Atmos vs 5.1

[Owen Smith]

Yes. I want to know more. @Owen Smith , can you describe the results better? What happens with those PF test tones with no Atmos processing?

I wrote it up in considerable detail somewhere on QQ, no idea where now. What I recall with no Atmos processing is each test tone came out of every speaker except the diametrically opposite one, which I suppose means it is at least vaguely in the correct position. So rear right came out of all of my 5.0 speakers except for front left for example. Same issue with the height channels ie rear right high came out of all speakers except front left. If I switched to Atmos decode the test tones came out of just the one speaker I'd expect. Obviously given I have a 5.0 system a number of the test tones all came out of the same speaker eg. three came out of rear right (rear right high, surround right, and rear right).

 

[Owen Smith]

I think we are too.

So if what Owen says is correct, the Atmos processor is doing something beyond just a simple combination of the rear and side surrounds.

I think you're saying the same thing too, and yes my Atmos processor is doing an actual rendering of the Atmos data to a 5.0 speaker layout.

 

[Owen Smith]

While I'm here I think it's worth stating my Arcam AVR31 also lets me select DTS:X and Auro decodes to my 5.0 system. However I don't have titles that make it easy to hear the difference for those formats like DSOTM allows. I have Ex Machina 4K for DTS:X and I have the UK Blu Ray of Pixels for Auro (just about the only title that has Auro on it). Both bought just for checking the formats work for me, though I admit I watched Pixels twice since it's such good fun.

 

[himey]

While I'm here I think it's worth stating my Arcam AVR31 also lets me select DTS:X and Auro decodes to my 5.0 system. However I don't have titles that make it easy to hear the difference for those formats like DSOTM allows. I have Ex Machina 4K for DTS:X and I have the UK Blu Ray of Pixels for Auro (just about the only title that has Auro on it). Both bought just for checking the formats work for me, though I admit I watched Pixels twice since it's such good fun.

I imagine that you know what DSOTM should sound like because you have heard it so many times over the years, you would be able to hear any flaws much easier. I noticed the difference between the streaming version and the Blu-ray much more than any other title, especially with the really complicated parts.

 

[Owen Smith]

I imagine that you know what DSOTM should sound like because you have heard it so many times over the years, you would be able to hear any flaws much easier. I noticed the difference between the streaming version and the Blu-ray much more than any other title, especially with the really complicated parts.

That's part of it, but playing the DSOTM Atmos mix as True HD and especially the test tones is just so all over the place mix wise that really anyone ought to be able to tell there is something wrong.

 

[barfle]

@gvl_guy, there’s been a lot of chatter about your issue, but what I didn’t see is that you might not be hearing the signals for the side surrounds when you’re set for Atmos.

I’m pretty sure the Atmos “bed” is 7.1, so if you aren’t folding those surround channels into 7.1, you are missing almost 1/3 of the music.

Just my first estimate from limited data.

 

[himey]

That's part of it, but playing the DSOTM Atmos mix as True HD and especially the test tones is just so all over the place mix wise that really anyone ought to be able to tell there is something wrong.

There were issues with some receivers and those test tones. Are those the only ones you have tried?

 

[himey]

@gvl_guy, there’s been a lot of chatter about your issue, but what I didn’t see is that you might not be hearing the signals for the side surrounds when you’re set for Atmos.

I’m pretty sure the Atmos “bed” is 7.1, so if you aren’t folding those surround channels into 7.1, you are missing almost 1/3 of the music.

Just my first estimate from limited data.

The receiver should do that if you set the channels to off.

 

[Owen Smith]

There were issues with some receivers and those test tones. Are those the only ones you have tried?

Those are the only ones I have, so yes they are the only ones I've tried. But it's clear from playing the album that the Atmos version of the album suffers exactly the same problem as the test tones. There are enough discrete sounds in DSOTM that it's easy to check.

 

[Owen Smith]

The receiver should do that if you set the channels to off.

Every sound in Atmos is in the 7.1 True HD, the receiver doesn't have to do anything to play back as 7.1.

 

[himey]

Every sound in Atmos is in the 7.1 True HD, the receiver doesn't have to do anything to play back as 7.1.

He is playing back as 5.1, so his receiver should mix the backs into the sides, not ignoring them as the person I quoted suggested.

 

[jimfisheye]

The object audio is all in the 7.1 True HD is my understanding. I'm aware it's all lossless on BD.

In the file as metadata! Not mixed into the bed audio. I think this is more telephone game gone wrong. I don't get the sense that you mean to dis on the format with that claim. It comes across that way if taken literally.

If you mix audio, "unmixing" after the fact is just not remotely possible with full fidelity. The recent stem separation or "de-mixing" tools are impressive on the surface. The results are still lo-fi when compared to the original audio. (Run an example where you have the original multitracks. See how well the stem separation sounds next to the originals after .)

The claim that the audio is truly mixed together and separated after the fact sounds like an accusation that Atmos is faking the whole thing and crude artifact riddled results would be the best it could do.

Again, I suspect this is miscommunication! I just don't want someone unfamiliar and asking questions to take away that this is a crude lo-fi gimmick and dismiss it. That's where I was coming from. Extraordinary claims (ie. Object audio is mixed into the bed audio and separated out on playback.) demand extraordinary evidence. That's a brutal accusation! There's plenty to say about using the decoder to limit access and being greedy! They really did get the technical end right on this though. The decoded results match the studio master and there's no audio stem separation or de-mixing going on in any way with this system. Object audio is delivered in the metadata.

Un-mix a cake and get the original egg back. A chemist might be able to take that pretty far... Might even be able to reconstruct an "egg" with all the right elements. It wouldn't be the exact shape, detail, etc of the unique original in the end. Those elements of nuance are still irretrievable. Un-mixing audio is the same deal right now. Full fidelity means to get the original egg back in all its uniqueness. Building a system that relied on un-mixing would be novelty indeed!

There are a lot of variables in a 12 channel sound system! Just bumping the volume control on one of your amps can throw a whole mix out of whack! The roots of the format itself are solid. This isn't 5.1 or 7.1 with a gimmick. These are full 7.1.4 mixes delivered in full fidelity.

 

[AYanguas]

I was under the impression that if your speaker layout was 5.1 and not 7.1 or greater, the Atmos decoding option wasn't available?

Yes, I can confirm this with tests done on my Denon 8500

 

[himey]

Yes, I can confirm this with tests done on my Denon 8500

Thanks. I believe sadly, most are like that, and the Arcam is the outlier.

 

[SeeMoreDigital]

Personally, if I had the choice I wouldn't opt to play an Atmos encode on a 5.1 channel set-up, I would play the disc's dedicated 5.1 channel mix. Which is usually encoded in DTS-HD MA...

 

[LuvMyQuad]

In the file as metadata! Not mixed into the bed audio.

Then please explain to me how I can play an Atmos 7.1 file or disk though equipment that was built well before Atmos came along, which simply ignores all the metadata. and not lose any of the signal.

 

[LuvMyQuad]

Personally, if I had the choice I wouldn't opt to play an Atmos encode on a 5.1 channel set-up, I would play the disc's dedicated 5.1 channel mix. Which is usually encoded in DTS-HD MA...

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

 

[himey]

Personally, if I had the choice I wouldn't opt to play an Atmos encode on a 5.1 channel set-up, I would play the disc's dedicated 5.1 channel mix. Which is usually encoded in DTS-HD MA...

It always depends on the mix, not the format. Also, many times you don't have a choice especially for streaming only titles.

 

[Owen Smith]

He is playing back as 5.1, so his receiver should mix the backs into the sides, not ignoring them as the person I quoted suggested.

Indeed, it's the same mix down as when playing back any True HD 7.1 material on a 5.1 system.

Note however that throwing away the backs is the correct process to play back DTS (any variant) 7.1 on a 5.1 system. This difference between DTS and True HD 7.1 to 5.1 handling causes much confusion. The other oddity being that when playing back DTS on an actual 7.1 system, the processor has to subtract the backs from the sides since they're already in there. This comes from DTS being "core plus extensions", the DTS core is 5.1 max and all sounds must be in it.

 

[jimfisheye]

Then please explain to me how I can play an Atmos 7.1 file or disk though equipment that was built well before Atmos came along, which simply ignores all the metadata. and not lose any of the signal.

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.

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!