Skip to main content
Jared_T
Jared_T
Participant
April 9, 2020
Question

Mixdown levels much different than Multitrack levels

  • April 9, 2020
  • 2 replies
  • 970 views

Hey guys, I need help!
I've been mixing a 2 host podcast with occasional guest. Before I begin, I've been matching clip loudness to -16LUFS, and then adding all my plugin effects until I get the desired levels and sound. I'm running FabFilter plugins, and the order of my chain is EQ, Compressor, DeEsser, and then Limiter. On my master bus, I'm also running a set of stacked Compressors and a Limiter as a final safety.

As I play and edit this show on my multitrack timeline, all the levels and tones sound great. I'm obviously not clipping anywhere because each track is running solid compression and limiting between -1 & -2 dBTP, as well as my master limiter at -1dBTP. 

The trouble is, when I export this multitrack mixdown either as .mp3 or .wav, the exported track levels in some areas are drastically different, and some are grossly clipping. I've tried adjusting the "default panning mode" in the multitrack preferences to no real avail. I've tried exporting in mono and stereo with no help either.

Why on earth does my export sound so much different?! It seems ridiculous to me that I can't mix my tracks in my multitrack editor and not get that same result in my export. What is happening?!

This topic has been closed for replies.

2 replies

J
Justin23024650qui7
Participant
February 11, 2022

Ugh, I'm having the same issue and can't find a solution!   Any luck on your end?   I keep having to export to mp3 to know what the levels actually sound like, which really, really isn't ideal 😕😕

    SteveG_AudioMasters_
    Community Expert
    SteveG_AudioMasters_Community Expert
    Community Expert
    February 11, 2022

    I'm not aware of any movement on this - and it certainly doesn't happen to everybody. If you have a lot of effects on a track - especially if they are processor-intensive ones - then one thing you could try is to pre-render the track so that those effects are processed before the final mix takes place. Hit the lightning bolt... (yes it's undoable).

    E
    emmrecs
    Legend
    April 10, 2020

    Yes, I have seen this also, albeit with only one project of a much smaller size than yours.

     

    Like you I had set up the needed FX for each clip, balanced the audio levels etc. and then exported the project.  One clip's level had increased to the point of distortion, almost as if the FX, including iZotope Ozone 9, had operated "twice", for want of a better way of describing it.  I checked the clip settings in the session, no problems.

     

    I then did a second export - no changes made to the multitrack session - and the export was fine!  No changes made but now everything was as it should have been!  I couldn't replicate the problem (any subsequent exports, made to "test" whether the problem could be repeated, they couldn't). 

     

    So, I'm very confused!   And I fully understand your frustration.  Did you try a second export?  With what results?

     

    Jeff

    SteveG_AudioMasters_
    Community Expert
    SteveG_AudioMasters_Community Expert
    Community Expert
    April 10, 2020

    No, on the face of it that shouldn't happen - certainly in emmrecs' situation; you shouldn't get a different result from two mixdowns of the same thing - that does need investigating further. But it is absolutely the case (whether it should be or not) that the more processing you put on a track, the more chance there is that something will go wrong...

     

    As far as the OP is concerned, I'd far rather normalise clips to the EBU R 128 recommendation of -23 LUFS and put a limiter on for emergencies, and not bother with an awful lot of processing, which almost inevitably makes everything sound worse, not better - ask any broadcaster!

     

    Also, Audition only displays dBTP values if you look at a file's amplitude statistics - none of the effects use that value for their operational calculations - they are all normalized to dBFS. If they used dBTP they'd take longer to run, as it takes rather more computation to establish them, and there's no real value to it; if you follow the EBU guidelines you'll end up with program content that falls within the accepted gamut without any recourse to dBTP at all.

    Terms & ConditionsAccessibility statement
    Using the community Terms of use Privacy Cookie preferences Do not sell or share my personal information ad choicesAdChoices