has the wider dynamic range and peaks can go up to about -2 or -3dBFS but HD has this whole dialnorm thing going on as well and frankly I haven't quite wrapped my mind around how that all works as yet.

Roshdi - good strategy. Also try it with about 5 dB compression and 5 dB make-up gain so you don't have to lose quite so much overall level.

[Roshdi Alkadri]

thanks steve good points there. I usually place the ultramaximizer on the output bus.

[Petri Kaipiainen]

The TV company I have worked for asked the levels to be -20 to -12 dBFS for dialog and sound in general, all peaks hitting -6 dBFS had to be listed on a report form, and the absolute tops allowed was -3 dBFS. This was for SD broadcast.

[Phil Bambridge]

My worry would be that if you actually tried to exploit this higher range with HD, you might run foul of the tv company more aggressively compressing the sound. Meh. At the end of the day someone might be watching it on a 2.5" screen on their mobile phone, sat on a crowded bus as it drives past a construction site...