During the mix, keeping levels right

Question

When you mix, do you have a certain db level that you try and keep all your levels at and then compress the peaks?

Are there certain levels for dialogue?

Does using the gain audiosuite plugin for intial level setting between different regions combined with a meters tool sound like the right idea? After that you would use vol. automation?

Is there a specific level for fx?

What level should you suggest to have the Dialogue Premix at??

I was also wondering, how much compression for dialogue you would say is TOO much. I'm under the impression that it is a subjective thing but after about 6-8 db compression would you say that it is too much?

…and finally…

I noticed the PAZ Meter doesn't recognize automation level changes so I'm guessing the only thing you can do is to put it on a master fader and is there a disadvantage to having the response set to a low value?

Answer 1

I mix for -24dB LKFS for all the television work I do, but I'm usually not looking at the meters to do that. If you're monitors are calibrated correctly, it shouldn't take you long to learn what the proper level sounds like. Mixing by ear is a lot more natural and usually turns out a better product.

I always start with the dialogue. Basically, my dialogue premix is very close to what it will be in the final mix. I know what it should sound like and get it there before I touch anything else. When I go to final, I mix everything else around it. I usually don't touch the levels for dialogue at that stage, but will make adjustments as needed to any compression and eq I'm using on the DX tracks.

I'll run my rough mix programs through the LM100 for reference after I've done the bulk of the work, and look for trouble spots that get outside of the +/- 2dB range from -24 LKFS. Then I'll go back and tweak if necessary, but I usually don't need to for levels. Most of the tweaks I make are for balance and subtlety, rather than overall levels.

In general, I avoid using the Gain plug-in unless something can't be brought to a reasonable level using just the typical volume automation. If the +12dB to -infinity dB within the mixer doesn't do enough, it's usually a good idea to consider replacing the audio to begin with.

Answer 2

Chris, just to add to what Shaun has already said, the 'Trim' plugin is your friend when it comes to pre-dynamics gain on the individual dialogue channels. It's like another volume fader working before your compressor \ de-esser \ whatever \ plugins.

Printing audiosuite 'Gain' will lock your regions in stone, and mean that you won't be able edit the handles later on (audio before \ after the region that is hiding in the wings) which is something that you may want to do if you change your mind about an edit.

Automating the Trim means that you can set your channel compression + de-essing to work pretty well with whatever you throw at it, and lets you regulate the level going in to them so that you don't need to constantly adjust their thresholds region by region. Of course you may only need to automate the Trim on occasion, depending on the consistency of the location record. To make audio from a lot of different locations work together, expect a bit more work.

Just one way of doing things

"I noticed the PAZ Meter doesn't recognize automation level changes so I'm guessing the only thing you can do is to put it on a master fader and is there a disadvantage to having the response set to a low value?".

Where is your PAZ in the chain? It's best to have meters sitting on their own Aux, and send to them from your main buss if you want to keep an eye on what is happening overall, post all automation and plugins. When it comes to levels, these are different for every network, go take a look for some delivery specs --- chances are you'll need to use a meter that gives you LEQ or LKFS levels, which are a way to measure perceived dialogue loudness (which is what Shaun mentioned in his post with the -24dB LKFS).

As far as "how much compression for dialogue you would say is TOO much" -- consider what you may be turning UP in the quiet patches. If you're hitting your dialogue with 8dB of compression then that's potentially 8dB of gain on the noise in between. No rules here really.... and there are techniques to minimise the quiet patches with downward expansion, but as a rule (hah..) you shouldn't need to slam your dialogue with "6-8dB" of compression to achieve control.

But it all depends on what you've got... and how long you have to mix!