Foley levels and Loudness track question

Nuendo
1

Hi everyone,

i hope it is ok to post non Nuendo-specific questions here ?
As i’m not too experienced in film mixing, i would be happy about some advice on handling levels for foley. My problem is that i’m always falling into the trap of reducing the levels too much, both for clips and in the mixer (most of my library sounds require that), that i end up with too low and/or inhomogeneous volumes in the mix (i.e. a big mess of different clip normalizations, clip levels and automation :confused: ) .
What is your approach of working with them regarding levels ? Is there some kind of approved level to normailze foley to (and/or LUFS target level ?) - do you use clip volume at all ?
To make it at least a bit Nuendo specific - i’m new to the Loudness track as well - is there a way to change it’s colour ? My eyes are probably too old for that blue :stuck_out_tongue:
Thanks in advance !

2

Well I mix foley as I mix anything else.
I never normalise anything, ever.

I realise most new to film sound may not have the financial resources to use a actual foley artist.
If you can’t have that the person creating and editing foley needs to think like a foley artist.
Think flow and sound in relation to the picture.

I use clip levels, faders, group fader, eq, reverbs and delays when I mix foley. The best foley artists really “mould” the foley in relation to what the shot needs, and they do it intuitively using the room, position and proximity to the mic and how they perform to assist the illusion of movement and distance.

Doing this with samples is not easy, you really have to work hard in foley editing and if left to the mix, you will need a lot of time to do a decent foley premix.
Make sure to use levels and Basic eq as a part of the workflow when creating and editing or the mix will be super slow.

No film mixer I know of uses loudness or meters to deal with foley mixing. Foley is mixed in relation to dialog. That is it.

Edit: foley is not just footsteps (feet is really the only part of foley that is “easier” to do using samples) moves and props will always have to be recreated as actual foley, if you don’t you will have a hard time making the feet work.

Make sure you keep and use pfx from production as much as possible as it will help to “glue” the fx/foley/pfx together.

3

Hello Erik, thanks a lot for the insight ! “I never normalise anything, ever.” - that’s very interesting. Is there a reason for this ? (i never do it in music production either, but thought it was very common for film, somehow…web knowledge :wink: )

About foley - actually i do record most of the sounds myself (and enjoy the creative side of it) and have also realized how important all the subtle sounds like clothes, zippers etc. are while characters move, also regarding the distances, to make it sound organic and realistic.
The relative levels are also fine, so this is rather not an issue. I’m actually very happy with the results so far (and so is the client), i’ve done a few episodes already in the past.

My problem is that i usually get the dialog only at the end of the production (of course only a couple days before the deadline), so all foley, atmo, music and sfx come in first…and i used to mix / automate levels as i go, which later often turns out as too low, forcing me do go back and adjust everything (hours or days of work) and/or crank up the limiter input (thank god for having low noise preamps).

My thinking was, as i’m coming from music mixing/production, that my usual approach doesn’t work here, and that an ideal workflow might be different - like recording and placing all the necessary sounds, set them at basic and constant levels, then do the mix and automation at a final, separate stage. But this is not how it’s done ?
I’ve also set up some kind of master VCA’s for all groups/categories - is it maybe a good idea to set them at like -10db right from the beginning to leave me some room to push upwards later ?

Make sure you keep and use pfx from production as much as possible as it will help to “glue” the fx/foley/pfx together.

Had to google what pfx means :slight_smile: - there’s no original sound in this case, it’s an animation, but still good to know.

Somewhere i read that it’s good advice to normalize the dialogue to -22 LUVS…is this at least a common level to start with/relate to ?

4

Perhaps you can sort of analyse the levels of dialog in your earlier mixes. No LUFS needed, just ordinary digital level. Maybe dialog has been between - 10db and -15db. And take that as a start.
No normalisation. Its rather the contrary what we try to achieve: A interesting range between low and high levels.

Oswald

5

Thanks Oswald.

No normalisation. Its rather the contrary what we try to achieve: A interesting range between low and high levels.

I see. I assumed that this dynamic is done by automation/mixing, but obviously it’s just more complex than starting from a certain level (and that i mabye have done it intuitively right before already…probably overthinking all of it).
The levels in my earlier mixes are in the same way “problematic” - i have them all open in one session. I wish i could start from scratch, but i need lots of the previous sounds.

Maybe dialog has been between - 10db and -15db.

Yep, it was in that range (after fixing) in the earlier mixes.
Back then i also compared levels to an imported audio of a big animation movie. I’m not sure how reliable this approach is, though.

By the way, ich habe Deine Seite besuchen wollen, ein Klick auf das Bild(Link?) führte aber nicht weiter…vielleicht ein Javascript Problem oder ist es nur eine Landing Page ?
Regards/Grüße
Christian

6

What dialog are you getting while you’re working? Just raw audio or edited but unmixed?

I mean the most ‘obvious’ solution is to simply mix your Foley louder than you think it should be if you often have to lower it when you get final dialog. The other solution is to measure the average loudness of the dialog you typically get in the end and then just do a quick and dirty leveling of your temp dialog so it’s easier to find appropriate balance of Foley.

My hunch though is that you’re going to have to deal with it this way no matter what, more or less. So just adapt to it as far as scheduling goes.

I agree 100% with Erik btw.

7

Mattias, thanks for the hints !
The dialog i get is edited/cut (whole stems), just needs some processing and levelling.
The problem is however not so much the relative levels. My main question was if there’s a good procedure to handle volumes (on clip level) right from the start, so that one can’t get into trouble at any point later. So far it seems there is none…just work with what you got, as i understood (i.e. no normalization of any kind).

So, setting the master VCA’s 10db down to spare some room is not a common thing i suppose ? (in theory, or in my case, it could work well, i may just try it next time)

8

Well I mean either you have a problem that’s even throughout the entire content, or it varies over time. If it’s the same throughout then you should be able to just do as you say using the VCAs, or you can just increase all clip gain. Same difference. If you want to ride levels again to tweak then sure, VCAs can be used. But I think Nuendo faders have at least 12dB gain so you can probably just leave them at “unity” and gain up from there rather than default them at -10dB. I mean, I don’t really see the difference there.

9

Actually i was seeking some general advice for future projects. With this one it’s too late…volumes vary and it has become too complex (for me), over a hundred of tracks, and i’m afraid to mess up everything. I will try the safe route with VCA level reduction on the next one.

But I think Nuendo faders have at least 12dB gain so you can probably just leave them at “unity” and gain up from there

The thing is - i need to add more than 12 db :unamused: :wink:

Ok, i think this subject is clarified, thanks a lot to everyone again.

10

Well ill add a few questions and some more feedback:

Dialog is cut (whole stems) what does that mean?
Would it be possible to adjust schedule/workflow?

I always have final dialog while mixing, otherwise it makes no sense. Why are you getting it late? In all our productions I prioritise dialog in the schedule so if at all possible dialog will be delivered to the mixer first. If not possible before the mix starts we will work in reels (feature films) or just divide the material up in useful time chunks, and assemble in the final mix.

Any production that relies heavily on dialog (most of them) the dialog is king, you mix the dialog and then set the levels of everything else in relation to the dialog.
I would never even want to try mixing a dialog film any other way. This doesn’t mean that I don’t set levels on foley fx and ambiance’s ahead of the mix. The sound editors will of course have a temp DX to work against, the editor needs to know what level that would reasonably play at (it won’t be perfect but will give a good basis for setting all other levels).

Yes there are times when I choose to mix fx ahead of dialog for creative reasons. But I think it best to avoid that workflow until you are comfortable mixing around dialog first. You will learn your room and monitoring and start to understand how you can use metering to get in the right ballpark level wise.
But there are no answers to “what level should foley/dx/fx/music play at” it doesn’t work that way.

What peak level a signal has is seriously affected whether the signal is raw or has been dynamically controlled. Thus any peak level is pointless unless you know and understand what dynamic the program has and how you choose to mix it.

Good luck on your future endeavours!

11

Erik,

maybe it would help if you explained what (or “if”) you pre-mix out of all the fx you have, prior to the final re-recording of all content.

12

Erik and Mattias,
really appreciate that you’re trying to help, thanks again.
Why do i get the dialog late - well, the animation studio is also new to this and not too familiar with established workflows (like me in that matter). Of course i tried to explain that i can only really start working if i had the dialog first, but this discussion was as fruitless as my requests for a timecode in the picture (do it myself now). At least i do get picture locked video.

“Dialog is cut (whole stems) what does that mean?”
It means i get each voice track as one file, ranging over the whole movie. There’s probably a term for this that i don’t know.
So i don’t have to make any edits there.

The sound editors will of course have a temp DX to work against, the editor needs to know what level that would reasonably play at (it won’t be perfect but will give a good basis for setting all other levels).

This is beyond my understanding…what would a sound editor do ? What is a temp DX ? :slight_smile:

In this project i’m doing all of the sound, mixing and music on my own. My approach until now is that only the result is what counts, no matter how i get there. But, since i might get more jobs of this kind, i’d like to get more into the subject and a bit closer to how it’s done professionally (i.e. fast, efficient). After all, the best would be an internship in a studio probably, but currently thats not possible.

Some more questions if i may:
What advantage would it have to premix things out ?
Do you know any good in depth video tutorial on film mixing/post production ?

13

Here’s my workflow when working on all steps of a project, meaning just one post engineer working throughout all the post sessions.
As said before, dialogue is king, so I start by editing dialogue and use clip gain to get the levels the closest I can to 0dbfs, there’s a minus 2 to 4 dbfs headroom that I tend to leave open to later dynamics and restoration, it all depends on the quality of the material and production sound that is sent to me.
Then I start building the sound design, again, only using clip gain and dynamics, or fx, but all with the dialogue as the center piece.
In the end, the score editing with attenuation to at least -4 dbfs, I tend to normalize this step to -4 or -6 because in this step the material has already been mastered and has way more dynamics than the dialogue.
Now I’m ready for mixing and jump on the mixing console using automation on the sound design and score, I never ride dialogue!!! or use the faders until this stage.
Now the “mastering part”, where I use the stereo out or whatever delivery mix I need, to insert dynamics and achieve unity gain.
Now you’re ready to ask your client for the loudness standards, either you reduce the level of the entire mix on the stereo out or let Nuendo do it for you.
I tend to deliver for TV with ebuR128 standards, meaning -23lufs. When I set the stereo out to -10dbfs, I only tend to miss by -1 or + 1 lufs when analyzing the whole mix, meaning something must be right with my workflow, then I’ll correct that minor loudness discrepancy or let Nuendo do it for me.
The reason I edit and mix dialogue close to unity gain is because I have had sessions with over 4 delivery standards, that way I have headroom for all of them just by attenuating the whole mix on the stereo out or surround groups.

Resuming:

-Start by editing and levelling dialogue/ VO ( in this stage I give 1/2 db plus to VO regarding production dialogue), only using clip gain.

-move on to Sound design, working within your dialogue range, again, only using clip gain

-move on to music score, normalize those masters to -4 or -6 dbfs, then Edit. ( If you’re also scoring, mix and master those scores as if they were original pieces, even if you’re scoring to video, then move them to the Post stage).

-move on to the mixing console and start mixing the sound design and the music score with the dialogue, never touch the dialogue levels.

-move on to the mastering stage where you insert (?) another compression or limiter to achieve 0dbfs

-your mix is ready, ask for delivery specs, unless its cinema ( there’s no loudness standard), you should go to the stereo out and enter -10 dbfs or let that stereo out fader untouched and choose loudness settings on Nuendo export menu.

You’re used to music production, audio post doesn’t work that way, you only touch a fader in the mixing stage, when all is already edited, and I highly recommend getting a control surface for that.

And of course, sends and vcas are part of the editing/mixing stage, but the workflow is the same.

DX= Dialogues.

Hope this short (?!) answer helps.

Cheers.

14

Hi Miguel,
thank you so much for the (way too short :slight_smile: ) insight ! It indeed helped a lot understanding of the whole process.
That you’re only using clip gain in the first stage is some important info…i guess my mistake was to set fader levels right at the beginning (using a mix of the two). Sounds logical to have them at somewhat homogeneous gain before touching any fader.

then I’ll correct that minor loudness discrepancy or let Nuendo do it for me.

Let Nuendo do it - can you explain that further please ? (And can i tell Nuendo to do the whole mix for me somewhere ? :wink: )


choose loudness settings on Nuendo export menu.

Just checked, i can’t find any setings regarding the loudness there, can you explain that too, please ?

Another thing is room calibration - from what i’ve read, working at a fixed 85dB SPL is obligatory ? (never done that either for example)

About the master limiter - are there any plugins or hardware you can recommend for that ? I’m currently using the Flux/Jünger Level Magic (does something very nice to transients) and a hardware NTP 179 400 limiter (never clipping, adds some body, set before the Level Magic)…although i’m quite happy with this combo, i have no real comparison.

Also, speaking of mastering, i use to insert an EQ on the main groups except DX (especially on music) that takes out a few db in the vocal range for better clarity of the dialog…is this something common to do ? After all it’s probably a matter of taste, but i’m curious.
If anyone wants to share their favourite go-to EQ’s i’d also be interested to hear. I know there’s a gazillion of choices…just recently i discovered the Kilohearts Slice Eq and it’s such a breeze for quick DOP for example.

I highly recommend getting a control surface for that.

I would very much want to use one…does there any large enough exist for a budget of under 2K Euro ? I was thinking of getting a modified Mackie d8b, but these things are so old and probably unreliable. Although, for automation rides i’m still using an even older Radikal SAC 2, eight channels…yippieyay :slight_smile:
Sorry for bothering with even more questions and thanks again :slight_smile:

15

Hi Chris.

Sorry for the delay, but the worldwide pandemic got in the way.
Im going to post some tutorials online regarding basic approaches to Audio Post Production, the Production content is low and got nothing else to do.
Ill post the links here, its gonna be on YouTube.
If you think its gonna help, check it out.

M.D.

16

I think you really only have an obligatory standard if you’re mixing to one, for example Dolby. So if you’re doing that then you’ll work in a room that’s been calibrated a certain way. For other facilities I believe the idea is to calibrate the setup so things play back at an appropriate level. In a small room you’re probably not getting 85dB SPL but less than that.

I’m not sure that makes sense to me. I would work on dialog and make that sound good, then I would work on other stuff like sfx and music. Now, for TV I’ll mostly leave music alone but if I need to EQ the fx then I will. But I’ll do that on the individual tracks that feed my groups. So if I have an EQ on the group and it’s dipping for dialog then when I play back my effects I’ll hear them through that EQ. And if I want to adjust the way the fx sound based on what I hear, and I’m using an EQ on the track, then there’s a chance I’m basically undoing what the other EQ did, or at least doing more or less work than I have to.

So to me it doesn’t really make sense.

17

Hi Miguel,
no worries, now it’s me being late :wink:
Creating tutorials sounds great, i’ll be excitedly watching out for them.


@Mattias

Thanks for sorting that out about the levels. Meanwhile i’m accustomed to working on a specific level (not measured) i can always jump back to…i think it’s only somewhat important/better to constantly work on the same level, no matter what it is exactly.

About EQ’ing the music to fit the dialog - i forgot to mention that i use to do this on projects where i only deliver sounds and music, and the voice is recorded and mixed elsewhere. It’s a blind shot so to say, hoping the voice will blend better, assuming the video editor doesn’t care much for it or doesn’t have the equipment to judge sound really well.
If i got the voice tracks as well, i’ll of course mix it all the way i think it sounds good, very much the way as you described it.

18

Hi, Your level will be high if you try doing complete focus. The loudness track is really so good.

19

There’s a tick box ‘Normalize to Integrated Loudness’ under File Type Settings.