Video: Why Dolby Atmos Music at Home Is Always Worse Than In The Studio (No Matter What)

[JediJoker]

YouTube immersive audio evangelist Michael G. Wagner gives a breakdown of the differences between Atmos as it exists in the studio and in the home. Unfortunately, the only "home" format he explores is the common streaming Dolby Digital Plus with Joint Object Coding (DD+JOC) lossy encode at 768kbps. I posted a comment asking about lossless TrueHD on Blu-ray, to which I hope he will reply.

 

[JediJoker]

I posted a comment asking about lossless TrueHD on Blu-ray, to which I hope he will reply.

Someone else replied, and Michael co-signed:

 

[humprof]

Well, fine, but . . . how many studios have more than 7.1.4 (or maybe 9.1.4 or at most 11.1.4) speakers to render the positions of those 128 potential objects? And how many mixers are really being that precise in their placement of however many objects they're actually using on a given track (which placement will, in real life, be a function of whatever playback rig they're using)?

I mean . . . for the vast majority of us, the stereo music we listen to on our home systems is an inexact, and probably "inferior," replica of what the mixers heard in their studio, too. And?

 

[sjcorne]

Someone else replied, and Michael co-signed:

I would challenge these people to re-render a 7.1.4 (or 9.1.6) .WAV from their ADM source, then print another such 12 or 16-channel PCM file from a TrueHD encode of that same material through the Dolby Reference Player. Stick both of these in a Pro Tools session and toggle between them using the X-OR solo mode. The differences are imperceptible.

Also not really sure why the '128 objects' is such a big talking point. It's rare even half of them get used for a music mix, especially since the use of 'object beds' is becoming an increasingly common workflow.

 

[CINERAMAX]

Maybe that is why concert atmos is so sucky (the natural immersive reverb of the venue). Snarky Puppy aside.

 

[jimfisheye]

BS! I can A/B the mix coming off the DAW with the encoded TrueHD+Atmos in the Dolby reference player and they sound exactly the same. It's only the lossy version currently used for streaming that doesn't work very well and should be avoided. If the mix doesn't sound right at home... well, it would have sounded just like that in the studio. Maybe do something about that instead of making excuses after. Live music sounds great mixed to 12 channels too! (FYI, the encoder can't tell it's live.) What started this gaslighting?

 

[jimfisheye]

Also not really sure why the '128 objects' is such a big talking point.

Because people think they're listening to "objects" now. A rare new kind of audio that sounds magical! Or something...

Apparently some don't realize it's simply the Dolby renderer finishing the panning for the mix. That is being monitored in 7.1.4. And meant to be played back in 7.1.4 for the most 1:1 experience. There's no consumer way to listen to raw isolated object channels. They are an intermediate part in this "split production system" that is transparent to the listener. Just like the kick drum mic from the multitrack recording can't be solo'd in the mix listening at home. The multitrack recording could be called "raw mix objects" using the same vernacular.

Objects are mix elements delivered to the renderer separately so they can be scaled up to larger theater arrays. The listener hears the final mix. Delivery to the same speaker array it was monitored on in the studio is 1:1. Up or down mixing is hopefully close and acceptable and often is.

Don't let clowns like that gaslight you folks! We're still making mixes to deliver to the listener 1:1 and you're listening to just that. Panning a mix element as an Atmos object vs with the panner in the DAW comes out exactly the same in the final mix. As it is designed to.

 

[MidiMagic]

So how do I get this mix delivered 1:1? What form is on a DVD or a Blu-Ray?

 

[JediJoker]

I would challenge these people to re-render a 7.1.4 (or 9.1.6) .WAV from their ADM source, then print another such 12 or 16-channel PCM file from a TrueHD encode of that same material through the Dolby Reference Player. Stick both of these in a Pro Tools session and toggle between them using the X-OR solo mode. The differences are imperceptible.

Better yet: try a null test. That will reveal what differences exist, if any. Bonus: print/bounce the result and observe the waveform to see if there are any inaudible differences.

 

[AYanguas]

Well, fine, but . . . how many studios have more than 7.1.4 (or maybe 9.1.4 or at most 11.1.4) speakers to render the positions of those 128 potential objects? And how many mixers are really being that precise in their placement of however many objects they're actually using on a given track (which placement will, in real life, be a function of whatever playback rig they're using)?

I mean . . . for the vast majority of us, the stereo music we listen to on our home systems is an inexact, and probably "inferior," replica of what the mixers heard in their studio, too. And?

The "exact" image location of an object sound, that can be different in the studio and at home, especially if we have different numbers of speakers in different positions, is almost NOT relevant at all.

What really matters is the quality of sound. It is nearly identical in TrueHD (relative to the ADM master file), but is degraded in the DD+ JOC, because of the lossy compression and the "imperfection" of the decoding Atmos algorithm for some musical content, such as the reverbs as the video mentions.

 

[AYanguas]

Maybe that is why concert atmos is so sucky (the natural immersive reverb of the venue). Snarky Puppy aside.

I understand that the degradation decoding algorithm for reverb would apply to DD+JOC, but NOT, or at least less, to TrueHD with Atmos in the concert Blu-rays.

I guess the sucky sound of some Atmos concerts would be due to the recording/mixing process.

 

[jimfisheye]

So how do I get this mix delivered 1:1? What form is on a DVD or a Blu-Ray?

Bluray

The mix is encoded in TrueHD+Atmos.
Rip to mkv or mka to pass to hardware decoding devices.
Extract the mlp from mkv to use the Dolby reference player.

You need to know what system the mix was made on (eg 7.1.4) and set your media player to that same speaker array. (It helps to have said speaker array.) That's what you might find weird about the format. Set a different speaker array and it starts up or downmixing. Some bluray discs omit the mix speaker array detail from the liner notes and that's a point of contention. Assuming 7.1.4 will be correct 99% of the time.

The lossy streaming version is akin to 64k mp3 fidelity. If fidelity comes first for you before more channels if they are compromised, you will likely agree that Atmos "doesn't work" or comes across as a gimmick. The linked article is pretending only lossy streaming Atmos is available to the consumer and choosing not to mention the actual full version that delivers lossless sound in 12 or 16 channel mixes 1:1.

 

[jimfisheye]

I guess the sucky sound of some Atmos concerts would be due to the recording/mixing process.

Yeah, people can make bad mixes in just about any speaker array format! Lossy Atmos is a compromised gimmick from the ground up but it's an aside. You've probably heard a good mix trounced on with 64k mp3 compression vs a lossless delivery of a really bad mix. There are things to discuss in what you hear but you probably still prefer the lossy good mix over the lossless bad mix. Or at least you might have heard some examples that hit that way. Same thing going on with the lossy Atmos. But it really is a compromise and it really is a failure from the perspective of fidelity before more channels for the audiophile surround listener.

 

[Owen Smith]

but you probably still prefer the lossy good mix over the lossless bad mix.

I've experienced that, definitely true.

 

[CINERAMAX]

I cancelled my Apple Music subscription for a reason. I now only use Blu-rays, which are expensive but worth it. It could be worse, as I dislike Church ( showing the birdie to NL) and concert hall Atmos; I prefer studio recordings.

 

[zdjh22]

Listening to my latest batch of discs, all Atmos via TrueHD on BluRay (Ziggy Stardust, Chicago IX, No Secrets, Luck and Strange) I have to say that I’m really jealous of the mixers if the in-studio sound is that much better. Because they are all damn good via my home systems.

Or, much more likely, in-studio is different, not better.

 

[MidiMagic]

- Is the difference audible?

- Which formats have this problem and which don't?

 

[jimfisheye]

There is no audible difference between the encoded Dolby TrueHD+Atmos and the mixing coming off the mixing desk.

The only format with the problem is the lossy streaming Atmos (DD+JOC). Correctly called out as not suitable to deliver 12 channel sound.

The video is insincere, acting as though not aware the full TrueHD format exists.

A misguided attack IMHO. He should be going after them for the gaslighting over the decoder codec. The refusal to license it to software media players to push sales of new hardware AVRs, etc with the decoder hidden in their firmware. Dude is probably connected to some hardware seller. "You aren't hearing Atmos correctly at home." (a lie) "So that's why you need our new product! bla bla bla..."

 

[Owen Smith]

The video is insincere, acting as though not aware the full TrueHD format exists.

I posted a comment pointing out TrueHD, and the video poster replied saying "yes but it has basically the same problem, just in higher quality". He's trying to emphasis the 128 objects max at the mixing desk vs the number of objects consumer Atmos allows. Personally I suspect this is a very minor issue given most stuff is mixed for 7.x.4 and lossy encoding of streaming Atmos is a much more serious problem.

 

[SeeMoreDigital]

It would be useful if somebody who actually creates Atmos muxes was able to comment...