|
|
This topic comprises 8 pages: 1 2 3 4 5 6 7 8
|
|
Author
|
Topic: How long will Dolby Atmos last?
|
|
|
Bobby Henderson
"Ask me about Trajan."
Posts: 10973
From: Lawton, OK, USA
Registered: Apr 2001
|
posted 04-15-2015 01:52 PM
quote: Marcel Birgelen
Well, that's why Atmos probably started off with the concept of beds, like I mentioned before, it is a transitional solution. But "beds" can be implemented as objects too, given the right tools, this can be a fully automated process.
A pre-existing sound element that has already been pre-mixed from numerous elements and "flattened" down into something like a stereo mix or 5.1 mix isn't something that can be changed into an individual object, not unless the whole thing is summed into a single monoaural audio track. Summing a stereo or 5.1 stem into a mono object could yield some unwanted, odd sounding results -kind of like noise from a TV set playing from a specific part of the room.
I can't open up a music CD .WAV file and surgically extract the vocals out of the sound mix or individually select any elements I hear in the mix and change them. Making such changes is difficult. It's no different than taking a Photoshop image with many layers, flattening it and saving it as a PNG or JPEG image. Once the image is flattened everything is permanently baked into place. To make any precise edits you have to go back to the layered Photoshop file, if you still have it.
Fundamentally, sound objects are supposed to be isolated sources not contaminated with any other audio. A vocal track recorded on an ADR stage is an isolated object. A gunshot sound can be recorded as an isolated object. Same goes for a shell casing hitting the floor. If you have a pre-mixed 5.1 stem of shell casings bouncing across a floor there's really nothing there to reposition. The sound panning work was already done on that element.
quote: Marcel Birgelen
In my opinion it's best to have a pure format omitting most of the legacy and tools that do the conversion for you.
That might be a pure approach, but it sure is not a practical approach and not flexible either. Creative people prefer to have multiple ways to skin the proverbial cat. The only thing that matters is how good the finished mix sounds.
A movie mix that is built up from pre-mixed stems in combination with discrete objects is probably going to sound more complex and lush than something that was built up solely with objects alone. That's because there's only so much time to dink around with dozens or even hundreds of individual objects from one scene to the next.
Perhaps as time goes on, if MDA is widely adopted in the sound editing community and is available to use in various sound editing systems and software applications we might see libraries of various editable sound effects and sound environments that can be modified in an 3D sound layout. We're not there yet.
quote: Marcel Birgelen
I'm by no means the biggest fan of Digital IMAX and I have my doubts they will upgrade every location, but this is, as of now, just speculation. Maybe they actually do have plans to overhaul each and every theater.
Some of what I'm talking about with IMAX and their new 4K laser projection has to do with comments from their people made recently in conjunction with the TCL Chinese Theater unveiling its laser-based system with the premiere of Furious 7.
quote: Marcel Birgelen
It worked before with two labels that became rather meaningless in the process: THX and IMAX. Why shouldn't it work with a new label that has some strong marketing behind it but will actually be relevant once more?
There's too many people with profits to protect standing in the way of full disclosure and educating the public. They're not going to let the public know fully just what's up in the projection booth and installed inside the theater. They're making more money just throwing around brand names.
| IP: Logged
|
|
Marcel Birgelen
Film God
Posts: 3357
From: Maastricht, Limburg, Netherlands
Registered: Feb 2012
|
posted 04-15-2015 04:58 PM
quote: Bobby Henderson
A pre-existing sound element that has already been pre-mixed from numerous elements and "flattened" down into something like a stereo mix or 5.1 mix isn't something that can be changed into an individual object, not unless the whole thing is summed into a single monoaural audio track. Summing a stereo or 5.1 stem into a mono object could yield some unwanted, odd sounding results -kind of like noise from a TV set playing from a specific part of the room.
You can still import multi-channel samples into your MDA projects. They will appear as separate objects in a group or you can flatten them down to monaural objects. You can even adjust the panning and other parameters on each imported "channel", just like any other object after you've imported them.
quote: Bobby Henderson
I can't open up a music CD .WAV file and surgically extract the vocals out of the sound mix or individually select any elements I hear in the mix and change them. Making such changes is difficult. It's no different than taking a Photoshop image with many layers, flattening it and saving it as a PNG or JPEG image. Once the image is flattened everything is permanently baked into place. To make any precise edits you have to go back to the layered Photoshop file, if you still have it.
The same is true for any audio mix, graphics project or practically any project you can think of. If you threw away your original sources, it's your fault and it's gonna be a lot of trouble to get them back, if at all possible.
quote: Bobby Henderson
Fundamentally, sound objects are supposed to be isolated sources not contaminated with any other audio. A vocal track recorded on an ADR stage is an isolated object. A gunshot sound can be recorded as an isolated object. Same goes for a shell casing hitting the floor. If you have a pre-mixed 5.1 stem of shell casings bouncing across a floor there's really nothing there to reposition. The sound panning work was already done on that element.
Like I mentioned before, a multi-channel source can be imported as multiple objects (those objects can be grouped to keep stuff maintainable, just like Photoshop and Illustrator let you group stuff) and panning can still be adjusted on those separate imported objects if required.
The same goes for background music. If you're stuck with a stereo mix, all you'll get when importing it will be two objects, but you can still freely pan them around like you please.
quote: Bobby Henderson
A movie mix that is built up from pre-mixed stems in combination with discrete objects is probably going to sound more complex and lush than something that was built up solely with objects alone. That's because there's only so much time to dink around with dozens or even hundreds of individual objects from one scene to the next.
The point is, everything that can be done with "stems" or "beds" can be done with objects too. It's up to the tools producing the mix how easy the workflow will be.
quote: Bobby Henderson
Perhaps as time goes on, if MDA is widely adopted in the sound editing community and is available to use in various sound editing systems and software applications we might see libraries of various editable sound effects and sound environments that can be modified in an 3D sound layout. We're not there yet.
If you're looking at what e.g. Fairlight is doing, I'd say we're already pretty much there already. Maybe it will take a release before it becomes a core attribute of ProTools, but Avid need to sell upgrades to keep their shareholders happy, so they're surely going to jump on this bandwagon with full force, before others eat their lunch.
quote: Bobby Henderson
Some of what I'm talking about with IMAX and their new 4K laser projection has to do with comments from their people made recently in conjunction with the TCL Chinese Theater unveiling its laser-based system with the premiere of Furious 7.
I'm not aware that IMAX ever officially acknowledged they don't intent to upgrade all of their deficient digital xenon setups. Somehow I also wouldn't expect IMAX to comment too much on this matter, because it could hurt their brand if they acknowledge some IMAX theaters aren't worth the "premium treatment".
quote: Bobby Henderson
There's too many people with profits to protect standing in the way of full disclosure and educating the public. They're not going to let the public know fully just what's up in the projection booth and installed inside the theater. They're making more money just throwing around brand names.
In the end the people with clue will educate part of the clueless public. Then they might recognize the value of "certification Y", not just because of the cool trailers, but because it was actually better than this other acronym most likely containing an X.
THX worked, because theaters that did have THX didn't sound like shit and people started to notice. Once THX got sloppy, meaningless and could be had for few bucks on your next toilet seat with integrated speakers, people didn't give a damn anymore.
I refuse to give up this idea, there are still a lot of people that actually do care.
| IP: Logged
|
|
Bobby Henderson
"Ask me about Trajan."
Posts: 10973
From: Lawton, OK, USA
Registered: Apr 2001
|
posted 04-15-2015 11:25 PM
quote: Marcel Birgelen
You can still import multi-channel samples into your MDA projects. They will appear as separate objects in a group or you can flatten them down to monaural objects. You can even adjust the panning and other parameters on each imported "channel", just like any other object after you've imported them.
The only thing object-based you'll get importing a multi-channel sample is each channel as an object. The problem is each channel has already been down-mixed from dozens or even hundreds of different audio elements. That's worthless as something to be specifically positioned or panned around a room. You can't really do much of anything with such elements other than change them globally. If you jack around with any of the individual channels with object-like changes you'll break the relationship with the rest of that 5.1 stem.
quote: Marcel Birgelen
The same is true for any audio mix, graphics project or practically any project you can think of. If you threw away your original sources, it's your fault and it's gonna be a lot of trouble to get them back, if at all possible.
That's true. Nevertheless I think you're missing the point about the differences between individual audio objects and channel-based audio that has been downmixed from many elements. You simply cannot import a flattened, down-mixed 5.1 track or other multi-channel source into an audio editor and expect to get back all the discrete separate objects that were used to create the original mix. That's just plain impossible. You gotta go back to the orignal, not yet down-mixed ProTools layout or whatever was used to generate the 5.1 mix in the first place.
I cannot open a flattened TIFF image in Photoshop and expect Photoshop to magically expand the image into a bunch of separate layers. The same thing is true for a multi-channel mix.
quote: Marcel Birgelen
I'm not aware that IMAX ever officially acknowledged they don't intent to upgrade all of their deficient digital xenon setups. Somehow I also wouldn't expect IMAX to comment too much on this matter, because it could hurt their brand if they acknowledge some IMAX theaters aren't worth the "premium treatment".
There is a big cost difference between a pair of 4K laser-based projectors versus a pair of 2K xenon-based projectors. They have stated pretty clearly the laser-based systems are only going to be deployed on the biggest screens.
quote: Marcel Birgelen
In the end the people with clue will educate part of the clueless public.
The people who understand have been trying to get the message through, apparently without much luck.
| IP: Logged
|
|
|
|
|
|
Marcel Birgelen
Film God
Posts: 3357
From: Maastricht, Limburg, Netherlands
Registered: Feb 2012
|
posted 04-16-2015 03:34 PM
quote: Bobby Henderson
The only thing object-based you'll get importing a multi-channel sample is each channel as an object. The problem is each channel has already been down-mixed from dozens or even hundreds of different audio elements. That's worthless as something to be specifically positioned or panned around a room. You can't really do much of anything with such elements other than change them globally. If you jack around with any of the individual channels with object-like changes you'll break the relationship with the rest of that 5.1 stem
But how is this different from importing a pre-mixed sample into any other Pro-Tools project? And you messing around with the panning of the imported objects is your own responsibility.
quote: Bobby Henderson
Nevertheless I think you're missing the point about the differences between individual audio objects and channel-based audio that has been downmixed from many elements. You simply cannot import a flattened, down-mixed 5.1 track or other multi-channel source into an audio editor and expect to get back all the discrete separate objects that were used to create the original mix. That's just plain impossible. You gotta go back to the orignal, not yet down-mixed ProTools layout or whatever was used to generate the 5.1 mix in the first place.
I guess do know the difference between source project files and compiled/rendered deliverables... But my point is, like I mentioned in other posts and in this post before: This situation is not different from a discrete channel based mix. Isolating a certain sound or other elements from a "flattened" sample is going to be surgical work with an often uncertain end-result.
And yeah, this object based paradigm shift will certainly require some investment in updating your library of pre-fabricated sound effects if you want to get the most out of it. But it's not like this is an entirely new thing. Every TV channel transitioning to HD had to redo quite a lot of boilerplate content and templates to get the most out of HD. But it's not like it's impossible to show SD content on an HD station.
quote: Bobby Henderson
There is a big cost difference between a pair of 4K laser-based projectors versus a pair of 2K xenon-based projectors. They have stated pretty clearly the laser-based systems are only going to be deployed on the biggest screens.
But they also made clear they considered this frankendigital thing to be a stop-gap solution, didn't they? Maybe they actually saved tons of cash for an all-out conversion? I'm pretty sure Barco would give them a nice discount. ![[Wink]](wink.gif)
quote: Bobby Henderson
The people who understand have been trying to get the message through, apparently without much luck.
A lot of people I know of do still know the difference between a shitty, ok and a great cinematic presentation and avoid the shitty ones, so not all hope is lost. Some also actually made it clear to me they felt cheated by this thing called "IMAX", which one described to me like going to a bad rock concert where the band had been replaced by a big, ugly LED wall.
quote:
and there's been confirmation re: 'Jurassic World' too - although the aspect ratio doesn't look correct:
I hope it does, this 1:2.0 aspect ratio is utterly stupid, inefficient and just serves the ego of the director with his super-cool defacto-exclusive aspect ratio.
| IP: Logged
|
|
|
|
Bobby Henderson
"Ask me about Trajan."
Posts: 10973
From: Lawton, OK, USA
Registered: Apr 2001
|
posted 04-20-2015 02:36 PM
Dolby added a few more titles to its upcoming Atmos Movie Releases list at the Dolby web site. Future releases into the Summer are as follows:
The Age of Adaline • April 24
Avengers: Age of Ultron • May 1
Mad Max: Fury Road • May 15 (Terry's prediction was wrong)
Main Aur Charles • May 15
Tomorrowland • May 22
San Andreas • May 29
Spy • June 5
Entourage • June 5
Traffic • June 15
Inside Out • June 19
Magic Mike XXL • July 1
Terminator Genisys • July 1
Minions • July 10
Marvel's Ant-Man • July 17 (Terry's prediction was wrong)
Pixels • July 24
Pan • July 24
Mission: Impossible-Rogue Nation • July 31
Fantastic Four • August 7
The Man from UNCLE • August 14
Internet Movie Database has Jurassic World listed as having both Dolby Atmos and Auro 11.1 mixes, as well as that odd 2.00:1 aspect ratio. I guess we'll have to see how that pans out on June 12. Jurassic World would also seem to be a very key title to hype DTS:X, given the original JP release in 1993 is what launched the original DTS format.
17 of the 19 titles listed above are North American releases. That's not too bad for a release slate through mid August. I was surprised to see movies like Entourage and the Magic Mike sequel getting Atmos treatment. I would not be surprised to see more titles added to the list as various movies get closer to their release dates.
Now, on to some other stuff:
quote: Marcel Birgelen
But how is this different from importing a pre-mixed sample into any other Pro-Tools project? And you messing around with the panning of the imported objects is your own responsibility.
Marcel, do you not understand how down-mixed sound works? There are no individual objects to strip out and pan individually. Some piece of audio already down-mixed into a flattened 5.1 layout has nothing individual to pan. If it's something that's been down-mixed into stereo or 5.1 that piece of audio is locked into that layout. You can layer 5.1 stems for music, sound effects, etc. into a final 5.1 layout. But to have anything that's been down-mixed into a conventional 5.1 layout try to act like it's former object based self prior to mixing is asking too much.
This makes me think of silly things that happen when some people don't understand the differences between certain kind of digital assets. For instance, I've had people bring me a "logo" they want to use on a sign. It's a low resolution JPEG image. I tell them I need vector-based artwork in Adobe Illustrator .AI format. What do they do? Put the JPEG image inside an Illustrator .AI document expecting the artwork to somehow magically translate into clean, scalable vector art.
![[Roll Eyes]](rolleyes.gif)
quote: Marcel Birgelen
Every TV channel transitioning to HD had to redo quite a lot of boilerplate content and templates to get the most out of HD. But it's not like it's impossible to show SD content on an HD station.
There is only one correct way to show SD on HD: pillar-box it. Interpolation technology can't invent HD levels of detail from nothing. A video image with 345,600 pixels of detail cannot be inflated into over 2 million pixels worth of detail. There is no way to avoid it looking fuzzy. HD itself is done wrong in many ways. I see that shit all the time, and it is only that: shit. There is zero excuse for any TV station to do on screen graphics for HDTV wrong. Many networks on local TV stations just extract the center 4:3 zone for re-broadcast in SD. But they often don't compose graphics for that safe title area. So you get things like names and phone numbers getting cut off at the left or right edge during the news or a commercial. Some networks air SD content in stretch mode to fill the 16:9 frame -now the camera adds 100 pounds! I really love the one where a TV station will get a letter-boxed SD copy of a commercial originally produced in HD. When it airs it is both letter-boxed and pillar-boxed. Then there's all the dip shits out there grabbing eyewitness video for the news, but holding their phones vertically. It's funny to watch networks try to blow-up the vertical video and do all sorts of other creative things to make the footage less shitty. I have yet to see any of it look better than shit quality. If I had my way, I'd run the damned footage sideways, properly filling the HDTV screen but forcing everyone at home to turn their heads 90 degrees. Maybe that would start giving people a clue about how stupid it is to make a vertical video.
The point with that whole rant is just because a lot of people are goofing up HDTV doesn't mean people have to do stupid shit with next gen audio mixes.
quote: Marcel Birgelen
But they also made clear they considered this frankendigital thing to be a stop-gap solution, didn't they? Maybe they actually saved tons of cash for an all-out conversion? I'm pretty sure Barco would give them a nice discount.
"Lie-MAX" has been around going on 7 years. That's not exactly a stop-gap thing. It appears the company will not convert very many of those existing 2K theaters to the new laser-based system. IIRC, laser-based system will bring back the taller screens we associate with 15/70 film-based theaters. Most multiplex theaters bearing IMAX logos won't be able to do that.
| IP: Logged
|
|
|
|
Marcel Birgelen
Film God
Posts: 3357
From: Maastricht, Limburg, Netherlands
Registered: Feb 2012
|
posted 04-20-2015 06:09 PM
quote: Bobby Henderson
Marcel, do you not understand how down-mixed sound works? There are no individual objects to strip out and pan individually. Some piece of audio already down-mixed into a flattened 5.1 layout has nothing individual to pan. If it's something that's been down-mixed into stereo or 5.1 that piece of audio is locked into that layout. You can layer 5.1 stems for music, sound effects, etc. into a final 5.1 layout. But to have anything that's been down-mixed into a conventional 5.1 layout try to act like it's former object based self prior to mixing is asking too much.
Maybe you're not reading what I'm writing then... I'm not talking about magical "object extraction software", but a very simple concept, I will try to explain it again, please try to follow...
If you import a 5 channel mix into a purely object based environment, the mixing software will create *five* objects, each object containing one of the five channels in your original mix. It will *not* magically extract objects from your flattened mix. Those sound objects will be placed on the *virtual* locations where the speakers are supposed to be in the 5 channel mix, maybe with a little spread applied to it. You probably can tune the default parameters during import or else you can adjust them by hand later on. The specifics are up to the implementation of your chosen tool.
If you have a 5.1 track, you could still argue what should happen with the ".1" track. Should it become a separate object, mixed into one of the other five objects or "averaged" over the other five objects?
quote: Bobby Henderson
This makes me think of silly things that happen when some people don't understand the differences between certain kind of digital assets. For instance, I've had people bring me a "logo" they want to use on a sign. It's a low resolution JPEG image. I tell them I need vector-based artwork in Adobe Illustrator .AI format. What do they do? Put the JPEG image inside an Illustrator .AI document expecting the artwork to somehow magically translate into clean, scalable vector art.
I know all to well about people doing a right-click save-as of their tiny JPEG or PNG logo on their website and expect them to stretch it out over massive proportions. And yeah, I also know the dickheads that manage to copy/paste it into an AI file and expect it to become vector graphics automatically. People generally have no clue how stuff works...
quote: Bobby Henderson
There is only one correct way to show SD on HD: pillar-box it. Interpolation technology can't invent HD levels of detail from nothing. A video image with 345,600 pixels of detail cannot be inflated into over 2 million pixels worth of detail. There is no way to avoid it looking fuzzy.
Well, yeah, but does it really matter if their TV set stretches it from SD to HD or if the broadcaster does it for them? The end result is a practically the same... People don't like black bars or borders, they usually stretch, zoom, tilt and rotate until it fits on the screen and don't mind stuff looking all wrong and blurry.
quote: Bobby Henderson
The point with that whole rant is just because a lot of people are goofing up HDTV doesn't mean people have to do stupid shit with next gen audio mixes.
People goof up all the time, technology will not stop them, just to the contrary, they will just find new ways to screw it all up again and again.
Dolby Atmos mixes also have been screwed up badly for example. Still, I don't understand what you're actually ranting about to be honest. The idea that future MDA formats will be largely speaker-agnostic, so they can better scale to any layout? The non-uniform branding? Or me not getting my point across about the obsolescence of channel-mapped audio in MDA formats?
quote: Bobby Henderson
"Lie-MAX" has been around going on 7 years. That's not exactly a stop-gap thing. It appears the company will not convert very many of those existing 2K theaters to the new laser-based system. IIRC, laser-based system will bring back the taller screens we associate with 15/70 film-based theaters. Most multiplex theaters bearing IMAX logos won't be able to do that.
They could use the lack of suitable technology as a defense. But due to their "patented room layout", even their smallest screens just requires 4K to be taken even remotely seriously as a premium format.
| IP: Logged
|
|
Bobby Henderson
"Ask me about Trajan."
Posts: 10973
From: Lawton, OK, USA
Registered: Apr 2001
|
posted 04-20-2015 11:12 PM
quote: Marcel Birgelen
If you import a 5 channel mix into a purely object based environment, the mixing software will create *five* objects, each object containing one of the five channels in your original mix. It will *not* magically extract objects from your flattened mix. Those sound objects will be placed on the *virtual* locations where the speakers are supposed to be in the 5 channel mix, maybe with a little spread applied to it. You probably can tune the default parameters during import or else you can adjust them by hand later on. The specifics are up to the implementation of your chosen tool.
Dolby Atmos' "beds" approach for any channel-based material would already treat something like this correctly. Each channel-based track in that 5.1 stem would be treated as it would normally be treated in a conventional 5.1 playback environment. The LCR stage channel audio would be rendered as-is. The split surrounds would be rendered as beds for all the surround speakers on the side and back walls, just divided in half the same way any normal split-surround mix would be played. There is no reason to treat a 5.1 stem any different than that.
quote: Marcel Birgelen
Dolby Atmos mixes also have been screwed up badly for example. Still, I don't understand what you're actually ranting about to be honest. The idea that future MDA formats will be largely speaker-agnostic, so they can better scale to any layout? The non-uniform branding? Or me not getting my point across about the obsolescence of channel-mapped audio in MDA formats?
"Speaker agnostic" sounds like a neat term, but unfortunately it can be twisted into "hey, we only need 3 speakers to do next gen object-based surround sound on the cheap! We're going to get rich screwing over these dumbass customers! Hooray!" Open MDA looks like a big opportunity for theaters to literally con customers into thinking they're getting something new and state of the art when that really isn't the case.
Marcel, are you willing to get specific with just how many speakers and amplifiers a theater needs in order to render object based surround sound correctly? From what I've read about the original Iosono prototype, Dolby Atmos itself is making some considerable compromises. But at least they are suggesting specific standards for numbers of speakers, amplifiers, subwoofers, etc. based on room size. Open MDA doesn't appear to even do that. From my perspective Open MDA just says any speaker layout will do, regardless of how many or how few there are in the system. It suggests to the theater operator to cut corners, cut costs, do as little as possible and MDA will fill in the gaps.
Non-uniform branding translates into don't even bother with any branding at all. This whole parade of silly acronym labels for premium priced big screen auditoriums may have actually helped IMAX' business. Casual movie goers see all these other high priced theater houses trying to act like they're IMAX or Lie-MAX, but have a lesser, meaningless brand name. A parade of private label brands re-using DTS:X/MDA technology will essentially be the same thing. And they could end up elevating Dolby Atmos' business. Because at least most people have heard of Dolby before. They're not going to know what to think of all the possible re-branding possibilities with MDA. As I said earlier, theaters can't really use "MDA" itself since it sounds like the Muscular Dystrophy Association.
| IP: Logged
|
|
Marcel Birgelen
Film God
Posts: 3357
From: Maastricht, Limburg, Netherlands
Registered: Feb 2012
|
posted 04-21-2015 02:22 AM
quote: Bobby Henderson
Dolby Atmos' "beds" approach for any channel-based material would already treat something like this correctly. Each channel-based track in that 5.1 stem would be treated as it would normally be treated in a conventional 5.1 playback environment. The LCR stage channel audio would be rendered as-is. The split surrounds would be rendered as beds for all the surround speakers on the side and back walls, just divided in half the same way any normal split-surround mix would be played. There is no reason to treat a 5.1 stem any different than that.
The reason is because you want to have a future-proof format, that isn't dependent on the concept of channels. If we find a way, in the future, to do "immersive 3D audio" without boatloads of speakers, the concept of fixed channels is more of a hindrance.
Dolby Atmos used this concept primarily to get traction in the MDA space. It's a hybrid solution, where you can take your basic 5.1, 7.1 or 9.1 mix, put those channels into the beds and add pour some object sauce over it and ready is your Dolby Atmos mix. They can do this, because they maintain the basic 7.1 layout as a reference and extend their setup from there.
A truly "next gen" format can do away with those concepts and still perfectly emulate "legacy".
quote: Bobby Henderson
"Speaker agnostic" sounds like a neat term, but unfortunately it can be twisted into "hey, we only need 3 speakers to do next gen object-based surround sound on the cheap! We're going to get rich screwing over these dumbass customers! Hooray!" Open MDA looks like a big opportunity for theaters to literally con customers into thinking they're getting something new and state of the art when that really isn't the case.
Well, yes. But screwing dumbass customers could be done with DTS, SRD and even SDDS. If Brand X wants to avoid associating their name with those kind of setups, it's their responsibility to put some minimal requirements in the specs and enforce them. It's not the failure of MDA as a format, if people cheap out.
quote: Bobby Henderson
Marcel, are you willing to get specific with just how many speakers and amplifiers a theater needs in order to render object based surround sound correctly? From what I've read about the original Iosono prototype, Dolby Atmos itself is making some considerable compromises. But at least they are suggesting specific standards for numbers of speakers, amplifiers, subwoofers, etc. based on room size. Open MDA doesn't appear to even do that. From my perspective Open MDA just says any speaker layout will do, regardless of how many or how few there are in the system. It suggests to the theater operator to cut corners, cut costs, do as little as possible and MDA will fill in the gaps.
Just like you can use multiple renderers to render the same 3D graphics scene (E.g. Mental Ray v.s. RenderMan), you can use different processes to render "3D audio". Besides the size and shape and acoustics of the room, it is highly dependent on the process you're using. VBAP (Dolby Atmos, DTS:X), DBAP and WFS (Iosono) all have specific requirements. And since the research in this space is alive and kicking, people will come up with all kinds of new "rendering" techniques. Of the three systems mentioned, WFS needs by far the biggest speaker density to achieve good results.
Maybe I could do some comparison on what is generally considered a minimum amount of speakers for each system for some "average room size", I will try to do so if I have some time to spare.
| IP: Logged
|
|
Bobby Henderson
"Ask me about Trajan."
Posts: 10973
From: Lawton, OK, USA
Registered: Apr 2001
|
posted 04-21-2015 09:57 PM
quote: Marcel Birgelen
The reason is because you want to have a future-proof format, that isn't dependent on the concept of channels. If we find a way, in the future, to do "immersive 3D audio" without boatloads of speakers, the concept of fixed channels is more of a hindrance.
Please provide some proof in how Atmos supporting both channels and objects is a hindrance. A real hindrance would be expecting a movie production's post production team to deliver a mix that's entirely object-based in the very limited amount of time they have to complete the project.
quote: Marcel Birgelen
Dolby Atmos used this concept primarily to get traction in the MDA space. It's a hybrid solution, where you can take your basic 5.1, 7.1 or 9.1 mix, put those channels into the beds and add pour some object sauce over it and ready is your Dolby Atmos mix. They can do this, because they maintain the basic 7.1 layout as a reference and extend their setup from there.
Get traction in the MDA space? Dolby Atmos was in development long before DTS' MDA and has had a product on the market for nearly 3 years. Meanwhile MDA is fixing to get a product into the market.
Dolby Atmos is indeed a hybrid solution. It has the capability of handling both channel-based and object-based audio simultaneously. A sound designer can choose to mix a movie in Atmos entirely based on objects. Likewise someone can generate an Atmos-mix with nothing more than 9.1 channels of totally "bed" based audio (but that obviously wouldn't be a good thing). An analogy to this would be how Adobe Illustrator accommodates both vector-based artwork and raster-based artwork into the same layout. It would be flat out stupid and counter-productive to make Illustrator a purely vector-only application.
There are valid reasons for providing support for both objects and beds in Atmos. No other sound format appears to be going object-only either.
quote: Marcel Birgelen
But screwing dumbass customers could be done with DTS, SRD and even SDDS.
With those sound formats most of the 5.1 surround done wrong has little to do with the sound formats themselves and mostly to do with the theater's speakers, amps, choice of crossover, etc. Another big thing is maintenance. On that level a theater could be screwing over customers by doing mono or stereo sound wrong.
But here's the thing with next-gen surround sound: it's almost always a featured ingredient in premium priced auditoriums. It is rare to find an Atmos sound system or Auro sound system in a standard priced auditorium. On that level there is a requirement to do the damned job right. Dolby Atmos is closer out of any of the systems being marketed right now in demanding standards for doing the job right. Open MDA doesn't appear to be suggesting any quality standards at all.
I won't be all that critical of a standard priced theater doing a slight amount of Open MDA enhancement to what is really only a conventional 5.1 surround system as long as I don't have to pay anything extra to see a movie in that room. But if I'm going to be paying $13 or $16 for a ticket and the theater is pulling the same cost cutting bullshit I'll get justifiably angry and not visit that particular theater again.
quote: Marcel Birgelen
Just like you can use multiple renderers to render the same 3D graphics scene (E.g. Mental Ray v.s. RenderMan), you can use different processes to render "3D audio". Besides the size and shape and acoustics of the room, it is highly dependent on the process you're using. VBAP (Dolby Atmos, DTS:X), DBAP and WFS (Iosono) all have specific requirements. And since the research in this space is alive and kicking, people will come up with all kinds of new "rendering" techniques. Of the three systems mentioned, WFS needs by far the biggest speaker density to achieve good results.
Fundamentally 3D object based sound requires enough speakers and amplifiers to properly render the audio. It's no different than a digital image requiring enough pixels of native detail to reach an acceptable level of quality. Some cheap approach that uses hardly any more speakers and amplifiers than a standard 5.1 sound system will be woefully inadequate. There's just no way around that.
| IP: Logged
|
|
Marcel Birgelen
Film God
Posts: 3357
From: Maastricht, Limburg, Netherlands
Registered: Feb 2012
|
posted 04-22-2015 03:55 AM
quote: Bobby Henderson
Please provide some proof in how Atmos supporting both channels and objects is a hindrance. A real hindrance would be expecting a movie production's post production team to deliver a mix that's entirely object-based in the very limited amount of time they have to complete the project.
Are we in court now? Like I already mentioned:
- A pure object based format can support "legacy channel mapped sound" via the methods I described earlier. It's also part of the mixing tools responsibility to create "intuitive" interface to do this. If everybody supports this open MDA format, it will eventually benefit those post production teams, because they only need to create a single mix, not countless other, tweaked mixes in whatnot formats.
- Like I also mentioned before, if you want a pure, speaker agnostic or actually a largely sound system agnostic delivery format, you want to get rid of anything specifically targeted at speaker or speaker group A, B, C, etc. Your sound system should map this dynamically, based on the metadata in the format and not because of some assumed speaker layout.
This has nothing to do with allowing others to cheap out, but just with a little bit of visionary that things might change in the future. You can't anticipate all possible scenarios, but you can at least try to look a little bit ahead.
quote: Bobby Henderson
Get traction in the MDA space? Dolby Atmos was in development long before DTS' MDA and has had a product on the market for nearly 3 years. Meanwhile MDA is fixing to get a product into the market.
Yeah, to get traction. Because before Atmos, there was already Iosono (recently acquired by Barco) with working installs (first working install in a theater in 2009) and Imm sound (which Dolby later acquired) with working installs and actual movie titles in the pipeline.
Although none of those systems made a really big impact, they were out there, so Dolby was not a leader in this space for quite a while. Dolby obviously could use their considerable weight to introduce their system to a broader market. Then again, they needed to get traction and did so using a hybrid solution, to lower the bars for post production teams to deliver Atmos mixes.
We're now a few years down the road and production tools continue to evolve. The next generation tools will start supporting object based sound natively. So production of those mixes becomes easier. If you can eventually do just one mix and maybe just one deliverable, that would only help the post production teams to get their work done more efficiently.
quote: Bobby Henderson
An analogy to this would be how Adobe Illustrator accommodates both vector-based artwork and raster-based artwork into the same layout. It would be flat out stupid and counter-productive to make Illustrator a purely vector-only application.
There are valid reasons for providing support for both objects and beds in Atmos. No other sound format appears to be going object-only either.
Let's not overstate what those sound objects actually are. They are just uncompressed audio samples with some metadata on a timeline attached to them. They are not some high level representation of the sound waves that should be rendered. So this analogy only works on a certain level. They're more like bitmapped sprites, which you can freely move around in a 3D space and apply mutations on. And I could stretch it so long and call a bed or stem an object too, because it actually is.
quote: Bobby Henderson
With those sound formats most of the 5.1 surround done wrong has little to do with the sound formats themselves and mostly to do with the theater's speakers, amps, choice of crossover, etc. Another big thing is maintenance. On that level a theater could be screwing over customers by doing mono or stereo sound wrong.
The same for MDA based installs. The quality is dependent on the chosen implementation and maintenance... And here it's also not the failure of the format, but the implementation.
quote: Bobby Henderson
Open MDA doesn't appear to be suggesting any quality standards at all.
No, because it's not the purpose of Open MDA to do this, it's not the purpose of AC3 to suggest quality standards. If there is anyone that can demand a quality standard here, it's DTS for DTS:X and Barco for their implementation of it. If they want some uniformity in quality, maybe they could come up with a new brand and put their collective marketing efforts behind it.
quote: Bobby Henderson
I won't be all that critical of a standard priced theater doing a slight amount of Open MDA enhancement to what is really only a conventional 5.1 surround system as long as I don't have to pay anything extra to see a movie in that room. But if I'm going to be paying $13 or $16 for a ticket and the theater is pulling the same cost cutting bullshit I'll get justifiably angry and not visit that particular theater again.
Isn't this just the general rant against the "Premium Large Format" stuff that's happening all over the globe and which Digital IMAX was at least partly responsible for? People seemingly pay premium prices for a 5.0 sound system too. Nothing is going to stop anybody to build a crappy, "below spec" setup with some big bang logo, brand and marketing campaign calling it the best thing since laser-sliced bread.
quote: Bobby Henderson
Fundamentally 3D object based sound requires enough speakers and amplifiers to properly render the audio. It's no different than a digital image requiring enough pixels of native detail to reach an acceptable level of quality. Some cheap approach that uses hardly any more speakers and amplifiers than a standard 5.1 sound system will be woefully inadequate. There's just no way around that.
I didn't have the time yet to come up with meaningful numbers. But yes, nobody here is denying that you need an adequate number of speakers and amplifiers with all current systems, whatever the adequate number is, depending on room size and chosen implementation.
| IP: Logged
|
|
|
|
All times are Central (GMT -6:00)
|
This topic comprises 8 pages: 1 2 3 4 5 6 7 8
|
Powered by Infopop Corporation
UBB.classicTM
6.3.1.2
The Film-Tech Forums are designed for various members related to the cinema industry to express their opinions, viewpoints and testimonials on various products, services and events based upon speculation, personal knowledge and factual information through use, therefore all views represented here allow no liability upon the publishers of this web site and the owners of said views assume no liability for any ill will resulting from these postings. The posts made here are for educational as well as entertainment purposes and as such anyone viewing this portion of the website must accept these views as statements of the author of that opinion
and agrees to release the authors from any and all liability.
|
Home
Printer-friendly view of this topic