Where to put my next set of surround speakers?

[dcrandon]

I have a 5.1 system. My couch sits near the back wall, so my surround speakers are up on the wall about 2-3 ft higher than my head, even with the side of the sofa since they can't be put behind the sofa. The sofa is the only listening position. The surrounds are for movies. For music I favor just the two front speakers (I like to be in front of the band, not within the band on stage). Where should my next set of speakers be? I've heard they should be Height speakers that go above my main L/R speakers, and I've heard that they should be side surrounds, put on the side walls of the room, halfway between my seating position against the back wall and the TV and L/R speakers at the other side of the room. What's the proper place? Thank you.

 

[DigitalDawn]

Atmos?

 

[dcrandon]

Probably not. The WAF won't allow multiple more speakers, especially in the ceiling. So, just this one more pair. But, if I can do it somewhat Atmos with one more pair, I will.

 

[nicoleise]

To put it in simpler terms (it'll have to be somewhat general as there's no floorplan or layout) there are a few options:

1 - Leave as is.

2 - Add side surrounds. These are meant to be directly to your side when at the MLP.

3 - Add front wides. These seem to be what you describe as sides. They're meant to go somewhere between your side surrounds and front L/R speakers, to essentially close this very large gap (in degrees).

4 - Add height speakers of some sort. This would be spatial formats/Atmos, etc. I think the most established way is to add four speakers in the ceiling essentially at 45° in either direction from the MLP. It can also be done with front heights and rear heights as you mention. But for optimum effect I think it should be symmetrical to the MLP, as in you'd want the same arrangement above and in front as above and behind you.


With the way your layout seems to be, I would personally avoid both options 2 and 4 as a direct addition.

With your rear speakers so high above you, you shouldn't add spatial/Atmos channels. They really depend on separation between the bed layer speakers (at ear level) and height channels, and especially this separation needs to remain somewhat constant around the room.

As far as option 2, it would actually be my normal recommendation, and of the three "do something" options, I think it's generally the most preferred. However, these must be directly to your side (90-115° ish from center axis) and at ear level. You seem to have your rear surrounds in this position already, and if I'm picturing your layout correctly in my mind then these rear surrounds are likely firing somewhat towards the front speakers, rather than directly at you. I think side surrounds will cause a very confusing experience in which objects panning around you front to back will essentially start at the front, move to the side and then seem to move slightly back towards the front.


That leaves the option of front wides. While those can have their use, they are not a discrete channel in any soundtrack yet, to my knowledge. This means they're not playing a mixed channel of audio content, but rather they only come into play with audio objects on spacial formats such as Atmos. Generally, I've found feedback to suggest that in most soundtracks they're dead silent, but they can offer some nice effects. Especially if the seating is relatively close to the screen and you may have viewers seated well off to the sides or in front of the MLP. For them, front wides can help put the sound effects at the direction it's meant to come from, rather than "jumping" from side to front speakers instantly as would often be the experience.

So while I plan to install front wides myself and think they have potential (future use) and make sense, with your layout, again provided I am understanding it right, I don't think they'd be worth the hassle at all.


With all that processed, I would either do nothing as probably my most favourable solution, or I would first "fix" the position of the rear speakers and MLP and then add side surrounds (and possibly heights).

Hope that helps you,
Nicolai

 

[lovinthehd]

To put it in simpler terms (it'll have to be somewhat general as there's no floorplan or layout) there are a few options:

1 - Leave as is.

2 - Add side surrounds. These are meant to be directly to your side when at the MLP.

3 - Add front wides. These seem to be what you describe as sides. They're meant to go somewhere between your side surrounds and front L/R speakers, to essentially close this very large gap (in degrees).

4 - Add height speakers of some sort. This would be spatial formats/Atmos, etc. I think the most established way is to add four speakers in the ceiling essentially at 45° in either direction from the MLP. It can also be done with front heights and rear heights as you mention. But for optimum effect I think it should be symmetrical to the MLP, as in you'd want the same arrangement above and in front as above and behind you.


With the way your layout seems to be, I would personally avoid both options 2 and 4 as a direct addition.

With your rear speakers so high above you, you shouldn't add spatial/Atmos channels. They really depend on separation between the bed layer speakers (at ear level) and height channels, and especially this separation needs to remain somewhat constant around the room.

As far as option 2, it would actually be my normal recommendation, and of the three "do something" options, I think it's generally the most preferred. However, these must be directly to your side (90-115° ish from center axis) and at ear level. You seem to have your rear surrounds in this position already, and if I'm picturing your layout correctly in my mind then these rear surrounds are likely firing somewhat towards the front speakers, rather than directly at you. I think side surrounds will cause a very confusing experience in which objects panning around you front to back will essentially start at the front, move to the side and then seem to move slightly back towards the front.


That leaves the option of front wides. While those can have their use, they are not a discrete channel in any soundtrack yet, to my knowledge. This means they're not playing a mixed channel of audio content, but rather they only come into play with audio objects on spacial formats such as Atmos. Generally, I've found feedback to suggest that in most soundtracks they're dead silent, but they can offer some nice effects. Especially if the seating is relatively close to the screen and you may have viewers seated well off to the sides or in front of the MLP. For them, front wides can help put the sound effects at the direction it's meant to come from, rather than "jumping" from side to front speakers instantly as would often be the experience.

So while I plan to install front wides myself and think they have potential (future use) and make sense, with your layout, again provided I am understanding it right, I don't think they'd be worth the hassle at all.


With all that processed, I would either do nothing as probably my most favourable solution, or I would first "fix" the position of the rear speakers and MLP and then add side surrounds (and possibly heights).

Hope that helps you,
Nicolai

Front wides are more base layer surrounds than something using spacial cues like Atmos, and while not in soundtracks, an upmixer/sound mode in some models of avrs/pre-pros can use them. I tried them for a while, but felt my room a bit narrow for them to do much (at least the way that room is arranged)

OP, currently seems you have your current surround speakers in the typical position of rear surrounds...do you use the surround or rear surround connections for them? No way to move the couch off the wall some? I'd probably try and put the surrounds you have now more to the sides....perhaps in the corners above and to the sides of your seating.

 

[nicoleise]

Front wides are more base layer surrounds than something using spacial cues like Atmos, and while not in soundtracks, an upmixer/sound mode in some models of avrs/pre-pros can use them.

They are bed layer speakers by placement, but to my understanding they act entirely by spacial cues (I'm referring to native mix content, not upmixing).

Atmos mixes (to sorts stay in one area/topic) can contain up to 128 "audio tracks", which can be assigned as either "audio channels" (similar to traditional channels that will play a "continuous" stream of audio intended for that one channel/speaker) or "audio objects". The latter can conceptionally be thought off as a sound clip appended information about the desired azimuth and elevation from the MLP at any given time stamp.

The decoder will map objects in accordance with the configured speaker layout and all configured speakers will participate in rendering these audio objects when relevant. So front wides will play back objects intended for that direction.

However, even if it seems pretty logical to simply reserve a further two "audio objects" for use as "audio channels" for the front wides in the mix, this seems to not be happening in practical use thus far. Perhaps a container limitation, where Atmos consumer mixes seem to be capped at 7.1.4 discrete "audio channels". Perhaps a consideration that its "not worth it" for the small subset of users actually installing these channels. And obviously also considering that the vast majority of people stream their content where data efficiency matters more, and with many streaming providers even charging extra for Atmos, which probably again only adds to the statistics of how diminishingly small percentages of users actually take advantage of the work the mixing studio does.

This behavior matches many user reports which indicate that the users only heard the front wides play sound during the Atmos demo clips but never or only very rarely in movies. As the usage of "audio objects" increases, front wides will likely get more to do, just as it's conceivable that future "Atmos" formats will support fx 9.1.6 as a maximum of channels. So in its "correct" usage, I believe it can make some sense to install these channels, depending on room layout.

I believe mixing studios still are wrapping their head around the whole object notion. It's a very different approach. Personally I expected animation films to be first to really embrace it, since it aligns well with their preexisting work flow. But I guess it's still quite a lot more time consuming.

As you point out, if listening to fx non Atmos content, the upmixer can make use of these channels. There are also other uses, for example Trinnovs feature to create "virtual speaker positions".

I feel that this usage (upmixers at least) doesn't really merit installing front wides since in all likelihood the content being upmixed would never have intended for sound to emit from the front wide placement/direction anyways.

(With this said, I guess it's an open question if we are spoiled or deprived. On one hand, if we buy gear we'd like to see some actual use for it. On the other hand, if you consider how most people consume audio - even 2ch music - the quality put out by mixing studios is so much better than the vast majority can appreciate. Same with movies, where the majority use TV speakers or soundbars - and maybe especially for this field since cinema mixes and consumer mixes are different. So in essence, maybe we're lucky they're even bothering in the slightest. )

 

[lovinthehd]

They are bed layer speakers by placement, but to my understanding they act entirely by spacial cues (I'm referring to native mix content, not upmixing).

Atmos mixes (to sorts stay in one area/topic) can contain up to 128 "audio tracks", which can be assigned as either "audio channels" (similar to traditional channels that will play a "continuous" stream of audio intended for that one channel/speaker) or "audio objects". The latter can conceptionally be thought off as a sound clip appended information about the desired azimuth and elevation from the MLP at any given time stamp.

The decoder will map objects in accordance with the configured speaker layout and all configured speakers will participate in rendering these audio objects when relevant. So front wides will play back objects intended for that direction.

However, even if it seems pretty logical to simply reserve a further two "audio objects" for use as "audio channels" for the front wides in the mix, this seems to not be happening in practical use thus far. Perhaps a container limitation, where Atmos consumer mixes seem to be capped at 7.1.4 discrete "audio channels". Perhaps a consideration that its "not worth it" for the small subset of users actually installing these channels. And obviously also considering that the vast majority of people stream their content where data efficiency matters more, and with many streaming providers even charging extra for Atmos, which probably again only adds to the statistics of how diminishingly small percentages of users actually take advantage of the work the mixing studio does.

This behavior matches many user reports which indicate that the users only heard the front wides play sound during the Atmos demo clips but never or only very rarely in movies. As the usage of "audio objects" increases, front wides will likely get more to do, just as it's conceivable that future "Atmos" formats will support fx 9.1.6 as a maximum of channels. So in its "correct" usage, I believe it can make some sense to install these channels, depending on room layout.

I believe mixing studios still are wrapping their head around the whole object notion. It's a very different approach. Personally I expected animation films to be first to really embrace it, since it aligns well with their preexisting work flow. But I guess it's still quite a lot more time consuming.

As you point out, if listening to fx non Atmos content, the upmixer can make use of these channels. There are also other uses, for example Trinnovs feature to create "virtual speaker positions".

I feel that this usage (upmixers at least) doesn't really merit installing front wides since in all likelihood the content being upmixed would never have intended for sound to emit from the front wide placement/direction anyways.

(With this said, I guess it's an open question if we are spoiled or deprived. On one hand, if we buy gear we'd like to see some actual use for it. On the other hand, if you consider how most people consume audio - even 2ch music - the quality put out by mixing studios is so much better than the vast majority can appreciate. Same with movies, where the majority use TV speakers or soundbars - and maybe especially for this field since cinema mixes and consumer mixes are different. So in essence, maybe we're lucky they're even bothering in the slightest. )

I've not seen anything along the lines of the object oriented metadata extending to affecting the base surround layer....got any links for such? My setup wasn't Atmos at the time (and still isn't, too much pain to install in the ceiling) and the wides always had content if the upmixer/sound mode was using them. My comments were also more aimed at the OPs use....not familiar with your system.

 

[nicoleise]

I've not seen anything along the lines of the object oriented metadata extending to affecting the base surround layer....got any links for such? My setup wasn't Atmos at the time (and still isn't, too much pain to install in the ceiling) and the wides always had content if the upmixer/sound mode was using them. My comments were also more aimed at the OPs use....not familiar with your system.

Objects must use all available channels to properly work. Otherwise you can't freely assign an audio object to a given azimut and elevation relative to the listener, in which the perceived direction will be "the same" regardless of the configured speaker layout.


I think a good example, which is also on topic, is this youtube video where some movie clips are visualised through the Dolby Atmos Viewer in Trinnov, specifically to show front wides.

Objects are illustrated by orbs that move about.

As you can tell in the first example, the front wides don't play their own discrete channel mix (like the rest of the bed layer), but rather only emit sound when called upon by audio objects being placed in these channels by the decoder.

Note that the video is showing its age though. I believe the DSU has since been updated to use front wides.

 

[lovinthehd]

Objects must use all available channels to properly work. Otherwise you can't freely assign an audio object to a given azimut and elevation relative to the listener, in which the perceived direction will be "the same" regardless of the configured speaker layout.


I think a good example, which is also on topic, is this youtube video where some movie clips are visualised through the Dolby Atmos Viewer in Trinnov, specifically to show front wides.

Objects are illustrated by orbs that move about.

As you can tell in the first example, the front wides don't play their own discrete channel mix (like the rest of the bed layer), but rather only emit sound when called upon by audio objects being placed in these channels by the decoder.

Note that the video is showing its age though. I believe the DSU has since been updated to use front wides.

Is that just Trinnov, tho? I'd not heard that surrounds would change content depending on the metadata for Atmos before in any case. Any particular literature on the subject instead of a video?

 

[nicoleise]

Is that just Trinnov, tho? I'd not heard that surrounds would change content depending on the metadata for Atmos before in any case. Any particular literature on the subject instead of a video?

No, it's how the format works. Essentially Atmos (and other formats) implemented two changes, where the least significant gets all the attention. It implements overhead speakers known as height channels, and it implements (or rather adds) an object oriented structure to the format.

For the objects to work, the decoder (processor) in the playback setup must assign these objects to the physical speaker channels as appropriate by the elevation, azimuth and size properties of that object. That is the only way for this to work, since this is the only device aware of the physical playback layout.

So a side surround channel for example, will play the discrete audio channel for that channel within the mix "plus" objects that require the channel to participate, due to the direction and size properties of the object. The discrete audio channel content cannot be changed under playback, so meta data about objects does not affect this. But if an object "happens to pass by", the channel, the channel will play the associated sound. To keep things simple, this assumes that the playback layout and the mixed/encoded layout are identical (e.g. no folding, etc. taking place to convert to target layout).

The only channel barred to audio objects is the LFE channel.

A simple explanation from a written source:

In Dolby Atmos, by contrast, sound can be freed from channels. It enables artists to treat specific sounds as individual entities, called audio objects. These can be precisely placed and moved by the soundtrack creator anywhere in the cinema's three-dimensional space, though the artist can continue to use channel capabilities as desired. The Dolby Atmos cinema processor then determines which of a cinema's huge array of front, back, side, and overhead speakers it will use to recreate this lifelike movement.

(My emphasis)

Dolby Atmos Cinema Sound - Dolby Professional

Dolby Atmos creates powerful, moving audio by utilizing audio objects and overhead speakers. Immerse the viewers in your theater with breathtaking sound.

professional.dolby.com

And from the FAQs (excerpts for brevity in some cases):

An object is a discrete audio element that can be placed anywhere in the three-dimensional soundfield. Objects can consist of mono or stereo content and are positioned via dedicated Dolby Atmos panning. Use objects for precision positioning of your content. An audio object can utilize as few, or as many, speakers as defined by the positional and size metadata for that object. Objects can be static or moving and are not constrained to the outside edges of the Dolby Atmos soundfield.

Dolby Professional Support Community

Dolby Professional Support

professionalsupport.dolby.com

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.

Dolby Professional Support Community

Dolby Professional Support

professionalsupport.dolby.com

As evident by in particular the last excerpt, objects "have access" to all the bed layer channels as well. Designers are free to decide if a given sound should be placed into the discrete audio channels or into an object. But Dolby encourages using objects because it gives you greater control over the perceived direction of the object (in different playback layout configurations).

 

[lovinthehd]

No, it's how the format works. Essentially Atmos (and other formats) implemented two changes, where the least significant gets all the attention. It implements overhead speakers known as height channels, and it implements (or rather adds) an object oriented structure to the format.

For the objects to work, the decoder (processor) in the playback setup must assign these objects to the physical speaker channels as appropriate by the elevation, azimuth and size properties of that object. That is the only way for this to work, since this is the only device aware of the physical playback layout.

So a side surround channel for example, will play the discrete audio channel for that channel within the mix "plus" objects that require the channel to participate, due to the direction and size properties of the object. The discrete audio channel content cannot be changed under playback, so meta data about objects does not affect this. But if an object "happens to pass by", the channel, the channel will play the associated sound. To keep things simple, this assumes that the playback layout and the mixed/encoded layout are identical (e.g. no folding, etc. taking place to convert to target layout).

The only channel barred to audio objects is the LFE channel.

A simple explanation from a written source:

(My emphasis)

Dolby Atmos Cinema Sound - Dolby Professional

Dolby Atmos creates powerful, moving audio by utilizing audio objects and overhead speakers. Immerse the viewers in your theater with breathtaking sound.

professional.dolby.com

And from the FAQs (excerpts for brevity in some cases):

Dolby Professional Support Community

Dolby Professional Support

professionalsupport.dolby.com

Dolby Professional Support Community

Dolby Professional Support

professionalsupport.dolby.com

As evident by in particular the last excerpt, objects "have access" to all the bed layer channels as well. Designers are free to decide if a given sound should be placed into the discrete audio channels or into an object. But Dolby encourages using objects because it gives you greater control over the perceived direction of the object (in different playback layout configurations).

[/QUOTE]

Thanks, so it's not particularly the processor or upmixer, but deliberate recording decisions particularly that would affect sounds used on the bed layer...interesting...or? Not that it is likely I'll install overhead speakers in any case....

 

[nicoleise]

Thanks, so it's not particularly the processor or upmixer, but deliberate recording decisions particularly that would affect sounds used on the bed layer...interesting...or? Not that it is likely I'll install overhead speakers in any case....

Essentially it is both. But the "deliberate recording decisions" is basically how it's always been.

The creative proces would typically dictate that we want X sound to sound as if it originates from A azimuth and B elevation, and we'd like the sound to appear to be C in size. Maybe a helicopter is approaching from front right, so we'd like the sound to come from that direction, right?

That part is unchanged.

The rest is just methodology. The mixer could achieve this in a traditional workflow, but the result would also have the same limitations as previously.

The new in object oriented audio is that the mixer can directly apply those A, B, C variables to the X sound, and place that sound on its own "track" as opposed to figuring out the panning and volume in individual discrete channels.

And the big advantage is that this allows the processor/decoder (which is configured to understand the configured playback layout) to read this information and place the X sound in the appropriate channels to best imitate the A, B, C parameters in that specific configured layout.

So its more accurate in different layouts, because the playback chain is aware of what the intended result is, rather than just replaying "a bunch of audio tracks" in different speakers.

A traditional workflow would fall short here, if the layout varies, since the audio from fx non configured surround rears, would have to be "folded" into the surround side channels, as otherwise this audio information would simply disappear. But this process inevitably adds some inaccuracy to the directions sounds are perceived as coming from.

Or to put it differently, if we all installed the exact same layout, same dimensions, same room acoustics, etc., then object oriented audio would make zero difference vs a mix of discrete channels. But when reality is obviously different, objects are a very proficient solution.


And in relation to the topic; since front wides don't (yet) have their own discrete channels in most mixes, they only play objects. So whether you hear these utilised or not very much depends if the mixer used objects (in that area) or not. I'm sure the usage will evolve over time, and that's also the (albeit very small) tendency so far.

 

[lovinthehd]

Essentially it is both. But the "deliberate recording decisions" is basically how it's always been.

The creative proces would typically dictate that we want X sound to sound as if it originates from A azimuth and B elevation, and we'd like the sound to appear to be C in size. Maybe a helicopter is approaching from front right, so we'd like the sound to come from that direction, right?

That part is unchanged.

The rest is just methodology. The mixer could achieve this in a traditional workflow, but the result would also have the same limitations as previously.

The new in object oriented audio is that the mixer can directly apply those A, B, C variables to the X sound, and place that sound on its own "track" as opposed to figuring out the panning and volume in individual discrete channels.

And the big advantage is that this allows the processor/decoder (which is configured to understand the configured playback layout) to read this information and place the X sound in the appropriate channels to best imitate the A, B, C parameters in that specific configured layout.

So its more accurate in different layouts, because the playback chain is aware of what the intended result is, rather than just replaying "a bunch of audio tracks" in different speakers.

A traditional workflow would fall short here, if the layout varies, since the audio from fx non configured surround rears, would have to be "folded" into the surround side channels, as otherwise this audio information would simply disappear. But this process inevitably adds some inaccuracy to the directions sounds are perceived as coming from.

Or to put it differently, if we all installed the exact same layout, same dimensions, same room acoustics, etc., then object oriented audio would make zero difference vs a mix of discrete channels. But when reality is obviously different, objects are a very proficient solution.


And in relation to the topic; since front wides don't (yet) have their own discrete channels in most mixes, they only play objects. So whether you hear these utilised or not very much depends if the mixer used objects (in that area) or not. I'm sure the usage will evolve over time, and that's also the (albeit very small) tendency so far.

Thanks again. Is there any plan for an Atmos recording to have native wide content? Can blurays even accommodate?

 

[nicoleise]

Thanks again. Is there any plan for an Atmos recording to have native wide content? Can blurays even accommodate?

Not so far as I know (speaking of movies).

In Dolby Atmos, the largest bed configuration that exists is 7.1.2.

https://professionalsupport.dolby.com/s/article/What-is-a-bed?language=en_US

I suppose this limitation could very well be due to capacity (Blu-ray and streaming).

I don't know whether it's desirable either, to have discrete FW channels. Maybe it is.

A designer/mixer could probably emulate this easily by simply creating an object at each FW position with a size of 0, and pour the full audio the designer wants that channel to play into it. But this might well behave in a very funky way if the layout is different than anticipated. I think this is why this is seemingly done mainly to add very subtle effect.

In general, I think that i) the overall creative result matters much more than the individual utilization of X speaker and ii) the "solution" (if one is needed) is likely a higher rate of adoption to the use of objects in e.g. Atmos mixes.

I think objects are challenging the industry a bit like when CGI came about. It's obvious that it's a very different process, both creatively and technically.

To use objects, you need to record audio files that contain (only and all) the sound an object should have. That is much different from wiring some microphones on a set and in a foley studio and recording all the sound together through one or more microphone arrays. Essentially solving the positioning issue by actually positioning the action where it's meant to go.

This challenge mounts when you consider that each action often would need to have multiple objects to accurately render.

Imagine for example a tennis court viewed from the side. A "ball cannon" on one side of the net serving a human player on the opposite side. Recording this with two microphones in stereo essentially, would likely provide a very believable experience in a very easy way. If one wanted to, one could even add reverbs of the recording in the studio (even into other channels) to simulate the echo in the venue. I think most people would feel as if they were there, if some thoughtful ambiance is added.

Now imagine this as objects (exclusively). Suddenly you need audio of the "ball cannon" firing (and only that). Separately, you need to record the sound of the human held racket hitting the ball (and only that). If you're going for effects, you'd also need to record a separate object (probably foley artist made) of the ball flying through the air. That's three objects already, and obviously many other interesting things could occur, each needing their own object. There could be the sound of the human, there could be other humans playing tennis in other locations, there could even be a plane flying overhead.

All this effort may seem fair if this is all very pivotal to the plot and action. But otherwise it's probably all effort that would be better spent elsewhere.

And this is also why I assume animation films will have an easier time adopting to this, since it's not as far from their normal (non-audio) workflow.


But much as with CGI, I think the industry learns what works and what doesn't, and how to better utilize the potential of this technology. And as they do, I think there's bound to be a wider adoption of the use of objects eventually, which will benefit all channels but obviously in particular channels like FWs that don't have discrete tracks to play.

Personally I would feel that a 9.1 discrete bed layer would make more sense than a 7.1.2, and the designers simply have to use objects for everything overhead and other desirable effects. But the format is designed with a different priority, obviously, probably matching better to what (home and "real") cinemas actually install.

 

[DIBstangguy]

Great info nicoleise! Thank you for your efforts.

 

[lovinthehd]

Not so far as I know (speaking of movies).

https://professionalsupport.dolby.com/s/article/What-is-a-bed?language=en_US

I suppose this limitation could very well be due to capacity (Blu-ray and streaming).

I don't know whether it's desirable either, to have discrete FW channels. Maybe it is.

A designer/mixer could probably emulate this easily by simply creating an object at each FW position with a size of 0, and pour the full audio the designer wants that channel to play into it. But this might well behave in a very funky way if the layout is different than anticipated. I think this is why this is seemingly done mainly to add very subtle effect.

In general, I think that i) the overall creative result matters much more than the individual utilization of X speaker and ii) the "solution" (if one is needed) is likely a higher rate of adoption to the use of objects in e.g. Atmos mixes.

I think objects are challenging the industry a bit like when CGI came about. It's obvious that it's a very different process, both creatively and technically.

To use objects, you need to record audio files that contain (only and all) the sound an object should have. That is much different from wiring some microphones on a set and in a foley studio and recording all the sound together through one or more microphone arrays. Essentially solving the positioning issue by actually positioning the action where it's meant to go.

This challenge mounts when you consider that each action often would need to have multiple objects to accurately render.

Imagine for example a tennis court viewed from the side. A "ball cannon" on one side of the net serving a human player on the opposite side. Recording this with two microphones in stereo essentially, would likely provide a very believable experience in a very easy way. If one wanted to, one could even add reverbs of the recording in the studio (even into other channels) to simulate the echo in the venue. I think most people would feel as if they were there, if some thoughtful ambiance is added.

Now imagine this as objects (exclusively). Suddenly you need audio of the "ball cannon" firing (and only that). Separately, you need to record the sound of the human held racket hitting the ball (and only that). If you're going for effects, you'd also need to record a separate object (probably foley artist made) of the ball flying through the air. That's three objects already, and obviously many other interesting things could occur, each needing their own object. There could be the sound of the human, there could be other humans playing tennis in other locations, there could even be a plane flying overhead.

All this effort may seem fair if this is all very pivotal to the plot and action. But otherwise it's probably all effort that would be better spent elsewhere.

And this is also why I assume animation films will have an easier time adopting to this, since it's not as far from their normal (non-audio) workflow.


But much as with CGI, I think the industry learns what works and what doesn't, and how to better utilize the potential of this technology. And as they do, I think there's bound to be a wider adoption of the use of objects eventually, which will benefit all channels but obviously in particular channels like FWs that don't have discrete tracks to play.

Personally I would feel that a 9.1 discrete bed layer would make more sense than a 7.1.2, and the designers simply have to use objects for everything overhead and other desirable effects. But the format is designed with a different priority, obviously, probably matching better to what (home and "real") cinemas actually install.

How is 7.1.2 a bed layer?

Fwiw what I saw for bluray it is a max of 8ch, so just the 7.1 bed layer plus metadata...

 

[nicoleise]

I think they (Dolby in the FAQs) mean bed as foundation, not so much as the elevation of the speakers. As in; you can mix a traditional "bed layer" of up to 7.1.2 and "build your objects" on top of that.

Yes, most audio formats for use on BD seem to cap out at 8 ch. The 7.1.2 Dolby mentions may be an option only pertinent to cinema renders.

 

[Xrazor77]

 

[dcrandon]

Thanks for the video. But, what is "Dolby Atmos Enabled" speakers in this video. It's just that they are angled down, correct?

 

[William Lemmerhirt]

Thanks for the video. But, what is "Dolby Atmos Enabled" speakers in this video. It's just that they are angled down, correct?

Nope. That means upfiring bouncy speakers.