2018-04-25, 02:55 PM
It's a fake (matrixed) 7th channel. It's not comparable to DTS-HD MA 7.1 or DTS-HD MA 6.1 and the hint is in the name. DTS-ES is different from DTS-HD MA. Just to demonstrate how weird it all is: Even DTS-HD is not always lossless, for example DTS-HD HRA is a lossy extension of the base DTS-HD standard. It's a confusing mess.
Here's from Wikipedia:
You have 6 actual channels. The 7th is just a matrixed rear one.
There's also a version called DTS-ES Discrete, which actually has 7 discrete channels and the rear is NOT matrixed. Here's again from Wikipedia:
So unless it is the specific "Discrete" variant, there is likely no 100% standardized way of extracting the rear channel I'm afraid and we have a similar situation as with Dolby Surround/Dolby Stereo. Although maybe there is a reference implementation of it somewhere in the chips/decoders they sell.
When you convert this DTS-ES to FLAC or WAV, you will get 5.1. That is the lossless conversion. The matrixed channel is still hidden in there in the two side channels, but you don't get it as a discrete one.
Why does MediaInfo say "7 channels"? I don't know. They must have figured that's the most fool-proof way of describing the format?
With that said, it's probably a bit pointless to convert it in the first place and smarter to just keep it in the format it is, as it is smaller and reencoding will only further reduce quality (unless you're delivering as lossless).
Also, personally, I am not sure how much point there is in a 96kHz lossy DTS over a 48 kHz PCM. I doubt the frequency response is actually that great up to the Nyquist frequency (around 48kHz for 96kHz I think). You could check it with a spectrum analysis of the original or the converted 5.1 FLAC/WAV tho.
Here's from Wikipedia:
Quote:DTS-ES Matrix provides 5.1 discrete channels, with a matrixed center-surround audio channel. DTS processors that are compatible with the ES codec look for and recognize "flags" built into the audio coding and "unfold" the rear-center sound from data that would otherwise be sent to rear surround speakers. DTS decoders which do not understand ES process the sound as if it were standard 5.1, and the matrixed audio for the center surround channel is output equally from the two surround speakers (very much as a sound intended to be in the centre of the sound field in a stereo recording is played equally by the left and right speakers). This is notated as DTS-ES 5.1.
You have 6 actual channels. The 7th is just a matrixed rear one.
There's also a version called DTS-ES Discrete, which actually has 7 discrete channels and the rear is NOT matrixed. Here's again from Wikipedia:
Quote:DTS-ES Discrete provides 6.1 discrete channels, with a discretely recorded (non-matrixed) center-surround channel; in home theater systems with a 7.1 configuration, the two rear-center speakers play in mono. To maintain compatibility with DTS decoders which do not support DTS-ES, the center-surround channel is also matrixed into the left and right surround channels, so that the rear center channel's sound is still present when played in 5.1 on a non-ES system; an ES decoder removes the matrixed audio from these two channels when playing back DTS-ES Discrete soundtracks. DTS-ES Discrete is sometimes notated as DTS-ES 6.1. Only a few DVD titles have been released with DTS-ES Discrete.
So unless it is the specific "Discrete" variant, there is likely no 100% standardized way of extracting the rear channel I'm afraid and we have a similar situation as with Dolby Surround/Dolby Stereo. Although maybe there is a reference implementation of it somewhere in the chips/decoders they sell.
When you convert this DTS-ES to FLAC or WAV, you will get 5.1. That is the lossless conversion. The matrixed channel is still hidden in there in the two side channels, but you don't get it as a discrete one.
Why does MediaInfo say "7 channels"? I don't know. They must have figured that's the most fool-proof way of describing the format?
With that said, it's probably a bit pointless to convert it in the first place and smarter to just keep it in the format it is, as it is smaller and reencoding will only further reduce quality (unless you're delivering as lossless).
Also, personally, I am not sure how much point there is in a 96kHz lossy DTS over a 48 kHz PCM. I doubt the frequency response is actually that great up to the Nyquist frequency (around 48kHz for 96kHz I think). You could check it with a spectrum analysis of the original or the converted 5.1 FLAC/WAV tho.