Atmos vs 5.1

[LuvMyQuad]

In short, their leveraging of their existing system for backwards compatibility is on point! The old codec 'thinks' it's adding a resolution layer to the core audio. The "true" part. Surprise! It's adding audio elements from object channels.

So True HD has resolution layers? It's the first I've heard of this. I thought it was basically just MLP repackaged. I know DTS has such layers, Can anyone else confirm what Jim is saying?

 

[sjcorne]

Extraordinary claims (ie. Object audio is mixed into the bed audio and separated out on playback.) demand extraordinary evidence.

The claim that 'older codecs' somehow omit object-oriented information upon playback also seems pretty extraordinary, and I've not seen any evidence to support it.

 

[himey]

I've not seen any evidence to support it.

I have not heard it either, and I have been listening for it. It would be very obvious.

 

[SeeMoreDigital]

Many release do not have a true dedicated lossless 5.1 mix. And just as often, the 5.1 that is there is inferior to what you get from a mixdown

Hmmm... I wouldn't call it many

 

[himey]

Hmmm... I wouldn't call it many

Most either doesn't have a 5.1 option or are from the same production as the Atmos and not a dedicated mix. Others are slightly tweaked from the Atmos production. Very few new 5.1 mixes that also include an Atmos mix, are completely a reworked version apart from the Atmos, and will sound more similar than different. When I listen to a good Atmos mix in 5.1, I usually think that it would be difficult for a dedicated 5.1 mix to beat it. Most Atmos mixes I listen to, will never have a 5.1 mix, ever, and that is fine by me.

 

[LuvMyQuad]

Hmmm... I wouldn't call it many

These are the last 10 Atmos releases I've purchased. The bold titles do not include dedicated 5.1 mixes, so 4 of 10.

The Floyd releases had 5.1 versions previously released. If that were not so, would they have been released as 5.1? Maybe. Maybe not.

Alanis Morissette, Jagged Little Pill
Big Big Train, From Where We Came
Fish, 13th Star
Mark Knopfler One Big River
Mark Knopfler' Guitar Heroes, Going Home (download)
Peter Gabriel, IO
Pink Floyd, Animals
Pink Floyd, DSOTM
Robin Trower, Bridge of Sighs

 

[AYanguas]

For me it’s easy to understand.

In a similar way to the fact that we assume that an upmix of a Stereo mix cannot give the same creative results as a discrete 5.1 mix, it is necessary to also assume that a “downmix” of an Atmos to a 5.1 channels may also not give the same creative result. (Ignoring the obvious reduction of channels/speakers, of course.)

That “downmix” is generated automatically by the Dolby tools when generating the final TrueHD or DD+ distribution file. It appears as an additional ‘substream’ in those files, which can be backwards decoded by older decoders that do not support Atmos and do not understand the other ‘Atmos substream’ with “Bed channels 7.1.2” and/or “Atmos Objects”.

That is, the artistic quality of an Atmos mix listened in a 5.1 system will depend on the type of mix that the artist mix engineer has made. It can sound good, average or bad, depending on how the overall mix was done in Atmos.

When the artist mixer or producer worries about this, if he considers that a particular Atmos mix does not sound correct in a 5.1 system, then he will worry about making an additional ‘reworked’ multi-channel 5.1 mix (non Atmos) that sounds good. And it will be included as a separate independent additional track on the Blu-Ray or as independent additional 5.1 files.

The latter may be a little more expensive to produce, but I think the main reason would be the love and concern that the mixer or producer shows for it, regarding the consumers that have 5.1 Home Theater discrete speakers.

As other post said, old mixers/producers who have grown with the previous market of surround 5.1, would have that in mind and will test very carefully if the Atmos translates well to 5.1 and, if not, will generate the additional 5.1 mix for that old consumers.

 

[himey]

That is, the artistic quality of an Atmos mix listened in a 5.1 system will depend on the type of mix that the artist mix engineer has made. It can sound good, average or bad, depending on how the overall mix was done in Atmos.

A good Atmos mix translates to 5.1 just fine most of the time and will sound better in 5.1 if the Atmos playback system is inferior, despite having more channels.

 

[kfbkfb]

https://professionalsupport.dolby.c...d-Stereo-downmix-settings-work?language=en_US

(might be of some help)


Any idea of how many Dolby Atmos physical media buyers (in my case, mostly movies, a few Blu-ray audios) listen to the DD 5.1 soundtrack only (I do)?

Is it worth the Atmos mixers time to fine tune the DD 5.1 downmix?


Kirk Bayne

 

[Owen Smith]

In short, their leveraging of their existing system for backwards compatibility is on point! The old codec 'thinks' it's adding a resolution layer to the core audio. The "true" part. Surprise! It's adding audio elements from object channels.

Please provide some hard evidence that this is how Atmos works, like some documentation from Dolby. Or some separated out streams eg an Atmos stream with test tones, then the raw 7.1 extracted from it manually not with a decoder, and that should show that the objects are simply missing if you are correct. But don't use audio software to do the 7.1 extraction, if you are correct it will add them back in knowing that is required. The 7.1 has to be extracted manually.

I have the software from the mouth of the beast itself. The first thing I did was render and encode something and then decode it and A/B it with the master. There's no bs here! If you are hearing something that sounds corrupt as though it was subjected to de-mix stem separation, something else is going wrong. The format is not doing that!

So your reason for saying Atmos doesn't remove the objects from the 7.1 True HD is because you think it would sound bad if it did. That is not evidence. Furthermore whenever DTS plays as 7.1 (no DTS:X involvement here) the rears data is removed from the side surrounds, since as encoded in DTS 7.1 the side surrounds contain all the sound needed for 5.1 playback. No-one has complained about side surrounds being corrupt whenever playing DTS 7.1 so clearly this is possible. This is also how DTS 6.1 ES Discrete works, the centre rear data has to be removed from the surrounds for correct 6.1 playback.

Someone else posted something a long time ago saying the data is in the 7.1 (which I agree with) and is cancelled by appropriate opposite phase sound being played from the height speakers to remove it from the 7.1 in the room. I don't believe that, because the phase required will vary depending on the relative positions of all the speakers and also the room acoustics, none of which the Atmos decoded knows.

 

[steelydave]

The claim that 'older codecs' somehow omit object-oriented information upon playback also seems pretty extraordinary, and I've not seen any evidence to support it.

You know how the old proverb goes, "a little bit of knowledge is a dangerous thing..."

Obviously Dolby is never going to give up their secret recipe of 11 herbs and spices, but an engineer with experience in Atmos told me that the Atmos metadata - or "sidecar" as he called it - didn't have unique audio information. From what he said (or my understanding at least) was that it was "like the old quad formats" (and I'm guessing he meant more CD-4 than matrix quad) in that the Atmos metadata basically includes audio that is to be deleted or reduced in volume from the core TrueHD (or DD+) stream and pushed toward the height speakers, along with metadata the tells it which channels are affected. So if you had a guitar in the top right rear speaker in Atmos, in the 7.1 it would be in the rear right speaker, and the Atmos metadata would have the recording of that element, along with the information to remove it (via phase cancellation or similar) and to place it in the top speaker.

At the end of the day, Atmos is an add-on to TrueHD and Dolby Digital Plus, so it's bound to be backward compatible with those codecs and to devices that support them that were created before Atmos was a thing. If height information was stored completely separately these pre-Atmos TrueHD/DD+ AVRs would completely ignore it and the mixes would simply be missing all the music from the height channels, which it simply isn't.

If you want more evidence that Atmos uses this subtract from the floor and push to the heights method, put your ear to one of the height channels in streaming Atmos. You'll hear all kinds of burbly artifacts, and this is because with DD+ the Atmos metadata for things being pushed to the heights isn't an exact duplicate of what's in the 7.1 bed - thanks to lossy compression it's just an approxmiation, and the burbling is thanks to the difference between the two. Sort of like if you took two 128kbps mp3s, inverted the phase of one of them and then mixed them together, you'd get all kinds of weird noise, whereas if you did the same with two CD-quality .wav files you'd get digital silence.

 

[Owen Smith]

So True HD has resolution layers? It's the first I've heard of this. I thought it was basically just MLP repackaged. I know DTS has such layers, Can anyone else confirm what Jim is saying?

I'll confirm that I don't believe what he's saying about this.

The claim that 'older codecs' somehow omit object-oriented information upon playback also seems pretty extraordinary, and I've not seen any evidence to support it.

Indeed. Extraordinary claims require extraordinary evidence, and no evidence at all has been provided. Not even which "older codecs" he's talking about.

 

[AYanguas]

A good Atmos mix translates to 5.1 just fine most of the time and will sound better in 5.1 if the Atmos playback system is inferior, despite having more channels.

Well, an "inferior" system will sound always probably worse that a better one, even if the better one is only Stereo.

In addition to the number of channels, we know that it is also important the quality of the system, calibration, speakers locations, room conditioning, etc.

That would mean that the quality requirements are bigger for a multi Atmos Speaker system than for a 5.1 system to sound "good enough".

 

[Owen Smith]

That “downmix” is generated automatically by the Dolby tools when generating the final TrueHD or DD+ distribution file. It appears as an additional ‘substream’ in those files, which can be backwards decoded by older decoders that do not support Atmos and do not understand the other ‘Atmos substream’ with “Bed channels 7.1.2” and/or “Atmos Objects”.

I think here you are talking about the Dolby Digital 5.1 substream, and that's not what we're talking about that. We're talking about playing the True HD 7.1 substream on a 5.1 system without decoding the Atmos objects.

 

[Owen Smith]

If you want more evidence that Atmos uses this subtract from the floor and push to the heights method, put your ear to one of the height channels in streaming Atmos. You'll hear all kinds of burbly artifacts, and this is because with DD+ the Atmos metadata for things being pushed to the heights isn't an exact duplicate of what's in the 7.1 bed - thanks to lossy compression it's just an approxmiation, and the burbling is thanks to the difference between the two. Sort of like if you took two 128kbps mp3s, inverted the phase of one of them and then mixed them together, you'd get all kinds of weird noise, whereas if you did the same with two CD-quality .wav files you'd get digital silence.

I agreed with everything else you said, but not this. This is merely evidence that streaming Atmos is lossy encoded. If the height objects sound worse than the main channels then that's only evidence that they're more lossy compressed in some way, which includes but is not limited to the extraction not being perfect due to lossy differences.

 

[himey]

I think here you are talking about the Dolby Digital 5.1 substream

Not just a sub-stream, but what you get as the 5.1 option on some Blu-rays.

 

[AYanguas]

Regarding the issues of what are bed channels (in Atmos), and how a TrueHD with Atmos is decoded on an old processor that does not support Atmos, I did some research on the internet a while ago and I actually did not find the details of the Dolby internals.

In this post you can find a summary of what I have found: https://www.quadraphonicquad.com/fo...tmos-parts-appear-in-stereo.35022/post-741118

Please, if you have any evidences for or against that post, I would like to hear them.

 

[Owen Smith]

Not just a sub-stream, but what you get as the 5.1 option on some Blu-rays.

Why would I want that? If I have a True HD decoder I want to play the lossless True HD 7.1 substream, not the lossy and rather low bit rate Dolby Digital 5.1 substream. If this is what the 5.1 option means on a particular Blu Ray they should make it clear it is sub optimal or just Dolby Digital.

 

[himey]

Why would I want that? If I have a True HD decoder I want to play the lossless True HD 7.1 substream, not the lossy and rather low bit rate Dolby Digital 5.1 substream. If this is what the 5.1 option means on a particular Blu Ray they should make it clear it is sub optimal or just Dolby Digital.

It is cheaper in man hours and can sound fine if the Atmos mix is good and the mixer takes the 5.1 output into consideration. It wouldn't be output as lossy.

 

[jimfisheye]

The "recording of that element" in the metadata would be a difference audio signal. The result of subtraction of one audio signal from another. Precisely the style of relationships and manipulation used in past systems, yes! That's still object audio in the metadata though. @steelydave does a much better job of accuracy than I did but we're not in disagreement.

My point was meant to be that the system keeps actual discrete audio channels throughout the process. Involving lossless phase inversions and mlp mysteries as it may be. I thought I read it the other way around when I read some tech papers a few years ago now. Maybe I read a 'simplified' version? But alright... difference signals in the metadata then from all the object channels. Same end result.