Setting proper mix and levels

[Rick R]

The important thing is knowing how your mixes sound on other systems, and once you know, you compensate. In my sound studio, I know a mix can easily become bass prominent if I don't just pull it back a tad in the mix. Your ears can cope with all this, with practice.

- This is why us audio pros have multiple sets of speakers, from very good to awful, including having a sub for checking very low frequency content that few will hear but can trigger other abnormalities downstream. If it sounds 'decent' on all the monitors, it's good to go out to the world. Personally, I prefer doing most of the relative level balances (dialog, music, S/FX) on 1970s Auratones which works 'for me' and translates well to other systems, especially the all important mid-range.

 

[J.Brown]

For all of my mixes I set a limit at -6db and just use my ears mixing levels on low sound listening level. When there’s a good balance I check on higher levels so that I didn’t miss anything odd. Voices should be easy to hear and music or effects should not compete with voices.

 

[Noiz2]

For all of my mixes I set a limit at -6db and just use my ears mixing levels on low sound listening level. When there’s a good balance I check on higher levels so that I didn’t miss anything odd. Voices should be easy to hear and music or effects should not compete with voices.

With the exception that my limiters are generally set at -3 that is what I do also. When the client want's the peaks at 0 or "as close as possible" (which happens way more than it should but...) I do a gain adjustment on the final mix to boost it up.

---

I shold add that my goal is to never hit the limiters, they are just a safety net.

I will sometimes, though not that often, use a multy band compressor on dialog. The bands are level based. Unfortunately the plugins got bought by another company and that company went bust so that plugin is not available anymore and I don't know of any others that work that way so if I ever upgrade to 21 century software that won't be in the mix anymore.

 

[Rick R]

I use limiters as 'a safety net as well'. In music mixes, I 'll use them for an "effect" and for level control. I also use multi-band compressors frequently in post. My go-to favorites are from Sound Forge Pro's iZotope mastering suite and the Waves C4.

 

[David Evans]

Thank you all for your help.

I have only one final question:

When you're using the NLE audio meters (in davinci resolve's case, they are full scale) and you're mixing everything to target -20dbs average, aren't you deciding loudness as well? I'm no sound engineer so please take my question with a grain of salt, but i assume that -20dbs will also translate to a certain spl measure, correct? Of course, loudness can be adjusted but if we mix dialogue to -15dbs and everything else around it, shouldn't the film be automatically adjusted for loudness levels as well?

Edit:
I'm assuming most people mix dialogues in the same range (-8 to -20). Maybe I'm mixing up concepts, so please correct me if I'm wrong.

 

[Rick R]

I'm assuming most people mix dialogues in the same range (-8 to -20). Maybe I'm mixing up concepts, so please correct me if I'm wrong.

The monitoring volume and mix levels are different parameters. Monitoring decibels are SPL (sound pressure level). Most everywhere these days, loudness is the measurement standard, typically LU/LKFS (integrated). Your final mix levels depends on where it's going e.g, broadcast TV, CD/DVD,BU, web streaming.
That said, prior to the 'CALM act', using full scale meters, audio submission spec for the major US networks was; a reference level (1kHz test tone) @ -20dBFS with program peaks not exceeding -10dBFS.
FWIW, w/o LU/LK meters , for usual web streaming, DVDs and such, the general consensuses is, program peaks around -4 to -6.0dBFS... but as long as it doesn't clip, (+.0dBFS) your ok. Broadcast submissions OTOH, are usually strict, if a program not within the network's technical parameters, rejection is likely.

 

[David Evans]

Thanks Rick. My film is for theatrical release (film festivals), so I've mixed it following the "film rules", with dialogues peaking at -10db and usually sitting around -15dbs and everything else around it. Average DB reading should be around -20dbs.

My thinking is: If I'm mixing my film with an average of -20dbs, with dialogues at -15dbs, and if most filmmakers are following this kind of rule of thumb (which I think are usually the numbers I see people mixing for film), then my film should have an expected loudness equal to most films out there. Is this more or less correct?

Thanks