The Atmos films on Bluray I have seen were mostly 4000 - 7000 kbps.
DTS-HD mixes are also 3000 - 5000 kbps for movies with 7.1 soundtracks. So Atmos mixes may need more bandwidth, but not that much.
Looking at the ranges you indicate, a change of 3 to 4 Mbps is an increase of 33% and from 5 to 7 is 40%.
The issue for the studios is not necessarily the bandwidth, but the total payload cost. At an additional 2 Mbps, a 2 hour blockbuster could require substantially more space allocated for Atmos.
A re-mix will need to be done for home Atmos and at an added expense for the studio.
Sorry but unless I greatly misunderstand object based audio (And I fully concede that this may be the case) but there is no 'mix' as such. There are objects with defined meta-data specifying characteristics such as the objects positional location. It is then up to the decoder/rendered/endpoint whatever you wish to call it to produce the 'mix' during playback.
There simply is nothing to re-mix in the traditional sense. As for the 30-40% added overhead for the meta-data I can't comment but it does sound like an awful lot for things like positional meta-data.
Tony,
In the film world the term mix includes both the combination and positioning of the thousands of individual sound elements that go into a sound track. The mixers also balance the relative levels and tone of the channels.
Atmos mixes include, not only the metadata required for positioning the additional Atmos sound elements, but those discrete sound elements as well. Depending on the choices made by the sound designer and director, the number of available channels can extend well beyond the traditional 5.1 or 7.1 we are used to.
In addition, the elements will need to be rebalanced for the smaller home environment. Most home Atmos systems have available, perhaps 4-6 height/ceiling channels rather than the larger numbers available in many cinemas. Relative levels and even positioning may need to be tweaked before new playback rendering metadata can be generated.
My original point is that this adds post production costs to a home media release, in addition to an increase in the audio payload on the discs. It's not always a given that Atmos will be available on a home release, even if it is a prominent part of a film's soundtrack. I know of a few cases of studio resistance, when a home Atmos relese is proposed.
I think the additional audio payload is irrelevant - just watch the extras disappear if the studios think they can make more money out of the release (or releases). The latter will be the sole driver IMHO. Also, we are on the verge of 4k BR so it's difficult to predict the future. My money is on object-based audio becoming the norm.
Main system - Sources (Kodi-HTPC, Vero 4k, FireTV, Sat) into Trinnov Altitude 32 (8-16) (+Roon), Trinnov Amplitude 8 amp & Vivid Audio (Giya G3, V1.5) and REL Stentor III speakers. Others: F80 + Raspberry Pi and other Pi systems.
I am afraid that I am still not getting it. Isn't one of the points of using objects and their corresponding metadata that the processor knows about the speaker layout, distances and configuration and uses this to produce the most accurate/representative output based on the object metadata?
If the number of channels, relative levels and positions need to be compensated for by a remix then what is the point of object based audio? If any 'rebalancing' needs to be done isn't that the job of the end processor? Surely we can't be back to having to produce a mix for each layout and having to tightly control the room and speaker layout to get the best out of a soundtrack. If so what is the point of object based audio if not to get the best out of varying speaker and room layouts and configurations?
At this moment, I'm afraid re-mixes are necessary. I know of 3 films in post that are waiting for the studios to OK a home Atmos, near-field mix.
While the rendering processor might do an adequate job of re-routing the individual assets and combining them into fewer channels, I don't believe they are sophisticated enough to determine if the sources being combined are correlated or uncorrelated. When you combine 2 assets into 1 speaker do you pad by -3dB (uncorrelated) or -6dB (correlated)? That's why the mixers re-balance the audio.
I think the additional audio payload is irrelevant - just watch the extras disappear if the studios think they can make more money out of the release (or releases). The latter will be the sole driver IMHO. Also, we are on the verge of 4k BR so it's difficult to predict the future. My money is on object-based audio becoming the norm.
Very true. They argue what will drive disc sales; more special features or an Atmos mix. I believe Atmos will come into its own when UHD Blu_Ray becomes a reality. They you will have a clear correlation between media and quality: DVD-Dolby digital, Blu-Ray-Dolby TruHD or DTS Master, UHD Blu-Ray-Dolby Atmos.
It's nice to hear some speculation about object based audio support. Meridian's position at the front of the line in cutting edge home theater in the 90's was what first drew me in. This slower cycle was just not them 15-20 years ago:
One of the first (if not the first) with Dolby Digital Again with DTS (when everyone else was using the outboard Millennium decoder) A free upgrade for the 561 being the launch vehicle for PL2.
Sure this was all building on the 8 digital channel support already in the 565, but the 565 itself was also cutting edge and ahead of its time. I'm sure EBA and MQA are wonderful, but I'm glad to have some HT excitement even if its purely speculation!
I'm slightly speculating here ... but I think part of this is a big change in how things are developed and implemented. In the 90s when HT itself was the cutting edge, people like Meridian took off the shelf DSPs and created their own code for processing. The large CE companies had to do likewise.
Now everything is just done with standard code for the likes of Denon and Marantz and Onkyo. And as the major customers they get the code for DSP chips as exclusives before everyone. Considering now there are only 3 major CE brands - Denon and Marantz; Onkyo, Integra and Pioneer and then Yamaha - each one probably has an exclusive from one IC giant. This leaves smaller more HiFi orientated companies such as Meridian, Classe / Rotel, Arcam, etc. having to either wait till the exclusivity period has ended, or developing their own code. Certainly I've heard comments from B&W representatives that they've not had access to the chipsets from their DSP suppliers to enable them to develop Atmos, etc. processors.
Differences from commercial installations Because of limited bandwidth and lack of processing power, Atmos in home theaters is not a real-time mix rendered the same way as in cinemas. The substream is added to Dolby TrueHD or Dolby Digital Plus. This substream only represents a losslessly encoded fully object-based mix. This substream does not include all 128 objects separated. This is not a matrix-encoded channel, but a spatially-encoded digital channel. Atmos in home theaters can support 24.1.10 channel, but it is not an object-based real-time rendering. Filmmakers need to remix and render the TrueHD and Dolby Digital Plus soundtracks with Dolby Media Producer. TrueHD is base with the help of MLP, so it would be not far fetched if Atmos also depends on same codecs with Meridian techteams help.
PerS
Last edited by PerS1; 2015-05-1518:18.
40th.Anv.DSP8000SE &861v8/ID41,Meridian Ultra Dac,C15 Black w/SSD,QNAP TS-669Pro 6X3 Tb w/Core,HD621,Oppo UDP-203 ,JVC DLA-X700R, 2xJL Audio F112v2, Marantz 8802A for Atmos with KEF Reference 5 and 4C withR700 sides & rears and 4xKEF R50 Atmos