Slashdot Mirror

There is no disk format for "DTS Neural Surround" as such.

Right, it's a phase matrix format, like Dolby Stereo or Pro Logic II. And like those, you can listen to the LtRt as a plain left-right and you'll still hear something that's tolerable stereo, it's just perilous for certain applications (it'll have bad mono compatibility).

See also this press release. A radio station was broadcasting in 5.1 using Neural Surround... broadcasting in ordinary stereo FM as well as HD radio. Anyone could listen in stereo, but those with Neural Upmix in their stereo receivers could hear 5.1 sound.

Right, it's a format. In a professional context this would be called a format. Format does not imply digital. Optical SVA Dolby SR is a "format" in this context.

Then I would say that DTS 11.1 is an actual format exactly the way Dolby Surround was an actual format.

Neural Surround is a format, it's a 7.1 format. DTS 11.1 is Neo:X, at least it's the only one I can find.

If you read their white paper it says nothing about encoding or encoding format, Neo:X is an upmixing system. This is what it is:

To match the user's speaker layout, DTS Neo:X separates input frequencies into sub-bands and then creates additional output channels. As it adds depth and intensity, it maintains the integrity of the original audio, keeping elements in their intended locations.

Okay, any box that "adds depth and intensity" is rank woo. You can't take a stereo mix from, say, the 1980s, run it through an upmixer and call that a high-fidelity process, you're making up a mix that the original mixer never intended. If the guy who made the mix didn't make it in DTS 11.1, there's absolutely no reason for you to play it back in 11.1, unless you're an audiophile or a horn-counting crackpot. But this is what the DTS gear is selling -- play any format through our box and we'll make it surround, that's the pitch. (More like, we won't make you feel like an idiot for buying all those speakers when nothing on cable, Internet or Blu-Ray uses them.)

I'm a feature film sound designer and mixer, DTS is completely out of theatrical and television -- the original theater format is owned by a different company now, and the DTS name is just used to sell stereo equipment. DTS has become is the Bose of home theater. Go through their website and there's basically nothing for filmmakers, film mixers or TV mixers. I can't find any documentation of a Neo:X encoder, or a licensing system for productions. All of their "for professionals" documentation looks like its for home theater installers. Neo:X is just a box in your house that sprinkles fairy dust on your speaker outputs.

The difference with Dolby Stereo is Dolby Stereo uses an encoder, an SEU4 or the AAX/VST/AudioUnits equivalent, and you need the decoder in order to hear the mix in surround, and the filmmaker has complete control over how sound is placed and presented, not a box in the viewer's living room. It's up to the filmmaker to add "depth and intensity" if he wants to, not some box.

The only movie I can actually find that's advertised as Neo:X is Expendables 2, and it doesn't appear to be encoded for Neo:X, it's just been "optimized" for Neo:X, which means they just listened to the mix it through a Neo:X receiver and made a whole new mix just for people with Neo:X receivers -- in other words, the mixer had to change his mix so the Neo:X box wouldn't screw it up the presentation. Any box that remixes my work is crap. We spend weeks getting this stuff right and getting the director, editor and producers to sign off on every little detail, and then some home theater box is going to re-pan everything, add nonstandard height channels and multiband crossovers? What's the point of that, except to sell speaker gear?

[DTS Neural Upmix is] marketed as a spatializing upmixer that can also decode Neural Surround (which is a third format not necessarily related to Neo:X).

No, there is no "Neural Surround" format as such. Neural Downmix uses phase encodings and the output is just an audio stream (can be analog, saved as a wave file, saved as DTS Master Audio, saved as MP3, etc.).

Look at this PDF. There are two columns: one shows different disk formats and how many bits per second each one needs; the other column has one thing in it, Neural Surround. This is because Neural Surround isn't a format as such.

http://www.dts.com/~/media/d5aad4e0d179439c8588ac3b61e37444/DTS_Broadcast_infosheet.pdf

See also this press release. A radio station was broadcasting in 5.1 using Neural Surround... broadcasting in ordinary stereo FM as well as HD radio. Anyone could listen in stereo, but those with Neural Upmix in their stereo receivers could hear 5.1 sound.

http://www.prnewswire.com/news-releases/wguc-fm-begins-51-broadcasting-with-neural-surroundtm-56267632.html

But this feature is sorta incidental, as literally nothing is mixed in Neural Surround.

I'm sorry but you are completely mistaken on this point. Let's look at the DTS web site again:

DTS Neural Surround DownMix technology reduces multichannel surround sound to a stereo mix that accurately represents the original intent of the content creator.

The DTS Neural Surround DownMix uses patented âoeActive Correctionâ technology. By analyzing the audio, the phase and intensity are rewritten, creating a pristine Lt/Rt stereo mix.

This process eliminates problems that traditionally occur in matrix surround downmix systems, such as comb filtering and spatial distortion. DTS Neural Surround DownMix creates a natural sounding stereo mix that is spatially true to the original multichannel localization.

http://www.dts.com/professionals/sound-technologies/audio-processing/dts-neural-surround.aspx#downmix

Note the phrase "pristing Lt/Rt stereo mix" and the concerns about comb filtering in the output mix. There is mixing going on here.

DTS Neural Downmix produces a stereo output stream which may be saved in any format. You can feed the result to DTS Neural Upmix, even as an analog waveform, and it will upmix using the encoded signals. There is no disk format for "DTS Neural Surround" as such.

My understand is that the height channels are encoded sum-and-difference with the main L-R channels, and a special decoder reads reads additional channel data to subtract out the height channels from the mains.

I think it is possible that there is some additional metadata embedded in the DTS Master Audio bitstream, because old DTS decoders do understand metadata tags and will ignore them. But there is no bitstream change from plain DTS 7.1 to DTS 11.1, and you can play the 11.1 stream on an old DVD player and you will get 7.1 out. (Just like you could play Dolby Surround on a stereo and get stereo out, if you didn't have the Dolby Surround decoder to upmix from stereo to surround.)

If you are still convinced that DTS 11.1 has additional discrete channels, please find a reference and show me. If I'm wrong, I'm wrong, but I think the DTS web page I referenced in the previous post backs me up.

By "actual format" I mean its a communications channel where the sender and recipient agree on what goes into the channel and what is supposed to come out.

Then I would say that DTS 11.1 is an actual format exactly the way Dolby Surround was an actual format. Both rely on specific, agreed-upon phase encodings

[DTS Neural Upmix is] marketed as a spatializing upmixer that can also decode Neural Surround (which is a third format not necessarily related to Neo:X).

No, there is no "Neural Surround" format as such. Neural Downmix uses phase encodings and the output is just an audio stream (can be analog, saved as a wave file, saved as DTS Master Audio, saved as MP3, etc.).

Look at this PDF. There are two columns: one shows different disk formats and how many bits per second each one needs; the other column has one thing in it, Neural Surround. This is because Neural Surround isn't a format as such.

http://www.dts.com/~/media/d5aad4e0d179439c8588ac3b61e37444/DTS_Broadcast_infosheet.pdf

See also this press release. A radio station was broadcasting in 5.1 using Neural Surround... broadcasting in ordinary stereo FM as well as HD radio. Anyone could listen in stereo, but those with Neural Upmix in their stereo receivers could hear 5.1 sound.

http://www.prnewswire.com/news-releases/wguc-fm-begins-51-broadcasting-with-neural-surroundtm-56267632.html

But this feature is sorta incidental, as literally nothing is mixed in Neural Surround.

I'm sorry but you are completely mistaken on this point. Let's look at the DTS web site again:

DTS Neural Surround DownMix technology reduces multichannel surround sound to a stereo mix that accurately represents the original intent of the content creator.

The DTS Neural Surround DownMix uses patented âoeActive Correctionâ technology. By analyzing the audio, the phase and intensity are rewritten, creating a pristine Lt/Rt stereo mix.

This process eliminates problems that traditionally occur in matrix surround downmix systems, such as comb filtering and spatial distortion. DTS Neural Surround DownMix creates a natural sounding stereo mix that is spatially true to the original multichannel localization.

http://www.dts.com/professionals/sound-technologies/audio-processing/dts-neural-surround.aspx#downmix

Note the phrase "pristing Lt/Rt stereo mix" and the concerns about comb filtering in the output mix. There is mixing going on here.

DTS Neural Downmix produces a stereo output stream which may be saved in any format. You can feed the result to DTS Neural Upmix, even as an analog waveform, and it will upmix using the encoded signals. There is no disk format for "DTS Neural Surround" as such.

My understand is that the height channels are encoded sum-and-difference with the main L-R channels, and a special decoder reads reads additional channel data to subtract out the height channels from the mains.

I think it is possible that there is some additional metadata embedded in the DTS Master Audio bitstream, because old DTS decoders do understand metadata tags and will ignore them. But there is no bitstream change from plain DTS 7.1 to DTS 11.1, and you can play the 11.1 stream on an old DVD player and you will get 7.1 out. (Just like you could play Dolby Surround on a stereo and get stereo out, if you didn't have the Dolby Surround decoder to upmix from stereo to surround.)

If you are still convinced that DTS 11.1 has additional discrete channels, please find a reference and show me. If I'm wrong, I'm wrong, but I think the DTS web page I referenced in the previous post backs me up.

By "actual format" I mean its a communications channel where the sender and recipient agree on what goes into the channel and what is supposed to come out.

Then I would say that DTS 11.1 is an actual format exactly the way Dolby Surround was an actual format. Both rely on specific, agreed-upon phase encodings

There's a fundamental difference between an encoded mix and an upmixer. Dolby Surround is intended to be decoded from 2 tracks into LCRS, the filmmakers mixed the film in Dolby Stereo and were listening to the surrounds so they know what's in them. The phase encoding is part of the channel spec.

I'm with you so far.

An upmixer takes a stereo or 5.1 mix and applies effects to it to make it sound like it was mixed in a wider format, but there's nothing really being decoded, it's just synthesizing or guessing what should be in the additional channels using heuristics, all-pass filters, delays, crossover networks and other stuff that sounds cool or "provide a good experience" but, in fact, interfere with the filmmaker's intent.

The original Dolby surround and DTS Neural Upmix can both be applied to any stereo recording and some sort of upmix will occur, but both were designed to be used with a mix that was intended to be upmixed. DTS also sells DTS Neural Downmix which can take a 5.1 or 7.1 stream and output stereo with intentionally encoded signals that decode back to 5.1 or 7.1 sound.

When DTS Neural Upmix is working from a stereo signal that was made using DTS Neural Downmix, you get a really clean surround sound with no leakage. I used to listen to the multichannel recording of "Money" by Pink Floyd, and the cash register and coin sound effects very cleanly came from all the different directions like the original multichannel mix.

Again, you can't fit 8 kilos of flour into a 2-kilo sack, so 7.1 audio sent through downmix, then upmixed back to 7.1, can never perfectly reproduce the original multichannel recording. But I was impressed by just how well it did.

Despite the name "Neural Upmix", it is designed to work with phase-encoded signals intentionally mixed using Neural Downmix.

Neural Upmix is an upmixer, DTS Neo:X is an actual format that decodes an 11.1. Neo:X home receivers also employ upmixing, mainly because no films are mixed in 11.1 Neo:X, it's a surround audiophile format, and it needs to do an upmix in order to justify people spending money on it.

My understanding is that DTS 11.1 audio uses intentionally encoded signals for the height channels, but the on-disk format is DTS Master Audio 7.1 (no additional discrete channels).

Just as the original Dolby surround could be listened to in stereo if you didn't have surround speakers, the 11.1 mix can be listened to in 7.1 if you don't have height speakers; in both cases, the downmix process is supposed to not add anything objectionable.

I don't know what you mean by "DTS 11.1 is an actual format"... if you mean that it has 12 discrete channels, I believe you are mistaken on this point.

Here's how DTS describes the 11.1 system:

An important goal of the DTS multi-tiered plan is to enable content creators to produce 3D audio and provide it to consumers without changing the delivery chain. With the DTS Neo:X capability for near discrete Height/Wide output, studios can produce directional cues intended only for these speakers, with no audible leakage into other channels. Studios can also produce soundtracks optimized for DTS Neo:X that offer a compatible listening environment in âoestandardâ multi-channel playback configurations.

From the "How it works" tab on this page:

http://www.dts.com/professionals/sound-technologies/audio-processing/dts-3d-audio.aspx

"without changing the delivery chain": no new audio format, disks play fine on older DTS decoders

"no audible leakage": there's no problem with leakage if you have discrete channels; if we are even talking about leakage, we are talking about an upmix.

I don't believe Imade any mistakes in my original post.

Forget what HD output? Component video cables provide equal (or better) quality output to HDMI (especially to those of us with CRT HDTV sets). HDMI is only barely interesting for newer resolutions and encryption (ugh) and its ability to carry audio as well (which no HiFi geek or audiophile wants anyway).

I'm currently using (and plan to continue using) component video cables to my TV and coax digital audio cables to my receiver for the near future. When I buy a new receiver with DTS Master audio support, I'll be going to new audio cables. If and when LCD or Plasma TVs look as good at variable resolutions as my CRT, I might use HDMI for video only.

Yes yes, I know, in a few years, some discs might be encoded so that I need HDCP to play them on my TV at all. For the reasonable future, it would seem that movie companies care more about sales figures though and won't be setting that bit.

Let us get the names of those entities that are threatening with software patents.

Here's one: www.dts.com