Start a new topic

LFE / subwoofer limiting applied to audio during offline re-render

 Hi there,

I noticed that when I imported an audio file created from an offline re-rendered master for comparison to the source material, that a limiter had been applied to the LFE / subwoofer with a threshold of -10dB.

When I playback the master from within the DAPS itself, there is no limiting, but it is definitely applied in the re-render; both in 7.1 & 5.1.

I have tried with spatial encoding switched on and off, and also with the subwoofer limiter option switched on and off (although the manual states that this only affects monitoring, not the re-renders), and nothing seems to make a difference.

I am working on a project remotely and need to send demo's of my work to supervisors, producers and most importantly the director, and a 10dB slice out of my LFE is an awful lot!

I am not sure if this is an intended "feature" or a bug, but either way it's a big problem for my workflow. Especially concerning is that there seems to be nothing written in the documenation about the processing that's applied to the LFE in the re-rendering process.

Help would be greatly appreciated

Sorry - for clarification software/setup used as follows:

DAPS: 3.3.0_5195857 via Dolby Audio Bridge

Pro Tools Ultimate Native 2019.12.0

Monitoring format: 7.1.4

Hi James, in v3.3 there is always a limiter applied in the LFE channel of re-renders with a threshold of -10db. The Subwoofer limiter switch in preferences allows you to either monitor that limiter in the speakers or not. The listening setup of your producers and director will not be expecting an LFE signal above -10db. Depending on their setup sending them a re-render with an LFE signal over -10db may result in distortion.

Thank you for your reply Fin - much appreciated. Is it good to know that this is not a bug at least!

However, my exported material will be imported into another DAW for playback - with all speakers calibrated correctly (as mine are). Their listening setup, as you mention, is in fact designed for an LFE signal that goes all the way 0 - but it is not an Atmos enabled room (7.1 only), hence the need to send a mix in a ‘generic’ format. Unfortunately now what they hear from me will not be as it was intended to play back.

I can’t help but wonder what the usage scenario for the re-renders is, if the LFE has 10dB of limiting on it by design? For what purpose is 5.1 or 7.1 re-render supposed to serve in this scenario?

It would make much more sense to instead attenuate the entire LFE signal by 10dB, so that in this usage scenario 10dB of gain could be added to the LFE on playback, and then the mix will play as expected. If 10dB of level is added to the limited signal on playback, any signal below the threshold will of course playback louder than intended - I’m not sure which is worse; loud sounds playing low, or low sounds playing loud!

It would be even better of course to make this a user selectable option in re-render dialogue box. You say that “depending on their setup sending them a re-render with an LFE signal over -10db may result in distortion” - that may of course be true, but surely this should be a setting that the user selects when creating the re-renders? The user will have a better idea of the intended purpose (and destined playback setup) than generically making an assumption “depending on their setup”. You exclude as many users as you protect with this ‘feature’.

It is also incredible that there is no documentation about this in the manual. I shudder to think how many people have exported re-rendered mixes without checking them, and as a result material has been delivered at a level other than what was intended.

For now, the most straightforward work around for me is to attenuate the level I send to the DAPS by 10dB when recording, so that the limiter threshold is never reached, and then 10dB of gain can be added back again on playback - but this is not an elegant solution for me or the playback operator. I could alternatively re-route all my objects to the bed temporarily and bounce inside of Pro Tools, and then mix the material down to 7.1 using a third-party plugin like Spanner, but again this is hardly practical.

I can’t emphasise strongly enough the importance of giving the user the option to toggle this on or off in future versions

Hi James, Yes I totally agree. I believe the initial reasoning behind it all was, as you deduce, that consumer equipment was set to boost the LFE by 10db by default, therefore everything ends up back at unity in the home. I will chase this up further, but I can also see it going away in coming updates to the renderer.

I've got my fingers crossed ;)

Just to note, the mix room and the playback room should ideally both be calibrated so the LFE amp is +10db. This way the chances of hitting the limiter are greatly reduced. Whilst the limiter was introduced as a "safety net", I agree it should be removed completely. I'm confident it will not be in future versions of the renderer.

Login to post a comment