Jump to content
vale

MIX and ISO Tracks levels

Recommended Posts

Hi all.

 

I recently had a long discussion with the Post-Production Sound team for a Netflix series I mixed on set.

They complained that my recording levels were too low.

Just to introduce you to our situation here in Italy, I have to say that - it may seem strange to you - it's not common to work with iso tracks pre-fader and mix track post-fader. The most part of sound mixers (?) like to work acting on the faders for the iso tracks and let the machine do the mix (it could be a blind sum of all the tracks or using an automix feature). I don't know why but everything almost always went this route. This maybe brought editors and sound editors to be used to work on pretty high leveled tracks.

I work with iso tracks pre-fader and use my faders for the MIX track (I like to do it mono) and always listen to the MIX.

So, usually I try to record the MIX track using the full scale, peaking even at -3dB or 0, having the iso tracks to be around 10-12dB lower.

 

I'm just curious to know which is your practice and how do you set your recorded track levels.

Thanks,

 

Vale.

Share this post


Link to post
Share on other sites

How low is low?  Being conservative is part of iso-tracking and post needs to be able to gain up if they need to and not be lazy about it.  Being SUPER low is not good, and means you need to work on your gain structure.

Share this post


Link to post
Share on other sites
1 hour ago, Philip Perkins said:

How low is low?  Being conservative is part of iso-tracking and post needs to be able to gain up if they need to and not be lazy about it.  Being SUPER low is not good, and means you need to work on your gain structure.

 

Indeed. I'm confident my gain structure is correct. 

In my case, they needed to rise up the gain of my iso tracks up to 12dB maximum (that is my max fader excursion on the channel strip to the Mix track), on the lower speech and just 5-6dB on normal dialog.

They're happy with the Mix track level, that is actually taking all the dynamic scale up to 0dB.

So I guess they simply don't want to rise up the gain on the tracks but having them already at the desired level.

 

Share this post


Link to post
Share on other sites

I’ve had a number of green posties come at me for this sort of thing, and often times it is someone pointing the finger so people don’t notice their own lack of experience. There often isn’t anything you can do about that. But for the benefit of the doubt, a short technical letter with references may be a good course of action. 

Share this post


Link to post
Share on other sites

First of all I will second what Phil said - too low can indeed be too low - not as much as signal to noise (though that can be also the case) but that the editorial workstations cannot really handle major gain shifts : less so now but significant in the last say 20 years.

 

Having said that, Vale, as a dialogue editor I would hope that my isos are lower and not at the distort level: I've worked with isos which are super quiet (too much so) but in extreme multi camera chaos - your own suggested levels would be ideal for me ... possibly even slightly hotter than I needed or expected ... but then I'm not cutting Netflix: was the criticism from the assistant editor or the dialogue editor/post sound dept?

 

Best, Jez Adamson

Share this post


Link to post
Share on other sites
6 hours ago, The Immoral Mr Teas said:

Having said that, Vale, as a dialogue editor I would hope that my isos are lower and not at the distort level: I've worked with isos which are super quiet (too much so) but in extreme multi camera chaos - your own suggested levels would be ideal for me ... possibly even slightly hotter than I needed or expected ... but then I'm not cutting Netflix: was the criticism from the assistant editor or the dialogue editor/post sound dept?

 

Best, Jez Adamson

 

Editor and his assistant are happy with my levels (they're editing the show just using my MIX track).

The problem came out on the dialogue editing side whe they start working with the iso tracks.

Share this post


Link to post
Share on other sites

Here it's common (for TV) to mix at 6 dBu line level, not depending if you're on cable or wireless to recorder/camera. It's always been that way since Nagra or so.

 

My ISO tracks are always lower, because Sound Devices mixers only provide 14 dB of headroom from the 6 dBu to 20 dBu, which is the right red end of the display (corresponding to 0dBfs). I guess it's better to have at least around 20 dB of headroom in ISOs, since you don't have much influence on their levels in case if someone's getting louder or loud things just happen.

So mix is more than 6 dB louder than the ISO tracks. Post got used to it, they prefer more headroom over working limiters.

Share this post


Link to post
Share on other sites

If your mix is peaking at 0 bBu (-20 DbFS), and your ISOs are -14dB below that, then it is too low.

 

I have had post friends call me to complain about mix and ISO levels on shows with some mixers being way too low, and they ask what I do. I always tell them they should contact the production mixer and ask for higher levels.

 

My mono mix is pretty hot (average conversational peaks at -12 dBFS), and knowing it's used for editorial, who don't want to be boosting gain for dailies, they and I seem to like it that way. But lately, knowing that I am on shows with extensive post sound schedules and budgets, I also keep the ISOs pretty hot too (peaks -20 to -12 dBFS). Recently I have been doing a lot of horror movies, so I let the limiters catch the screaming so the dialog isn't too low. I don't have enough hands to affect gain for multiple screamers. The show I am on now has three girls who run around screaming together a lot.

Share this post


Link to post
Share on other sites

It's possible that I am not writing well what I mean (my english is not too good).

 

My MIX track peaks, at its higher, even at -3dBFS or rarely even at 0dBFS - I use a limiter only on my mix track. For the same reasons you described (editorial doesn't like to boost).

So, my iso tracks are around 12dB lower. I.e. when the dialogue editor showed the session to me, he said he needed to rise up the gain 12dB to get what he consider the proper level. And he needed to gain up 12dB just on the quieter dialogues, but for normal dialogue he went just 5-6dB up, for screams sometimes the iso tracks were at the same levels if compared to the mix track.

Share this post


Link to post
Share on other sites

Sounds like an inexperienced post person...

It will be hard explaining that post production needs another take, when the director decides to only do one.

Share this post


Link to post
Share on other sites

Thanks for explaining that you meant dBFS. I was also scared when I though your were talking about 0 dB on a VU meter.

 

I'd still worry a bit about going to 0 dBFS even with a limiter. Look-ahead limiters, which are common in post, can catch even a single sample that tries to go over... but you can give them 10 ms or even 20 ms latency -- plenty of time for a gentle attack -- and the host software can compensate when you use it. 

 

General query: is anybody using a look-ahead with a significant delay in production? Does the recorder compensate so the track still agrees with timecode?

Share this post


Link to post
Share on other sites
3 hours ago, RPSharman said:

If your mix is peaking at 0 bBu (-20 DbFS), and your ISOs are -14dB below that, then it is too low.

 

I have had post friends call me to complain about mix and ISO levels on shows with some mixers being way too low, and they ask what I do. I always tell them they should contact the production mixer and ask for higher levels.

 

My mono mix is pretty hot (average conversational peaks at -12 dBFS), and knowing it's used for editorial, who don't want to be boosting gain for dailies, they and I seem to like it that way. But lately, knowing that I am on shows with extensive post sound schedules and budgets, I also keep the ISOs pretty hot too (peaks -20 to -12 dBFS). Recently I have been doing a lot of horror movies, so I let the limiters catch the screaming so the dialog isn't too low. I don't have enough hands to affect gain for multiple screamers. The show I am on now has three girls who run around screaming together a lot.

I do almost exactly as Robert...  -12 dBFS on my mix track, sometimes if I know it's tame dialog, a bit hotter...

-20 or a tad hotter on my ISOs...  If for whatever reason I think they may actually use the ISOs  and I'm not recording a true protection track, I'll go a bit hotter..

I use a limiter on my ISOs and my mix track..

Share this post


Link to post
Share on other sites

I'll quote myself first:

23 hours ago, The Immoral Mr Teas said:

 - your own suggested levels would be ideal for me ... possibly even slightly hotter than I needed or expected

 

Then afewmoreyears quoting RP:

5 hours ago, afewmoreyears said:

I do almost exactly as Robert...  -12 dBFS on my mix track, sometimes if I know it's tame dialog, a bit hotter...

-20 or a tad hotter on my ISOs...  If for whatever reason I think they may actually use the ISOs  and I'm not recording a true protection track, I'll go a bit hotter..

I use a limiter on my ISOs and my mix track..

 

Then finally your reply to my question:

17 hours ago, vale said:

 

Editor and his assistant are happy with my levels (they're editing the show just using my MIX track).

The problem came out on the dialogue editing side whe they start working with the iso tracks.

 

To me it sounds very very likely that this is not being 'properly' dialogue edited, but dialogue edited in an environment (with such folk) which is based on (probably cheap) general editing. Who maybe work only with super hot tracks for live reality or docu and this is their first bash or successful low quote on sound for a drama series. My guess only - but having to boost isos as hot as your's seem to be on that kind of scale (rather than the 'one in ten or so' we might expect) sounds suspect.

 

If it's really a problem, for the existing job or future work with the same production team, I might suggest looking at solving the 'problem' with the cooperation of the editorial dept, who are already happy, and if the problem indeed lies with sound post might be better suited to help correct the situation. If the problem is elsewhere they may be better placed to help sort 'what is expected'.

 

For me, at the end of the day, if sound is usable/clean (not distorted or obscured with noise) then it's good sound. But it might be just an unfortunate combination of cheap quotes, tight schedules, non-fitting workflows etc. So good luck - try editorial and try to be understanding to post's needs.

 

Best, Jez

Share this post


Link to post
Share on other sites

I try to keep my RMS level between -10dBU and 0dBU, which would be -30dBFS to -20dBFS on my 633. My peaks will be around -12dBFS most of the time, the limiters catch anything higher well enough without sounding bad at all. I haven't done proper post on my own tracks, but I've occasionally given them a listen and tried out how they'd deal with processing, sounds all good to my ear.

I don't often get asked for a mix track, but I'm starting to feel more like I should provide one anyway, I could/should keep my ISOs a bit lower in that case. I'm currently debating whether to use the line inputs or the mic inputs on the 633 for my wires. The line inputs don't have analogue limiters, which would mean that I should be a little bit more conservative with levels.

Share this post


Link to post
Share on other sites

Here in Germany most mixes for TV are done in Post with a loudness of -23lufs, for cinema release its even less, so no need to smash the sound to the limit while recording... compression/ limiting and distortion at the recorded files are not reversible. I prefer to have the limiter at my recorder acting only if absolutly needed.

The companies i am working for often dont want a mixtrack at all and work with the isos.

In Pro Tools and Avid MC its no problem to add gain in post, we have 2018, times of analog tape are over...

 

Share this post


Link to post
Share on other sites
4 hours ago, Ilari Sivil said:

I try to keep my RMS level between -10dBU and 0dBU, which would be -30dBFS to -20dBFS on my 633. My peaks will be around -12dBFS most of the time, the limiters catch anything higher well enough without sounding bad at all. I haven't done proper post on my own tracks, but I've occasionally given them a listen and tried out how they'd deal with processing, sounds all good to my ear.

I don't often get asked for a mix track, but I'm starting to feel more like I should provide one anyway, I could/should keep my ISOs a bit lower in that case. I'm currently debating whether to use the line inputs or the mic inputs on the 633 for my wires. The line inputs don't have analogue limiters, which would mean that I should be a little bit more conservative with levels.

 

Radio mics systems have their own limiting - if they are also line out then (and I need them on 4-6) I might set the gain on the TX higher and the line I/P trim low (633) to see where that leaves me with something loud enough to hit Rm limiter. Generally I push the faders over unity on all tracks, the adjustment is finer past 12 o'clock and it leaves safe ISO's.

Share this post


Link to post
Share on other sites

Speaking of dialogue levels at production stage, I tend to have it around -6 dBFS / -12 dBFS (not RMS).

Call me crazy, but I prefer PPM metering scale.

0 dBu = -18 dBFS seems low.

Share this post


Link to post
Share on other sites

Here's the deal (in my opinion, experience, and observation):

Prefader iso tracks, when done properly, will be, on average, at least 6dB lower. There are good reasons for this:

1) An important benefit of prefader iso tracks is to guard against surprise peaks that distort the postfader mix. To accomplish this, the iso track 0VU reference  is aligned with the mixer's prefader 0VU reference (typically -20dBfs). The input trim of a mixer normally falls into a position so that normal level dialog results in the fader knob being between "unity" (usually "0" on the fader's scale) and the top throw of the fader (usually +12dB), which makes the fader knob hover around +6 on the fader scale most of the time. These settings allow an additional 6dB of "reach" when the dialog level gets lower, and up to 14dB of gain reduction with the fader before the mix track goes past max, which would finally happen at exactly the same time the iso track reaches maximum. Some mixing boards like the Sonosax SX-8D and SX-ST have a +12 fader feature that essentially lowers the average iso track by even another 12dB. Fully utilizing the "never clip" feature of Zaxcom devices will lower the average iso track levels even more.

 

When the aid of compressor/limiters are factored into the post fader mix, even lower levels can result in the prefader iso tracks. The third factor is the reluctance of many sound mixers to record in the area between -20dBfs and 0dBfs, which can reduce the iso track levels by another 10dB-20dB, at which point the post production people of a very legitimate complaint.

 

Solutions:

1) Make sure the mixer's prefader levels are aligned evenly with the recorder's iso tracks (typically -20 to -20). Once that's done, the only adjustment that should be necessary is the input trim on the mixer and then riding gain on the fader for the mix.

2) Don't be afraid to use the mixer's and the recorder's dynamic range. In fact, be afraid to not use it. The audio quality does not degrade when going into the "yellow" and then degrade further when going into the "red" (technically, it actually improves). The perfect recording should have the highest peak reach 0dBfs.

3) Get familiar with how the use of compressor/limiters affect the relationship between the postfader mix and prefader iso tack levels.

 

Putting all this into practice will keep "The Calls" from postproduction to a minimum, and understanding the process will allow you to stand your ground and solve such issues when the calls do come in.

Share this post


Link to post
Share on other sites
2 hours ago, VAS said:

... Call me crazy, but I prefer PPM metering scale. ...

 

Vas, me too: it's what I grew up with mixing broadcast. But it is an analogue (and transmission) standard (12ms to 99% peak if I remember - it will be embarrassing but understandable if I'm wrong).

 

Still, it worked, and I agree it works within digital parameters: I am pretty sure that most digital spikes we have to deal with in dialogue post would not be there if a reasonable level PPM was the reference for recording: but is a digital PPM for mixers and recorders just a model on an analogue one, not the 'subtle real thing'?

 

Still, for me too, I use the 'ppm+peak' on my 302, ... And hang onto my Neve (in a box) and BBC (not plugged in) PPM meters for my editing setup!

 

Cheers, Jez

Share this post


Link to post
Share on other sites

Yes, peak meters are all that matters for recording original tracks. Still, the reference levels (typically -20dBfs for 24-bit digital original tracks) are there to align with 0 on a VU meter.

Share this post


Link to post
Share on other sites
2 hours ago, Glen Trew said:

The audio quality does not degrade when going into the "yellow" and then degrade further when going into the "red" (technically, it actually improves). The perfect recording should have the highest peak reach 0dBfs.

🎛️

 

Thanks @Glen Trew

Share this post


Link to post
Share on other sites

Low levels were a problem in the 16 Bit days. There was a reduction of higher frequencies when recorded too low and amplified afterwads. Remember the "Preemphasis" switch on DAT recorders to achieve a work-around?

In 24 Bit not a problem any more.

Share this post


Link to post
Share on other sites
8 hours ago, Glen Trew said:

Yes, peak meters are all that matters for recording original tracks. Still, the reference levels (typically -20dBfs for 24-bit digital original tracks) are there to align with 0 on a VU meter.

 

It depends on the meter ballistics actually. A PPM will give you the maximum values, so it would be safe to align them almost with -0 dBFS. 

 

However you must take the dynamics into account. If the program level increases and you are not using limiters you will have an overload. Or, even with limiters, you might hit them too hard.

 

With 24 bit recording systems there's no need to run everything so tight. With 16 bits its was undesirable to record at somewhat lower levels because soft signals were subject to quantization noise. Remember that the number of bits determines the resolution. 


At 24 bits it's safe to leave a big headroom, you will still have plenty of usable resolution. Leaving 8 bits of headroom at 24 bit will still give you full resolution 16 bit signals. Of course you can't enjoy that luxury recording at 16 bits, you would have only 8 bits which would really suck.

 

As a rough guide you can check this table of S/N ratios for different bit depths. 

 

https://en.wikipedia.org/wiki/Audio_bit_depth

 

 

 

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×