AC3 Mixes of Laserdiscs do Not Contain True Theatrical Masterings

[Kado_2020]

As a simple example

Terminator 2 Laserdisc AC3 center channel | Near field mix

Integrated LUFS: -24.5 dB
LRA: 21.2dB
Highest peak level: -4.9 dBFS


Terminator 2 Live/Artisan DVD AC3 center channel | Original theatrical (CDS) mix

Integrated LUFS: -21 dB
LRA: 22.8dB
Highest peak level: -0.1 dBFS

[Kado_2020]

The same goes for other AC3 5.1 laserdisc film mixes, which have serious limitations on their dynamic range and gain levels. I analyzed most of them.

[dvdmike]

Yeah, this doesn't prove anything unless you are directly comparing to a direct AC-3 stream from a print.

[Kado_2020]

Yeah, this doesn't prove anything unless you are directly comparing to a direct AC-3 stream from a print.

I compared the audio recordings i had directly with ac3 streams taken from laserdiscs; then, I extracted the center channels from these recordings one by one and performed frequency analyses

[Kado_2020]

Some people will probably get mad at me for sharing this, but if you don't believe it, you can test it yourself.

[commandrbond]

I'm interested in your hypothesis but I'm not sure I understand what your proof is.

[Kado_2020]

I'm interested in your hypothesis but I'm not sure I understand what your proof is.

The reason for T2 is as follows

The integrated LUFS value is -21 dB, which means it is 35% more effective in gain separation and natural punch.

The LRA value is 22.8 dB, which means it has a 16% wider dynamic range and the center channel of LD are 16% more compressed.

The peak level is 0.1 dBFS; this is the maximum limit achievable without clipping.

Furthermore, regardless of which authentic theatrical master you examine, the peak levels in the recording will always be close to 0 dBFS. If this value exceeds -1.5 dB (which is already high), it is not the genuine original recording. Moreover, the issue is not limited to this; the LD's overall audio profile sounds somewhat little more muffled compared to the Live edition, and—much like the mixes found on Blu-ray discs—there are limitations regarding the extension of sonic details.

[dvdmike]

I'm interested in your hypothesis but I'm not sure I understand what your proof is.

The reason for T2 is as follows

The integrated LUFS value is -21 dB, which means it is 35% more effective in gain separation and natural punch.

The LRA value is 22.8 dB, which means it has a 16% wider dynamic range and the center channel of LD are 16% more compressed.

The peak level is 0.1 dBFS; this is the maximum limit achievable without clipping.

Furthermore, regardless of which authentic theatrical master you examine, the peak levels in the recording will always be close to 0 dBFS. If this value exceeds -1.5 dB (which is already high), it is not the genuine original recording. Moreover, the issue is not limited to this; the LD's overall audio profile sounds somewhat little more muffled compared to the Live edition, and—much like the mixes found on Blu-ray discs—there are limitations regarding the extension of sonic details.

This only holds true if you have compared to the CDS track decoded from a print.

[Kado_2020]

I'm interested in your hypothesis but I'm not sure I understand what your proof is.

The reason for T2 is as follows

The integrated LUFS value is -21 dB, which means it is 35% more effective in gain separation and natural punch.

The LRA value is 22.8 dB, which means it has a 16% wider dynamic range and the center channel of LD are 16% more compressed.

The peak level is 0.1 dBFS; this is the maximum limit achievable without clipping.

Furthermore, regardless of which authentic theatrical master you examine, the peak levels in the recording will always be close to 0 dBFS. If this value exceeds -1.5 dB (which is already high), it is not the genuine original recording. Moreover, the issue is not limited to this; the LD's overall audio profile sounds somewhat little more muffled compared to the Live edition, and—much like the mixes found on Blu-ray discs—there are limitations regarding the extension of sonic details.

This only holds true if you have compared to the CDS track decoded from a print.

Keep fooling yourself; it's not a problem for us

[LucasGodzilla]

I imagine this sort of comparison would be interesting to make with something that had a Cinema DTS mix to compare to, since captures of those seem to float around depending on the title.

That way, you can properly cross-reference it with something you can confirm is theatrical and untampered (barring semi-standardized channel attenuation) and would be in an age where digital sound mixes started taking hold.

Unfortunately, Terminator 2 is kind of a rubbish example since we have no idea how the CDS mix was handled and I believe I've heard once or twice that the LD AC3 was stemming from the 70mm 6-track mix rather than the CDS mix (as I believe the two were handled differently to account for format capabilities; not to mention the former being analog).

Sticking to Cameron, True Lies or Titanic might be a better tests for this if you can get your hands on the right assets. Granted, I never did any research on what sound elements were used for those LDs, but given Cinema DTS mixes were more standard by the time those came out, odds are they may have just ported those mixes.

Also, whilst not at all implying you made any error in your audio measuring process, make sure the AC3s have the dialnorm stripped out. Once in a while I forget about that being a factor, so nowadays I just try and automatically convert my AC3s with it auto-removed with FFMPEG as part of a preset BAT file I have around.