G4 processors Dolby Atmos / DTS:X Pro speaker configurations

[KeithL]

The answer is both simple and complicated.

In general we are required to follow their guidelines as regards decoding.
But we are allowed to then add our own "proprietary additional post-processing" if we like.
(Just like in the old days, Yamaha added "cathedral" and "dark basement" "DSP modes" to their standard surround sound modes.)

The bottom line is that creating new and interesting DSP modes isn't especially difficult...
And neither would it be especially difficult to re-route or duplicate channels...

It is worth noting that, with Dolby Atmos content, the beds and objects are actually parts of the content.
So whoever made the content gets to decide things like how many beds they want to use.
So, while the decoder can make decisions, its options are limited to what's present in the content itself.

The VERY general guidance for mastering reads (I'm paraphrasing a bit here):
"how beds are rendered will depend on your speaker configuration"
"if you want precise control over how and where something will render then you should use an object"

Also... from Dolby... and a direct quote this time:
"Depending upon the position and size metadata applied to an object, objects and bed channels can be sonically identical.
For instance, an object placed in the left front with size set to zero will be identical to placing the audio in the Left channel bed."
(Note that, while they say "CAN be sonically identical" this is not always the case with different channels in different configurations.)


If you want to see Dolby's "guidance" on when to use beds or objects... :
professionalsupport.dolby.com/s/article/What-s-the-difference-between-beds-and-objects?language=en_US
professionalsupport.dolby.com/s/article/How-many-loudspeakers-do-I-need-to-mix-content-in-Dolby-Atmos-for-Home-Entertainment?language=en_US

And finally...   Dolby Atmos Home Entertainment Studio Technical Guidelines (really interesting):
dolby.my.salesforce.com/sfc/p/700000009YuG/a/4u000000lFHc/UYA0IZeD632SUXVmEPmUcr.wIuhpHp6Q7bVSl4LrbUQ
(In a total of 45 rather densely packed pages see if you can find a single direct reference to multiple listener seats...  )

Oct 23, 2024 14:44:52 GMT -5 lrobertson said:

www.youtube.com/live/L4n1zf2Y8vc?si=tDaYAp6SJIdAOEeV

22 minutes in. It sounds like they utilize the native dolby decoder for the 11.1.x playback. I guess a very relevant question is how much of the processing is dictated by dolby vs how much your Emotiva engineers have to implement dolby's "guidelines"? Hopefully if it's the ladder this can make it to Lonnie. I would assume it's easier to implement due to the channel expansion only focusing on 4 additional speakers that happen to be associated to bed layers.

[lrobertson]

Oct 23, 2024 15:26:52 GMT -5 KeithL said:

The answer is both simple and complicated.

In general we are required to follow their guidelines as regards decoding.
But we are allowed to then add our own "proprietary additional post-processing" if we like.
(Just like in the old days, Yamaha added "cathedral" and "dark basement" "DSP modes" to their standard surround sound modes.)

The bottom line is that creating new and interesting DSP modes isn't especially difficult...
And neither would it be especially difficult to re-route or duplicate channels...

It is worth noting that, with Dolby Atmos content, the beds and objects are actually parts of the content.
So whoever made the content gets to decide things like how many beds they want to use.
So, while the decoder can make decisions, its options are limited to what's present in the content itself.

The VERY general guidance for mastering reads (I'm paraphrasing a bit here):
"how beds are rendered will depend on your speaker configuration"
"if you want precise control over how and where something will render then you should use an object"

Also... from Dolby... and a direct quote this time:
"Depending upon the position and size metadata applied to an object, objects and bed channels can be sonically identical.
For instance, an object placed in the left front with size set to zero will be identical to placing the audio in the Left channel bed."
(Note that, while they say "CAN be sonically identical" this is not always the case with different channels in different configurations.)

Oct 23, 2024 14:44:52 GMT -5 lrobertson said:

www.youtube.com/live/L4n1zf2Y8vc?si=tDaYAp6SJIdAOEeV

22 minutes in. It sounds like they utilize the native dolby decoder for the 11.1.x playback. I guess a very relevant question is how much of the processing is dictated by dolby vs how much your Emotiva engineers have to implement dolby's "guidelines"? Hopefully if it's the ladder this can make it to Lonnie. I would assume it's easier to implement due to the channel expansion only focusing on 4 additional speakers that happen to be associated to bed layers.

If Lonnie gives me bed layers in the expansion channels when the beds are present I'll be sure to yell at the audio engineers and not you guys for a bad atmos mix haha. Thanks Keith.

[lrobertson]

Oct 23, 2024 15:26:52 GMT -5 KeithL said:

The answer is both simple and complicated.

In general we are required to follow their guidelines as regards decoding.
But we are allowed to then add our own "proprietary additional post-processing" if we like.
(Just like in the old days, Yamaha added "cathedral" and "dark basement" "DSP modes" to their standard surround sound modes.)

The bottom line is that creating new and interesting DSP modes isn't especially difficult...
And neither would it be especially difficult to re-route or duplicate channels...

It is worth noting that, with Dolby Atmos content, the beds and objects are actually parts of the content.
So whoever made the content gets to decide things like how many beds they want to use.
So, while the decoder can make decisions, its options are limited to what's present in the content itself.

The VERY general guidance for mastering reads (I'm paraphrasing a bit here):
"how beds are rendered will depend on your speaker configuration"
"if you want precise control over how and where something will render then you should use an object"

Also... from Dolby... and a direct quote this time:
"Depending upon the position and size metadata applied to an object, objects and bed channels can be sonically identical.
For instance, an object placed in the left front with size set to zero will be identical to placing the audio in the Left channel bed."
(Note that, while they say "CAN be sonically identical" this is not always the case with different channels in different configurations.)


If you want to see Dolby's "guidance" on when to use beds or objects... :
professionalsupport.dolby.com/s/article/What-s-the-difference-between-beds-and-objects?language=en_US
professionalsupport.dolby.com/s/article/How-many-loudspeakers-do-I-need-to-mix-content-in-Dolby-Atmos-for-Home-Entertainment?language=en_US

And finally...   Dolby Atmos Home Entertainment Studio Technical Guidelines (really interesting):
dolby.my.salesforce.com/sfc/p/700000009YuG/a/4u000000lFHc/UYA0IZeD632SUXVmEPmUcr.wIuhpHp6Q7bVSl4LrbUQ
(In a total of 45 rather densely packed pages see if you can find a single direct reference to multiple listener seats...  )

Oct 23, 2024 14:44:52 GMT -5 lrobertson said:

www.youtube.com/live/L4n1zf2Y8vc?si=tDaYAp6SJIdAOEeV

22 minutes in. It sounds like they utilize the native dolby decoder for the 11.1.x playback. I guess a very relevant question is how much of the processing is dictated by dolby vs how much your Emotiva engineers have to implement dolby's "guidelines"? Hopefully if it's the ladder this can make it to Lonnie. I would assume it's easier to implement due to the channel expansion only focusing on 4 additional speakers that happen to be associated to bed layers.

Page 27 sure seems like they are supposed to operate as bed arrays... page 27-34 are all relevant. I think focusing on the practical use of how it will make multiple rows more efficient isn't as important as seeing it's clearly bed arrays. Labeled as such. Which happens to work better with more rows*

"Panning a sound to the surround bed reproduces the sound in all speakers of the array" Same with tops.

"Where wider sound coverage is needed for specific seating positions"..

That sounds like referencing multiple rows in a round about way.

I really like you guys at Emotiva but just so we are clear after reading this I really expect bed channels to play through the expansion channels... Any other playback is ignoring those relevant pages and not conforming to their guidelines and it wouldn't technically be 11.1.8 atmos decoding. I'm happy to hear a reference to correct this but after reading it I honestly think you guys have a problem if the coding is done.

It sounds to me like the dolby representative is terrible. They should be more like nvidia.

Farming is going to sh** in California. I want to be part of the team haha. Or put in a word for me at dolby. That's local.

I'd also like to reiterate that it seems fair to allow for a later trade in period for a gen 4 system with expansion modules installed so we don't have to trade in our units till things are ironed out.

[PaulBe]

Oct 23, 2024 15:26:52 GMT -5 KeithL said:

The answer is both simple and complicated.

In general we are required to follow their guidelines as regards decoding.
But we are allowed to then add our own "proprietary additional post-processing" if we like.
(Just like in the old days, Yamaha added "cathedral" and "dark basement" "DSP modes" to their standard surround sound modes.)

The bottom line is that creating new and interesting DSP modes isn't especially difficult...
And neither would it be especially difficult to re-route or duplicate channels...

It is worth noting that, with Dolby Atmos content, the beds and objects are actually parts of the content.
So whoever made the content gets to decide things like how many beds they want to use.
So, while the decoder can make decisions, its options are limited to what's present in the content itself.

The VERY general guidance for mastering reads (I'm paraphrasing a bit here):
"how beds are rendered will depend on your speaker configuration"
"if you want precise control over how and where something will render then you should use an object"

Also... from Dolby... and a direct quote this time:
"Depending upon the position and size metadata applied to an object, objects and bed channels can be sonically identical.
For instance, an object placed in the left front with size set to zero will be identical to placing the audio in the Left channel bed."
(Note that, while they say "CAN be sonically identical" this is not always the case with different channels in different configurations.)


If you want to see Dolby's "guidance" on when to use beds or objects... :
professionalsupport.dolby.com/s/article/What-s-the-difference-between-beds-and-objects?language=en_US
professionalsupport.dolby.com/s/article/How-many-loudspeakers-do-I-need-to-mix-content-in-Dolby-Atmos-for-Home-Entertainment?language=en_US

And finally...   Dolby Atmos Home Entertainment Studio Technical Guidelines (really interesting):
dolby.my.salesforce.com/sfc/p/700000009YuG/a/4u000000lFHc/UYA0IZeD632SUXVmEPmUcr.wIuhpHp6Q7bVSl4LrbUQ
(In a total of 45 rather densely packed pages see if you can find a single direct reference to multiple listener seats...  )

Oct 23, 2024 14:44:52 GMT -5 lrobertson said:

www.youtube.com/live/L4n1zf2Y8vc?si=tDaYAp6SJIdAOEeV

22 minutes in. It sounds like they utilize the native dolby decoder for the 11.1.x playback. I guess a very relevant question is how much of the processing is dictated by dolby vs how much your Emotiva engineers have to implement dolby's "guidelines"? Hopefully if it's the ladder this can make it to Lonnie. I would assume it's easier to implement due to the channel expansion only focusing on 4 additional speakers that happen to be associated to bed layers.

I think object size and position can be distorted by bad speaker position. Dolby has a large range of acceptable HT speaker positions. Objects may appear smaller or larger than they were actually recorded, and in slightly different positions than intended.  Now, add moving object distortion because of bad speaker placement.

BTW - we have had object placement control for decades with analog stereo audio consoles. The controls are called Pan and Width. 

[PaulBe]

Oct 23, 2024 15:26:52 GMT -5 KeithL said:

The answer is both simple and complicated.

In general we are required to follow their guidelines as regards decoding.
But we are allowed to then add our own "proprietary additional post-processing" if we like.
(Just like in the old days, Yamaha added "cathedral" and "dark basement" "DSP modes" to their standard surround sound modes.)

The bottom line is that creating new and interesting DSP modes isn't especially difficult...
And neither would it be especially difficult to re-route or duplicate channels...

It is worth noting that, with Dolby Atmos content, the beds and objects are actually parts of the content.
So whoever made the content gets to decide things like how many beds they want to use.
So, while the decoder can make decisions, its options are limited to what's present in the content itself.

The VERY general guidance for mastering reads (I'm paraphrasing a bit here):
"how beds are rendered will depend on your speaker configuration"
"if you want precise control over how and where something will render then you should use an object"

Also... from Dolby... and a direct quote this time:
"Depending upon the position and size metadata applied to an object, objects and bed channels can be sonically identical.
For instance, an object placed in the left front with size set to zero will be identical to placing the audio in the Left channel bed."
(Note that, while they say "CAN be sonically identical" this is not always the case with different channels in different configurations.)


If you want to see Dolby's "guidance" on when to use beds or objects... :
professionalsupport.dolby.com/s/article/What-s-the-difference-between-beds-and-objects?language=en_US
professionalsupport.dolby.com/s/article/How-many-loudspeakers-do-I-need-to-mix-content-in-Dolby-Atmos-for-Home-Entertainment?language=en_US

And finally...   Dolby Atmos Home Entertainment Studio Technical Guidelines (really interesting):
dolby.my.salesforce.com/sfc/p/700000009YuG/a/4u000000lFHc/UYA0IZeD632SUXVmEPmUcr.wIuhpHp6Q7bVSl4LrbUQ
(In a total of 45 rather densely packed pages see if you can find a single direct reference to multiple listener seats...  )

Oct 23, 2024 14:44:52 GMT -5 lrobertson said:

www.youtube.com/live/L4n1zf2Y8vc?si=tDaYAp6SJIdAOEeV

22 minutes in. It sounds like they utilize the native dolby decoder for the 11.1.x playback. I guess a very relevant question is how much of the processing is dictated by dolby vs how much your Emotiva engineers have to implement dolby's "guidelines"? Hopefully if it's the ladder this can make it to Lonnie. I would assume it's easier to implement due to the channel expansion only focusing on 4 additional speakers that happen to be associated to bed layers.

Thanks for the Dolby links. Dolby has a lot of new papers, guides, and teaching material.

[lrobertson]

Oct 24, 2024 9:53:37 GMT -5 PaulBe said:

Any new information about G4 expansion modules would be appreciated. My interest is in a Sub expansion module. Is it ok to say this at the G4 thread?

How dare you haha.

I'd also appreciate a little more focus on the bed array info from the dolby home atmos papers.

I do wonder with all this focus on coding and the difficulties and costs if there would be a way to make the layout code layer open source. I know I've spent time on research as a hobby. I wonder if there is a way to allow hobbiests to take all expansion modules possible and work in different layout options. Maybe even cross platforms dts and dolby etc. Probably too complex and could create speaker damage if done wrong but just a curiosity on my part. At some point you would think ai could take it and run. Making this work alongside dirac I could see as a complex endeavor as well.

[KeithL]

Yes... and sort of... and pretty much good enough.
Although EXACTLY what that "width" control actually did is more complicated than you probably think.
(If you're panning a stereo source then you can adjust the separation by mixing the channels together.
However, if you're starting out with a monaural track, which has "zero width", you would need to "synthesize some width" before you can "adjust" it with that knob.)

Panning things between left and right by adjusting their relative levels worked well enough for stereo.
Although, in technical terms, it really doesn't work well at all... but, luckily for use, our brains ignore most of the parts that are wrong.

I'll digress a bit and explain...
Let's assume you have an actual stage; it's twenty feet wide; and stretches across the entire front end of the room.
And we have a lead singer who is the sort who walks back and forth a lot as he sings...
And let's make it easy and assume we're sitting front row center...
Now remember that sound travels about 1000 feet per second; so there is a delay of about one millisecond per foot as it travels through the air.
As that singer walks from left to right the amount of his voice we hear in our left and right ears will vary...
And that will be the "primary cue" our brain uses to tell us where he is...
And that is what we're adjusting with out pan pot.
BUT there is a whole bunch of "other stuff" going on....
For example, when he walks to our left, his voice will be louder in our left ear...
But his voice will also arrive directly at our left ear sooner...
And the reflection of his voice we hear as it bounces off the left wall will also arrive sooner...
While the reflection we hear of his voice from the RIGHT wall will arrive significantly LATER...
And, to make things more complicated, we won't notice shorter delays, but delays above a certain threshold will be audible.

Now imagine what will happen if we record him with a nice microphone pinned directly to his lapel...
And then use that "pan pot" to move him from left to right...
The LEVELS will be exactly what they would be if he actually walked from stage right to stage left...
But the DELAYS, and the delays from the reflections from the two walls, will be entirely WRONG.
And that pan pot ONLY adjusts the levels (at least on that old analog console)...
(And, since we used a lapel mic, we probably don't have any real wall reflections anyway, so you're probably hearing fake ones that were added later in the mix.)
LUCKILY FOR US, in this situation, our brains focus on that primary cue (level) and ignore the discrepancies with those other cues (the delays).
(And modern reverb plugins can make more realistic fake reverb that actually does take the dimensions of the "imaginary" room into account.)

But, with lots of channels, and lots of speakers, these other cues become more important if you want something to "sound natural and real". 
(The proof of that is that things won't sound right if your speakers are at different distances from the listener and the adjustment for that is too far off.)

The other thing that makes this example a lot simpler is that our singer is always going to be a point source.
We're never going to be in a situation where "we want him to stay in the same place but get BIGGER or SMALLER"...
"Bigger and smaller" is NOT the same as "louder and quieter".
(He's not a giant so we're never going to need to make it sound like he's singing from a mouth that's twenty feet wide.)
And, if you did want to do that. how would you DO it with a pan pot? (The answer is that you cannot.)

And this is something that you very well might want to do with an object in a movie sound track... like a big-ass spaceship landing on top of you.
(In fact, in that Atmos mastering app, you can simply specify that the SIZE of that object gets larger.)

And, in order to do that, you can't simply play that sound from more speakers.
If you did that then it would just be a point source that seemed to come from somewhere in the middle of all the speakers playing it.
What you're going to have to do is to somehow "shred the sound apart"... so that slightly different versions or parts of it are coming from each of those speakers.
And you're going to have to do this in a way that your sound "still sounds the same but like it's coming from a larger source".
As I said, in the mastering app, you can simply set a "size" for the object... but I don't know what trickery they use to "make an object spread out bigger"...
(The mastering engineer can treat this like just another knob on the mixer... but how it works internally is one of those "proprietary secret sauce" things.)

Oct 23, 2024 15:54:44 GMT -5 PaulBe said:

Oct 23, 2024 15:26:52 GMT -5 KeithL said:

I think object size and position can be distorted by bad speaker position. Dolby has a large range of acceptable HT speaker positions. Objects may appear smaller or larger than they were actually recorded, and in slightly different positions than intended.  Now, add moving object distortion because of bad speaker placement.

BTW - we have had object placement control for decades with analog stereo audio consoles. The controls are called Pan and Width. 

[marcl]

We really need a dedicated Atmos technology thread to talk about all aspects of Atmos production and delivery.

But while I'm here, a couple things .... I'm an AES member so I can read papers and see local chapter meetings that sometimes stream from various parts of the world. Over the past couple years I've attended lots of these sessions and also video sessions that Dolby puts on specific to Atmos mixing. What I conclude so far is that virtually NONE of the mixing engineers/producers speaks in a way that would indicate that they understand and use objects vs channels.  They talk a lot about turning knobs until it sounds good to them.   They may say the word object, but when they describe the way they mix, it sounds like channels and in their head they're still thinking "I want to put this sound in that speaker or between those speakers" ... not "I want to put this sound at this location in 3D space". In fact, one of the proverbial Grammy Nominated/Winning guys actually chuckled with a somewhat guilty expression and said "ya know what? I don't use objects."

And ... every one said they have to be conscious of how it will render on earbuds.

And ... in every case where they discussed Dolby certified mixing studios, the setup was 7.1.4 per Dolby spec.

[ttocs]

Oct 24, 2024 10:47:49 GMT -5 marcl said:

And ... in every case where they discussed Dolby certified mixing studios, the setup was 7.1.4 per Dolby spec.

This. This is what the default spec is from Dolby. 7.1.4. This is what Dolby starts with in ATMOS guides for professionals, and is what Dolby recommends for mixing setups.

[lrobertson]

Oct 24, 2024 10:47:49 GMT -5 marcl said:

We really need a dedicated Atmos technology thread to talk about all aspects of Atmos production and delivery.

But while I'm here, a couple things .... I'm an AES member so I can read papers and see local chapter meetings that sometimes stream from various parts of the world. Over the past couple years I've attended lots of these sessions and also video sessions that Dolby puts on specific to Atmos mixing. What I conclude so far is that virtually NONE of the mixing engineers/producers speaks in a way that would indicate that they understand and use objects vs channels.  They talk a lot about turning knobs until it sounds good to them.   They may say the word object, but when they describe the way they mix, it sounds like channels and in their head they're still thinking "I want to put this sound in that speaker or between those speakers" ... not "I want to put this sound at this location in 3D space". In fact, one of the proverbial Grammy Nominated/Winning guys actually chuckled with a somewhat guilty expression and said "ya know what? I don't use objects."

And ... every one said they have to be conscious of how it will render on earbuds.

And ... in every case where they discussed Dolby certified mixing studios, the setup was 7.1.4 per Dolby spec.

I think as systems are built to purpose around that atmos paper then eventually it will become more common knowledge and we may see better mixes into the future that would utilize arrays correctly. I'm happy to discuss that on a focused thread but this will be the thread I'll come back to when wanting to know if this system is 11.1.8 atmos or watered down. Two wrongs won't make a right. We can at least demand our playback in the g4 be to dolby spec.

[lrobertson]

Oct 24, 2024 10:53:02 GMT -5 ttocs said:

Oct 24, 2024 10:47:49 GMT -5 marcl said:

And ... in every case where they discussed Dolby certified mixing studios, the setup was 7.1.4 per Dolby spec.

This. This is what the default spec is from Dolby. 7.1.4. This is what Dolby starts with in ATMOS guides for professionals, and is what Dolby recommends for mixing setups.

This is true but their choice in mixing layout shouldn't affect 11.1.8 expansion in anything but a good way if they utilize the correct principals outlined in the paper at least theoretically.

If they have the ability to playback without any objects and just beds they should never hear anything that plays like an object in their surround, surround back or top beds.

[PaulBe]

Oct 24, 2024 10:16:35 GMT -5 KeithL said:

Yes... and sort of... and pretty much good enough.
Although EXACTLY what that "width" control actually did is more complicated than you probably think.
(If you're panning a stereo source then you can adjust the separation by mixing the channels together.
However, if you're starting out with a monaural track, which has "zero width", you would need to "synthesize some width" before you can "adjust" it with that knob.)

Panning things between left and right by adjusting their relative levels worked well enough for stereo.
Although, in technical terms, it really doesn't work well at all... but, luckily for use, our brains ignore most of the parts that are wrong.

I'll digress a bit and explain...
Let's assume you have an actual stage; it's twenty feet wide; and stretches across the entire front end of the room.
And we have a lead singer who is the sort who walks back and forth a lot as he sings...
And let's make it easy and assume we're sitting front row center...
Now remember that sound travels about 1000 feet per second; so there is a delay of about one millisecond per foot as it travels through the air.
As that singer walks from left to right the amount of his voice we hear in our left and right ears will vary...
And that will be the "primary cue" our brain uses to tell us where he is...
And that is what we're adjusting with out pan pot.
BUT there is a whole bunch of "other stuff" going on....
For example, when he walks to our left, his voice will be louder in our left ear...
But his voice will also arrive directly at our left ear sooner...
And the reflection of his voice we hear as it bounces off the left wall will also arrive sooner...
While the reflection we hear of his voice from the RIGHT wall will arrive significantly LATER...
And, to make things more complicated, we won't notice shorter delays, but delays above a certain threshold will be audible.

Now imagine what will happen if we record him with a nice microphone pinned directly to his lapel...
And then use that "pan pot" to move him from left to right...
The LEVELS will be exactly what they would be if he actually walked from stage right to stage left...
But the DELAYS, and the delays from the reflections from the two walls, will be entirely WRONG.
And that pan pot ONLY adjusts the levels (at least on that old analog console)...
(And, since we used a lapel mic, we probably don't have any real wall reflections anyway, so you're probably hearing fake ones that were added later in the mix.)
LUCKILY FOR US, in this situation, our brains focus on that primary cue (level) and ignore the discrepancies with those other cues (the delays).
(And modern reverb plugins can make more realistic fake reverb that actually does take the dimensions of the "imaginary" room into account.)

But, with lots of channels, and lots of speakers, these other cues become more important if you want something to "sound natural and real". 
(The proof of that is that things won't sound right if your speakers are at different distances from the listener and the adjustment for that is too far off.)

The other thing that makes this example a lot simpler is that our singer is always going to be a point source.
We're never going to be in a situation where "we want him to stay in the same place but get BIGGER or SMALLER"...
"Bigger and smaller" is NOT the same as "louder and quieter".
(He's not a giant so we're never going to need to make it sound like he's singing from a mouth that's twenty feet wide.)
And, if you did want to do that. how would you DO it with a pan pot? (The answer is that you cannot.)

And this is something that you very well might want to do with an object in a movie sound track... like a big-ass spaceship landing on top of you.
(In fact, in that Atmos mastering app, you can simply specify that the SIZE of that object gets larger.)

And, in order to do that, you can't simply play that sound from more speakers.
If you did that then it would just be a point source that seemed to come from somewhere in the middle of all the speakers playing it.
What you're going to have to do is to somehow "shred the sound apart"... so that slightly different versions or parts of it are coming from each of those speakers.
And you're going to have to do this in a way that your sound "still sounds the same but like it's coming from a larger source".
As I said, in the mastering app, you can simply set a "size" for the object... but I don't know what trickery they use to "make an object spread out bigger"...
(The mastering engineer can treat this like just another knob on the mixer... but how it works internally is one of those "proprietary secret sauce" things.)

Oct 23, 2024 15:54:44 GMT -5 PaulBe said:

I think object size and position can be distorted by bad speaker position. Dolby has a large range of acceptable HT speaker positions. Objects may appear smaller or larger than they were actually recorded, and in slightly different positions than intended.  Now, add moving object distortion because of bad speaker placement.

BTW - we have had object placement control for decades with analog stereo audio consoles. The controls are called Pan and Width. 

Yes. I do know about the other stuff. The Width control in an audio console is also used to control the pattern in M-S recording. My experience in pro audio is a little more extensive and complicated than you think it is. 

I did some stereo M-S recording in the studio with my Shure KSM-44 multi-pattern large diaphragm condenser mics -  using the Neve BTC audio console that I installed and commissioned - in the audio suite that I built and integrated with 4 studio locations in the House - 

I was trying to keep the conversation at a layman level for the sake of the general audience here – and make it germane to the discussion of accurate object placement in reproduction, and how the tool is not new.

The tool is obviously most sophisticated when used in 3D (or can be) – like with Atmos. And, the tool can include Pan, Width, Time, Intensity, and Phase, in all stereo recordings from 2 channels to Infinity. It can be done at the microphone system level. It can be done at the post production level.

Listen to an LP of Tomita’s Planets, on a good system, and marvel at the object placement effects coming from a 50 year old 2 channel, vinyl LP. It’s a mind blower to people who think Object effects were invented by Dolby Labs.

I assume the production sound stage meets the standards for speaker position layout. The latest Dolby papers still show some allowable deviation from a fixed standard. I’m not surprised. Dolby wants to sell licenses.

The latest Dolby HT layouts show some allowable compromise from a fixed standard that is different from that used on the sound stage, even when the respective x.x.x speaker layouts are the same position count.  I.E., the source production layout and destination layout standards don’t match; and the allowable source production deviations don’t match the allowable destination layout compromises. The standards change over time. My current recordings do not change. I won't be moving built-in height/ceiling speaker positions to accept 'enhanced' Atmos - though perhaps 'enhanced' Atmos includes new object information in the base bed channels.

The production details you are adding are part of the Art. Some artists and engineers are better at using the technical tools to create the Art they want. Sound that is ‘natural and real’ May be the goal. But, art is not always a Portrait, Landscape, or Still Life. Sometimes it’s Picasso; or Jackson Pollock; a scene from nature; or a surreal trip. I want interpretative accuracy across the full spectrum of productions that is true to the Art.

I want to hear the Art as intended; Reproduced as best I can. Some people want to enhance the reproduction, or don’t care as much about reproduction as I care about it.

Object information is very fragile. It’s just Not a new thing. Approved layout deviations and compromises sell licenses and product – and also mess with the Art in unpredictable ways. Deviations at the production end are unacceptable. Compromises at the reproduction end are unavoidable.

[PaulBe]

Oct 24, 2024 10:53:02 GMT -5 ttocs said:

Oct 24, 2024 10:47:49 GMT -5 marcl said:

And ... in every case where they discussed Dolby certified mixing studios, the setup was 7.1.4 per Dolby spec.

This. This is what the default spec is from Dolby. 7.1.4. This is what Dolby starts with in ATMOS guides for professionals, and is what Dolby recommends for mixing setups.

How many loudspeakers do I need to mix content in Dolby Atmos for Home-Entertainment?

May 18, 2021 Knowledge

Dolby recommends at least a 5.1.4 speaker layout to mix content in Dolby Atmos for Home Entertainment:

5 speakers on the listener plane (left, center, right, one surround pair)

1 subwoofer

4 overhead speakers (left top front, right top front, left top rear, right top rear)


A 7.1.4 speaker layout is the preferred minimum setup:

7 speakers on the listener plane (left, center, right, one side-surround pair, one rear-surround pair)

1 subwoofer

4 overhead speakers (left top front, right top front, left top rear, right top rear)


In some larger studios a 9.1.6 speaker layout can be found:

9 speakers on the listener plane (Left, Center, Right, a wide-surround pair, a side-surround pair and a rear-surround pair)

1 subwoofer

6 overhead speakers (left top front, right top front, left top middle, right top middle, left top rear, right top rear)

professionalsupport.dolby.com/s/article/How-many-loudspeakers-do-I-need-to-mix-content-in-Dolby-Atmos-for-Home-Entertainment?language=en_US

[lrobertson]

It makes sense the recommended minimum would allow for all 10 bed channels. I think the disconnect here is going to be that minimum recommended mixing layout does not mean they are supposed to focus on effective playback from only that layout. Objects will be known to be utilized in a 9.1.6 to them without having to have a 9.1.6 ti hear the wides. Beds will be known to operate in arrays to them with 11.1.8 or 13.1.10 without the need to have those layouts etc.

The mixing studio layout chosen should have little to no effect on proper audio mixing. Expanding the layout should only expand the arrays and reorient the object playback. The notion this was only a focus on the cinematic mix i think is clearly false after reading the home audio dolby paper.

[PaulBe]

Oct 24, 2024 11:09:28 GMT -5 lrobertson said:

Oct 24, 2024 10:53:02 GMT -5 ttocs said:

This. This is what the default spec is from Dolby. 7.1.4. This is what Dolby starts with in ATMOS guides for professionals, and is what Dolby recommends for mixing setups.

This is true but their choice in mixing layout shouldn't affect 11.1.8 expansion in anything but a good way if they utilize the correct principals outlined in the paper at least theoretically.

If they have the ability to playback without any objects and just beds they should never hear anything that plays like an object in their surround, surround back or top beds.

The short answer is this is not true. 

[lrobertson]

Oct 24, 2024 12:23:08 GMT -5 PaulBe said:

Oct 24, 2024 11:09:28 GMT -5 lrobertson said:

This is true but their choice in mixing layout shouldn't affect 11.1.8 expansion in anything but a good way if they utilize the correct principals outlined in the paper at least theoretically.

If they have the ability to playback without any objects and just beds they should never hear anything that plays like an object in their surround, surround back or top beds.

The short answer is this is not true. 

If something plays like an object then it will not play well with the arrays, this is simply true. I can see if the total number of objects are already taken for the home atmos mix but if not things that sound like they should be objects should be objects. The way they balance this is their decided tradeoff..

I'm of course assuming we can't keep ignoring the beds are arrays per the dolby official home mixing guidelines..

To your point I'm basically saying 2 wrongs don't make a right. You're correct in many cases but they are wrong in those cases for proper mixing per the guidelines. This is America and people are free to cut corners or do things improperly. I don't think Emotiva is technically allowed to call it 11.1.8 atmos without the arrays though without some serious concern. This is my take.

This could very well be why you preferred the 2 tops instead of your 6 from the past. Even our current processor should use the top beds as arrays. I'm not sure how they are currently mapped. I don't know if it was your source material, your preference for a clearer separation of audio, or maybe just improper audio extraction... But I am curious. I don't mean this with any disrespect but I don't have confidence we've really been experiencing true atmos anymore.. I'm not trying to unnecessarily stir the pot but instead solve what I think is a problem that may have taken root on gen 3.

[lrobertson]

Our current systems the bed layer stuff is relevant for 5.1.4, 7.1.4, 7.1.6, 9.1.4 to 9.1.6. How the top bed arrays to the top soeakers. We can focus on any of these layouts unless we want a better understanding of all beds as we get larger layouts. No trolling from me at least.

[PaulBe]

Oct 24, 2024 13:20:56 GMT -5 lrobertson said:

Our current systems the bed layer stuff is relevant for 5.1.4, 7.1.4, 7.1.6, 9.1.4 to 9.1.6. How the top bed arrays to the top soeakers. We can focus on any of these layouts unless we want a better understanding of all beds as we get larger layouts. No trolling from me at least.

We have 10 beds with Dolby Atmos. That is it.

[lrobertson]

Oct 24, 2024 13:34:49 GMT -5 PaulBe said:

Oct 24, 2024 13:20:56 GMT -5 lrobertson said:

Our current systems the bed layer stuff is relevant for 5.1.4, 7.1.4, 7.1.6, 9.1.4 to 9.1.6. How the top bed arrays to the top soeakers. We can focus on any of these layouts unless we want a better understanding of all beds as we get larger layouts. No trolling from me at least.

We have 10 beds with Dolby Atmos. That is it.

Yes and 2 of those are meant for top arrays. Page 34

Refer to 2.4.2 on page 27

By definition we are creating an array by adding speakers to tops, sides, or rears.

There's no way for me to brush past this one. For our understanding I think abandoning the word bed channels for array channels would be entirely accurate for those bed surrounds, rears and tops. That redefinition might actually help home atmos mixers as well.

[PaulBe]

Oct 24, 2024 13:36:28 GMT -5 lrobertson said:

Oct 24, 2024 13:34:49 GMT -5 PaulBe said:

We have 10 beds with Dolby Atmos. That is it.

Yes and 2 of those are meant for top arrays. Page 34

Refer to 2.4.2 on page 27

By definition we are creating an array by adding speakers to tops, sides, or rears.

There's no way for me to brush past this one. For our understanding I think abandoning the word bed channels for array channels would be entirely accurate for those bed surrounds, rears and tops. That redefinition might actually help home atmos mixers as well.

Initially, I didn't know what reference you were looking at. I think I found it:

For complex designs including loudspeaker layouts of 9.1.6 and more, surround-speaker arrays, or fully equidistant layouts, please refer to the full technical guidelines accessed through the link below.

Dolby Atmos Home Entertainment Studio Technical Guidelines
dolby.my.salesforce.com/sfc/p/#700000009YuG/a/4u000000lFHc/UYA0IZeD632SUXVmEPmUcr.wIuhpHp6Q7bVSl4LrbUQ

This focus on 'arrays' is meant for larger mix rooms of 18' wide or wider, and is meant for a large room with back seating for clients.  The 'array' concept is Not the preferred arrangement, and doesn't apply to the RMC-1+ for home use.

The best RMC-1+ match to a standard Atmos mix studio is 7.1.4.

For example from 2.3.5 in the document link above, page 21:

~~~

2.3.5 Speaker configuration design guidance for 5.1.4 speaker layouts in a

Dolby Atmos Home Entertainment Studio

Although 7.1.4 is the preferred layout for a Dolby Atmos Home Entertainment Studio, you can alternatively

use a 5.1.4 layout. However, with a 5.1.4 layout, the absence of rear surround speakers will lead to a

difference in spatial resolution.

Some key factors and expectations:

• Dolby Atmos 5.1.4 speaker layout sizes are often smaller than typical 7.1.4 layouts, but overall size

guidance remains the same as described in Room layout design .

• Extended monitor layouts should not be used with a 5.1.4 configuration.

~~~

It's a good document. I think you are getting too hung up on this 'array' concept.

Also, note that the top speakers in the mix studio are not positioned like the Dolby HT layout diagrams. See pages 18 and 19 in the doc for examples.