LEVELS in post

[Roshdi Alkadri]

Hi guys, we recorded dialogue on location at -12DB and we're about to post the audio, what are the general levels to peak at. There will be dialogue, music, foley and ambience. I did read through the subject and know to start with dialogue and then add to it....so

In post should dialogue be also peaking at -12, then what levels would you set music, fx, etc.

thanks

[Jarrod Whaley]

I'd actually recommend this: set your dialogue tracks so that they average around -20dB with occasional peaks above that. From there, set the levels of effects, foley, music, etc. wherever the mix sounds best to you and/or makes sense. A big explosion should be mixed louder than dialogue, dialogue should be mixed louder than room tone, etc.

Mixing dialogue to average at -20dB will do a couple of things for you, mainly: 1) it will give you ample headroom for occasional louder-than-speech effects (like explosions), and 2) it will lower the apparent noise floor of your -12dB recording a bit.

Also, here's a pretty good introduction to EQ'ing the various elements of your mix. There are some good tips in here, and it will get you going in the right direction. If you EQ things properly, you can do things like pull up the volume of your music without it detracting from the intelligibility of the dialogue, etc.

[Roshdi Alkadri]

thanks for the reply jarrod. Im confused, we recorded tone at -20 and dialogue at -12 on location, so why do i peak dialogue now to -20?

[Jarrod Whaley]

Originally Posted by Roshdi Alkadri

thanks for the reply jarrod. Im confused, we recorded tone at -20 and dialogue at -12 on location, so why do i peak dialogue now to -20?

First of all, don't let your dialogue peak at -20dB--you want it to be there on average, but sporadic peaks can go quite a bit higher.

I explained my reasoning in my last post. -20dB brings down any noise or background sounds, and gives you more headroom for loud stuff in your mix. Why record at -12dB then? Because it's better to get a hot signal (that isn't clipping, of course) and mix it low than it is to record it low and then mix it louder. Make sense?

These are just my suggestions, of course.

[Roshdi Alkadri]

hmm, sound good of course. While i was posting this thread, i was also testing your theory, i found that when peaking dialogue at -20, you gotta turn up the speakers a lot. So generally, the meter pots are pretty low?

[Roshdi Alkadri]

alright, i got tone at -20, averaging dialogue at -20 but peaking at -12, music is under aat about -7. thanks jarrod

[Steve House]

The levels and range you should aim for depend on the intended destination of the show - conventional broadcast, HD or digital broadcast, DVD release, or theatrical release.

When balancing your mix by ear you need to establish some standard levels. You mentioned "you have to turn your speakers up" but adjusting your monitor levels and changing it as you mix is not a good idea - it leads to mixes that are all over the place. Get a cheap sound pressure meter like those from Radio Shack (~$45). In your NLE generate a track of tone recorded at -20dBFS and one of pink noise also recorded with an RMS average level of -20dBFS. One channel at a time, playback the tone and adjust the output level slider in your NLE so the NLE's output meter reads -20. Switch to the pink noise without changing any gain settings and adjust your speaker volume until it reads 82dBSPL on the sound pressure meter when held where your head would be at your mixing desk. Put a mark on the speaker volume so you can always return it to that exact spot.

[Jason Robinson]

This is all completely new stuff for me (fascinating to learn I might add). I've been recording audio in camera and just making sure I don't peak but trying to keep it around the "middle" though I must confess to not shooting for a specific number. And again I haven't calibrated my speakers either. I perform a complete review of the finished production and watch the audio meter and just try to make sure nothing hits above -1. So I am sure nothing peaks through out the project, but now this discussion is worrying me..... like I am missing something vital in my mix, but I don't know if I would recognize what it is. I'll be reading up on that link provided above. Thanks for the info.

[Roshdi Alkadri]

Originally Posted by Steve House

The levels and range you should aim for depend on the intended destination of the show - conventional broadcast, HD or digital broadcast, DVD release, or theatrical release.

When balancing your mix by ear you need to establish some standard levels. You mentioned "you have to turn your speakers up" but adjusting your monitor levels and changing it as you mix is not a good idea - it leads to mixes that are all over the place. Get a cheap sound pressure meter like those from Radio Shack (~$45). In your NLE generate a track of tone recorded at -20dBFS and one of pink noise also recorded with an RMS average level of -20dBFS. One channel at a time, playback the tone and adjust the output level slider in your NLE so the NLE's output meter reads -20. Switch to the pink noise without changing any gain settings and adjust your speaker volume until it reads 82dBSPL on the sound pressure meter when held where your head would be at your mixing desk. Put a mark on the speaker volume so you can always return it to that exact spot.

steve, thanks for the reply, i did read about this topic in a holman book.
what i was trying to say is that with the recorded dialogue level i had was -12 and tone reference was at -20....so now that im posting the audio to i just calibrate my DAW to the -20 ref tone and leave the dialogue peaking just like the recorded file of -12?

or should i mix dialogue at a different level than originally set?

thank you

[Steve House]

Originally Posted by Roshdi Alkadri

steve, thanks for the reply, i did read about this topic in a holman book.
what i was trying to say is that with the recorded dialogue level i had was -12 and tone reference was at -20....so now that im posting the audio to i just calibrate my DAW to the -20 ref tone and leave the dialogue peaking just like the recorded file of -12?

or should i mix dialogue at a different level than originally set?

thank you

Dialog peaks at -12 are the generally accepted levels. But note that it does depend on the final destination - DVDs can accept a wider dynamic range than would be acceptable for analog TV broadcast, for instance.

[Roshdi Alkadri]

fair enough, so eventhough my peaks are at -12, a whisper is gonna be lower than that, an occasional yell might be higher. What i was trying to get at, is the general level that dialogue is mastered at.

[Thomas Johansson]

In music production we call this setting up your monitoring at a K-20... that is after Bob Katz K-System. :)

K-20 is recommended for classical music and film work.

[Roshdi Alkadri]

got ya. thanks

[Wayne Nakamura]

Originally Posted by Steve House

The levels and range you should aim for depend on the intended destination of the show - conventional broadcast, HD or digital broadcast, DVD release, or theatrical release.

When balancing your mix by ear you need to establish some standard levels. You mentioned "you have to turn your speakers up" but adjusting your monitor levels and changing it as you mix is not a good idea - it leads to mixes that are all over the place. Get a cheap sound pressure meter like those from Radio Shack (~$45). In your NLE generate a track of tone recorded at -20dBFS and one of pink noise also recorded with an RMS average level of -20dBFS. One channel at a time, playback the tone and adjust the output level slider in your NLE so the NLE's output meter reads -20. Switch to the pink noise without changing any gain settings and adjust your speaker volume until it reads 82dBSPL on the sound pressure meter when held where your head would be at your mixing desk. Put a mark on the speaker volume so you can always return it to that exact spot.

Steve:

I followed your instructions and with Edirol MA 15D 15 watt speakers I can only get a reading of 78db. Is this still acceptable.

When editing the vol turned up full is way too loud. USing your method above, can i do most of my editing and audio tweaking at comfortable volume levels through the edirols knowing that there is additional volume just by adjusting the volume control on the speakers?

cheers

[Steve House]

Originally Posted by Wayne Nakamura

Steve:

I followed your instructions and with Edirol MA 15D 15 watt speakers I can only get a reading of 78db. Is this still acceptable.

When editing the vol turned up full is way too loud. USing your method above, can i do most of my editing and audio tweaking at comfortable volume levels through the edirols knowing that there is additional volume just by adjusting the volume control on the speakers?

cheers

Yes, do your EDITING at whatever playback levels work best for you. But do your MIXING of the final tracks at the calibrated playback level. Since your mixing by ear and setting relative levels between tracks then, you need to have consistency so the last part of the show sounds like the first part of the show.