Sign In

DTS - Volume Loss

Author
Time

I’m having difficulty re-encoding a DTS audio track. Quick overview of process:

  • Rip the source m2ts file.
  • Extract the .dtshd file.
  • Demux the .dtshd file into 7 mono WAV files.
  • Edit the video/audio for the film in Adobe Premiere.
  • Export 7 mono WAV files.

The goal, obviously, is to do as little damage to the original DTS mix as possible. And up to this point everything seems to be working just fine. Audio quality is consistent throughout each stage of the process.

But I then re-encode the audio using the DTS-HD Master Audio Suite and the resulting file has a massive loss of gain/volume, to the point where the resulting file is virtually unusable for many people. (The audio is still all there: If I play it through my surround sound system I’m able to crank the volume up, with no apparent loss in quality, to the point where it’s audible again. But for many computer/speaker setups we’ve found that people literally can’t crank the volume high enough to have an acceptable experience.)

I’ve tried all manner of troubleshooting, but nothing seems to have any effect on this audio loss and I’m out of ideas. It feels as if I’m taking the same sound out and then putting the same sound back in, so I should get the same result in the final DTS file. But that’s not happening for some reason. Really hoping someone with expertise will be able to tell me what I’m doing wrong.

My default settings for the DTS encode:

  • DTS-HD Master Audio
  • 6.1 Discrete
  • Sample Rate: 48000 Hz
  • Dialog Normalization: -31 dBFS (No Attenuation)
  • Core Bit Rate: 1509 kbps
  • TC Frame Rate: 23.796
Author
Time

This problem was diagnosed through multiple steps over at fanedit.org: https://forums.fanedit.org/showthread.php?tid=16140

It turned out the the WAV files produced by Premiere were resulting in a broken DTS file, despite the fact that (a) nothing could detect anything wrong with the WAV files and (b) DTS’ own tools verified the dtshd file. (It was just every other decoder that identified the DTS file as broken.) Very weird.

For the reference of anyone finding this thread in the future having the same problem, here’s the “solution”:

  • Creating an Audacity project file with all the tracks.
  • Doing a Gain adjustment across all tracks.
  • Exporting as new WAV files.
  • Encoding a DTS file using the new WAV files.

The resulting DTS file has no volume/gain loss and is no longer “broken”. My best guess is that making the Gain adjustment forced Audacity to re-encode the file in a way that (a) conversion didn’t and (b) eliminated whatever element was causing the DTS suite to glitch out. But since there’s no clear reason why Premiere would be exporting a buggy WAV file, none of the tools I’ve used have been able to detect anything wrong with the WAV file, and DTS’ own tools don’t detect anything wrong with the dtshd files that all other decodes consider broken, that’s just a wlid guess at this point.

But if it works, it works.

Author
Time
 (Edited)

I don’t have experience with DTSHD encoding, because the suite never did anything but freeze up when I tried to use it.

But I do have experience with Premiere exporting things with massive gain reduction for whatever reason. The issue, as far as I could tell, seemed to always be with what it was importing. Even though they were perfectly fine wav/mp3/ac3 files, its import function seems to be bugged. I fixed it by simply taking whatever source I was using, importing it into audacity, exporting again as wav, and then bringing into Premiere. I could then compare it directly to the previous, bugged import and see directly that the peaks were higher. And that carried over into what I exported from Premiere, now correct.

I don’t know if this is the same problem that you’re having with DTSHD. I assume that you know how to export the lossless audio from it and not just the lossy core (I’m not sure that I know how to do that). You say ‘demux’, but I don’t know if that means losslessly re-encode in this context. In any case, you could try re-encoding the wav audio and importing that into Premiere. Perhaps you will luck out.

Edit: Reading over in the other thread, it seems even less likely that this is the same problem, as you don’t actually notice the drop when looking at the directly exported Premiere audio. It only shows up after the next step.

Author
Time

What all is involved with doing a gain adjustment in Audacity? Never having done it before I’m not sure if it’s something automatic or if it needs to be fine tuned.

Author
Time

Basically automatic.

  1. Add all the tracks to Audacity. (Drag and drop.)
  2. Make sure you select all tracks.
  3. Go to Effect -> Amplify.
  4. Define “New Peak Amplitude” as -0.1. This will automatically set the “Amplification” value to the maximum value possible without introducing clipping.
  5. Click “OK”.

You want to make sure you select all of the tracks and Amplify them at the SAME TIME. Audacity will look at all the tracks, calculate the maximum amplification that can be applied to ANY track without introducing clipping, and then apply that amplification to ALL the tracks. (And you want the amplification to be applied equally to all tracks, because otherwise you’ll lose the original sound balance. And your LFE channel will be totally insane.)

To export the WAVs:

  1. Go to File -> Export
  2. Select “Export Multiple”
  3. Set “Header” = WAV
  4. Set “Encoding” = Signed 24-bit PCM
  5. Set the export folder to whatever you want.
  6. Click “Export”.

And that’ll give you the WAV files you need.

Author
Time

Thank you so much for the instructions. You’re a gentleman and a scholar. 😃