(2026-06-03, 05:24 PM)stwd4nder2 Wrote: Gladiator and Fight club are DTS-ES mixes. Maybe it's just messing up those?

0.3.13 creates the 7th channel for "Gladiator", too. And the channel order was correct. Only 0.3.14 screws it up.

I think I need more DTS CDs for testing to find common patterns. I'm currently working on comparing different sound tracks from cinema and several home releases. Especially, comparing to the German sound tracks is sometimes interesting, because they usually don't get new mixes on new releases and stay more "original". For "Fight Club" the German tracks on Blu-ray have the least amount of clipping. The english tracks show a high amount of clipping and a compressed LFE. Only the new english UHD mix is nearly flawless in this regard.

The audio mixes on a number of Nolan movies were confirmed to be the far-field mixes, surely comparing the decoded C-DTS tracks would confirm the correct levels?

I own the Criterion Blu-ray of "The Game" (1997) which contains the original theatrical mix as stated on the Criteroin home page and inside the blu-ray booklet. I compared it to the Cinema DTS CD.

I applied a high pass to the surrounds of the Blu-ray track at 120 Hz with 24 dB/oct. Otherwise amplitude and RMS are higher than on the already cut-off cinema track. All tracks were normalized so that the center channels had the same amplitude. I chose the scene with the loud music for comparison. The cinema track was decoded with +6 dB on the LFE and -3 dB on the surrounds with version 0.3.13.

The result are:

Contrast (RMS difference) Cinema - Blu-ray:

LFE: 0,78 dB (cinema is a bit louder)
Surrounds: -2.7 dB (Blu-ray is louder)

The front channels have a similar difference as the LFE. It varies between 0.2 - 1.1. If we assume that the Blu-ray track has the correct levels, the cinema track has to be decoded with -1 dB for the surrounds. -3 dB seems to be too much decrease.

One more thing: I always analyze the lower end of the spectrum with BEQDesigner, because it features a high resolution there. The high pass (usually around 20 - 30 Hz) often differs from track to track. Sometimes the Blu-ray has an steeper high pass or a higher cut-off frequency. And I have also seen it the other way around. In this case the Blu-ray is filtered a bit stronger, but not much. In any case a different spectrum makes comparison of amplitude and RMS difficult.

(2026-06-03, 10:23 PM)FoLLgoTT Wrote: I compared it to the Cinema DTS CD.

Many thanks for that.

(2026-06-03, 10:23 PM)FoLLgoTT Wrote: The cinema track was decoded with +6 dB on the LFE and -3 dB on the surrounds with version 0.3.13.

And with which Foobar variant - 64-Bit or 32-Bit? Because that also seems to make a difference as if we wouldn't already have enough variables in the equation. Currently, I can't test it for myself, hence the question.

(2026-06-03, 10:23 PM)FoLLgoTT Wrote: If we assume that the Blu-ray track has the correct levels, the cinema track has to be decoded with -1 dB for the surrounds. -3 dB seems to be too much decrease.

Hmm, can we assume that? Then the plugin v0.3.13 would be doing funny things because that -3dB attenuation for Cinema DTS seems to be pretty much a given, doesn't it?

(2026-06-03, 10:23 PM)FoLLgoTT Wrote: In this case the Blu-ray is filtered a bit stronger, but not much. In any case a different spectrum makes comparison of amplitude and RMS difficult.

So no "weaker bass impression" in this case from your experience?

(2026-06-03, 11:51 PM)little-endian Wrote: And with which Foobar variant - 64-Bit or 32-Bit? Because that also seems to make a difference as if we wouldn't already have enough variables in the equation. Currently, I can't test it for myself, hence the question.

I used 64 bit floating point format. I noticed that there are sometimes high amplitudes which would be clipped with 16 bit integer. They are not on the Blu-ray tracks. Maybe, because of the lossy cinema DTS format or the high pass on the surrounds. This makes level comparison even more complicated.

Quote:Hmm, can we assume that? Then the plugin v0.3.13 would be doing funny things because that -3dB attenuation for Cinema DTS seems to be pretty much a given, doesn't it?

I don't think that we can assume that, yet. I will do more comparisons and move on with the Nolan movies. I'll write my findings here.

Quote:So no "weaker bass impression" in this case from your experience?

I had no time to listen to "The Game" in my home theater. I guess the different high pass is not audible at all in this case. Regarding "Fight Club" this is completely different, because the lower end is much more extended on the english home video tracks. I listened to them and noticed the tactile difference, because my (FIR filtered) riser with voice coil motors is very good in reproducing frequencies < 30 Hz. Of course, without the riser the difference may be questionable.

Decoder 0.3.15: channel order is fixed now and 7th channel is back again.

I analyzed "Inception". The english and german tracks on the UHD BD match the Cinema DTS track very closely. -3 dB on the surrounds seems to be the correct setting here.
But there is a very steep low pass on the LFE at about 60 Hz which leads to a very different spectrum. I have never seen such a big difference on the upper end of the LFE. Usually the difference is on the lower end. When I apply a similar low pass the RMS difference is only about 1 dB.

And regarding "The Game": I have to correct myself. The high pass is audible.
And I noticed (only by listening in stereo downmix) that the right channel has a delay of about 8 samples. This leads to a cancellation with the center channel when the music plays. It sounds muddy and left-shifted. This is extremely audible in stereo downmix with enabled center (not without center), but nearly invisible when analyzing the waveform. The delay is fixed on the Criterion Blu-ray and the German Blu-ray. It looks like the Cinema DTS track has a bug.

(2026-06-04, 07:29 PM)FoLLgoTT Wrote: And I noticed (only by listening in stereo downmix) that the right channel has a delay of about 8 samples. This leads to a cancellation with the center channel when the music plays.

Are you sure that it's only 8 samples? At a sample rate of 44100 Hz, that would equal not even 0,2 ms or about 6cm of speaker distance difference. 🤔

Filtering the low frequencies from the surround channels will have the effect of lowering their overall dB level. Do the blu/UHD mixes have full frequency surround channels (ie content below 80Hz) vs the filtered surround channels present in C-DTS?

(2026-06-04, 08:57 PM)zoidberg Wrote: Filtering the low frequencies from the surround channels will have the effect of lowering their overall dB level. Do the blu/UHD mixes have full frequency surround channels (ie content below 80Hz) vs the filtered surround channels present in C-DTS?

Would that filtering be an inherent part of the mix, or just something introduced in the DTS encoding process?

(2026-06-04, 08:40 PM)little-endian Wrote: Are you sure that it's only 8 samples? At a sample rate of 44100 Hz, that would equal not even 0,2 ms or about 6cm of speaker distance difference. 🤔

I shifted the channel a few samples and it sounded much better and more centered. I didn't tweak it to the max.
When played with a center speaker the effect is much less audible, because of the three speakers, reflections, HRTF etc. in comparison to the stereo downmix.

(2026-06-04, 08:57 PM)zoidberg Wrote: Do the blu/UHD mixes have full frequency surround channels (ie content below 80Hz) vs the filtered surround channels present in C-DTS?

Yes, always. So, for comparison a high pass is mandatory.