How to decode 6-track APTX-100 (cinema DTS) with the correct channel levels

[Turisu]

Also, I don't remember any clipping when doing my own syncing and adjusting levels on any of LOTR movies. Odd.

Odd indeed. Here's ROTK with levels untouched:



If I boost the LFE channel by +6dB those peaks are going to be 2dB louder than all of the other channels.

EDIT: I opened the track in Audacity as well to make sure it's not Audition screwing things up. Same levels.

[zoidberg]

I know memory is fickle but I distinctly remember all 3 LOTR films had extremely aggressive bass in cinemas and then there was the original DVD of FotR which had the 'hot' LFE controversy. 

Technically the LFE channel is the 'loudest' of the channels in terms of pure sound energy/dB but due to the limited bandwidth the effect is felt, not heard. The main speaker channels will always sound louder and in the case of poor amplification painfully so. 

I was doing some research into the LFE levels during the lockdown of last year along with little-endian which kind of fizzled out, I do think there is more work to be done in order to be able to say for certain the definite values needed for adjustment, for example a rip of one of the calibration/test CD-ROMs along with a RTA to analyse the sound on playback. But the research done by others here and on OT does point to +3dB and +6dB for pre/post SMPTE RP200. These adjustments put the LFE channel into the correct relative level so that when played back on home systems (with the in-built +10dB gain applied) the levels are correct.

Just out of curiosity exactly how much gain is needed to bring the LFE just below the point of clipping? Is something along the lines of 5.25dB?

[Turisu]

ROTK has a lot of loud LFE compared to other theatrical DTS mixes I've seen. FOTR is tamer by comparison and can withstand a +6dB boost without clipping.

I can boost ROTK around +4dB without clipping. But no more than that.

[zoidberg]

Interesting, in theory that LFE level is higher than what is capable with AC-3 or SDDS, I wonder if their LFE is clipped or if it's a mastering flaw with the DTS track

[Turisu]

I think it's as you said; that we don't yet have the full picture of what adjustments need to be applied to which movies and more research is needed. I also think that home video mixes where all the main 5 channels match the levels of the cinema DTS track should not be totally disregarded as a potential guide to the correct LFE level.

[zoidberg]

Agreed, by all accounts the UHD Nolan Batmans have far field mixes as does the Criterion blu of The Game, comparing the cinema DTS tracks for these should help

[Turisu]

LOTR is really making my head hurt now. So the cinema DTS of FOTR has the same LFE level as ROTK (I can boost by +4dB without clipping) *BUT* the BD LFE level is far lower than the ROTK BD. The FOTR BD LFE is only a little louder than the untouched cinema DTS level.

Why would they neuter the LFE on FOTR for the BD?

What on earth do I do with these movies?

[allldu]

Agreed, by all accounts the UHD Nolan Batmans have far field mixes as does the Criterion blu of The Game, comparing the cinema DTS tracks for these should help

Are you sure about Nolan's Batmans? They say "original theatrical mixes" on the cover. Actually, all Nolan's movies on UHD have that on the cover.

[Turisu]

I just took a look at Batman Begins. And I think it provides the answer.

The UHD 5.1 is indeed identical to the Cinema DTS track. In order to make the untouched Cinema DTS match the UHD levels we need to attenuate the surrounds by -3dB and boost LFE by +6dB just as zoidberg said. The key thing is that this causes the same clipping as with LOTR but that same clipping is on the UHD track.

So the reasonable conclusion to this is that zoidberg's +6dB LFE adjustment is correct for all post-1999 movies and any resulting clipping would have been present in the hardware decoded theatrical presentation.

[zoidberg]

I can't take credit for the adjustments, schorman did the real work and pipefan413 put this thread together.
It sounds like the clipping (ie peaks exceeding the max. level) was present in the original mix, although of the 3 digital theatre formats the DTS wouldn't have the clipping due to the headroom in the encoding. Technically it would have the loudest LFE but only to 90Hz not the 120Hz of AC-3/SDDS.
I wonder if the clipping can be detected by the human ear at that frequency range?