I’m not a mixer but I’m a home theater enthusiast. Objects are very useful when listening in a home theater. It makes use of all speakers in that plane no matter if you have a 5,7, or 9 speaker base/bed layer. Same of height speakers. Objects allow the decoder to pan to all speakers in the object path. So objects for sounds that actually move is really important imo.
great video Dave ! honestly thx for putting out content like this, I feel like the world needs more clarifications on dolby in general. So , my take on this concept of the OBEDS is that I use them not because of any sonic differences but because of flexibility and ease. So let's take an example of purely using beds. Beds unfortunately as of now are limited to 7.1.2 (hopefully this changes in the near future :D). So that being said, if we were to bus elements to a 7.1.2 AUX in PT then to the bed or, directly to the bed in dolby, its ,limited to 2 height speakers. This becomes a problem as the output from protools is only sending out information for top left and top right, so even when you monitor in 7.1.4 (we are still inputting 7.1.2) , you will notice that the renderer is just duplicating the same signal for both the top front and top rear because it has no way of knowing the discrete information since its only outputting 2 heights. So , thats why most mixers would have to reach for objects when panning information on the heights especially if there are more than just 2 height speakers. So long story short is that I'd rather have an Object BED so I have full access to panning as much elements as I need to and have discrete spatialising possibilities of all 4 or 6 heights. This example can also be applied when we start to mix in formats that got beyond bed capabilities (7.1.2). Instead of sacrificing objects which can be very limited especially when working on films , it's more beneficial to have OBEDS set up to represent the format and mix through that. I hope this makes sense, but I also love your explanation on it and I totally agree with you in the sense that there is absolutely no sonic differences.
Thanks! I'm glad you liked the video. One of the things I forgot to mention about the height in the bed is, height speakers are also part of an array in a larger venue which is why it is .2 in the bed and will probably always be .2. I use a quad set of objects for hitting the corners so I guess I do use a little bit of a OBED. :)
@@goingto11 I don't get that part. If you have a speaker array, doesn't that array get the same signal anyway the way it is connected to and configured in the processor? That's why it is setup like an array? Otherwise, you could just set it up as individual speaker channels and it would work perfectly fine with Atmos objects (being able to pan across the individual speakers)? As I commented in a separate comment, 7.1.2 mixes are detrimental to all larger systems.
really good points. I use Object Beds For time based EFX. and when I want to pinpoint a sound/ EFX in space. An Example is I can set up a 4-point object bed and insert a quad reverb. Then place that 4-point bed anywhere in the space, Beds also Blead into other speakers, But like you said making them very important also.
I was going to make a comment about separating the FX and objects, due to compatibility with certain plug-ins only being up 7.1.2 or lower. In my current workflow, I used OBEDs for "dry" instruments, typically without reverb or delays, and then used the FX on the Dolby Bed. Especially being able to hear the before and after just at the click of muting the objects and default bed within the Atmos renderer is oddly super fun to me!! It will definitely be a consideration in my own mixes from here on out. I don't disagree with anything within this video, but I wonder how far we have to go to make sure how our mix translates to these mediums and still respecting the artists intent. This is where I think the idea of being a "mix engineer" in 2024 REALLY needs to be stressed to even bedroom producers. Because a speaker array system having single instruments coming from multiple speakers on the same side with the same content seems like it could create a different intent than what the artist had when making the song originally. A listening party of their song for the first time in a theater compared to in headphones or in a smaller Atmos studio could lead the artists to ask questions, such as "why does it sound like that??" And maybe this is just me, but if I go to see a movie in theaters, I really try to get in the center of the action as much as possible. If you're sitting where stuff is "hard panned" to the left or right side, you're still potentially missing some information that would otherwise be different than if you were sitting in the "sweet-spot", eventually giving you a different experience to everyone in the theater regardless. Hopefully the artist would be the one in the midst of the action hearing their album to the highest quality and intent as both them and mixing engineer had. But like you mentioned, could be a REALLY different experience for other sections of the room depending on how this mix was approached. I'm not an expert on speaker arrays or theater sound setup, this is all coming from my experiences of being a very non-religious movie goer. So please correct me if I am wrong on these ideas about public theaters and speaker arrays (would be a great video!!) But I am also making comparison as if headphones and a decent 9.1.4 setup are anything close to being similar (as of 2024). There is no comparison how good one is over the other, but we still have to consider it in translation. We haven't even scratched the surface when it comes to the quality and size of movie theaters that aren't 100% up to the Dolby Spec, but are the only ones available for the artist. I would pay good money to see/hear this demonstration and comparison in person, I feel like this would be a great educational experience for newcomers into Dolby Atmos!! Great video Dave.
Excellent video! When mixing music I use the 7.1.2 bed with a few objects as well. Sometimes I do find in Apple playback panning objects the relationship between the bed and the objects shifts depending on panning, which can be annoying.
Great video, sir! Object beds are pretty fascinating- but I’m with you. Utilizing the beds makes better use of the arrays. If future-proofing a mix is the goal, the trend is that engineers and listeners will have more speakers in the future rather than fewer.
I really love your vids David thank so much for taking the time to share your knowledge. I’m from Costa Rica and highly appreciate all the info you bring to the table jejeje In regarding the usage of OBEDS I would like to say that I started to use them about 2 months from now and I quit to them because I was having the problem that ok I mix using the obeds but later I won’t have option to manage the binaural settings of those instrument but just for the bed which IMO limits a lot of the creative stuff you can get in binaural using different distances…. Is funny what you mentioned about the adm Size cause indeed that was one of the first things that came up to visibility all of the sudden my files increased the size and sometimes I created 2 obeds but for some reason actually used 1 obedient for the whole mix so I was creating adm with a phantom OBED doing NOTHING jajajaja so yeah I agree that the upload of those files should be charged based on the size otherwise people won’t care and like I my case I sent 1 or 2 files like that I then realized it was a waste of space so fixed it for the rest of files I created ever since….
Thanks Dave - your videos are always superb references! BTW: The bed still is 7.1.2 , what is the best way to access the 4 top channels individually? Im not on the latest PT version. Thanks!
@@goingto11you can access all 6 ceiling positions with an object bed. You can also access the wide speakers by placing a stereo object about 25% back from the front toward the side. I mix into a full 9.1.6 object bed for stationary instruments even though my actual array is 5.1.2 I then use objects for anything that moves and the bed to access the LFE channel.
Thanks for the great video! I have a question regarding the difference between a bed and object-based audio in Dolby Atmos, especially when considering playback on AirPods and Apple Music’s Spatial Audio. While I understand that using a bed might be better for theaters because more energy is coming from the speakers - I’m curious about the potential impact on headphones. When panning between speakers or placing sound within the virtual space, does using a bed introduce more phase issues or «smearing» compared to using object-based audio? I imagine that objects, with their precise control over spatial placement, might reduce phase artifacts in a headphone environment, whereas a bed might lead to less accurate spatial imaging. What is your take on that?
Good question. If I am panning between 7.1.4 speaker locations, I will use an object because you can have phase issues in binaural if you have the same information in multiple bed channels. Objects sound smoother to me in headphones when panning in between speaker locations. But for static panning to a speaker position, it's sonically the same if you use an object or a bed channel.
Hi Dave, great channel. A coworker recently turned me on to it and this is the best immersive mixing content out there. Keep up the great work! I recently had the opportunity to hear a mix I did using a 9.6 Object Bed in a 15.8 room. I immediately noticed the same thing you express in this video - some speakers (the ones that didn't have Objects snapped to them) were not engaged in the playback. I don't know if I would have noticed a difference if I wasn't staring straight at the Renderer (and its unused output meters), but the mix felt thin to me in that space. So in terms of scaling, I sort of came to the same conclusion - that the Object Bed method is inferior for scaling UP. You mentioned in the video that there is no sonic difference between using the Bed and an Object Bed. I assume you are referring to when you A/B on your speakers. Is there any discernible difference in the downmix? Do you get different results using the Bed vs. Object Bed if you create a 5.1, stereo, or binaural re-render? We've established that Object Bed mixes may not scale UP as well, but what about the more likely scenario of scaling DOWN? Another aspect to consider is the LFE. Since the LFE is not directly addressable via an Object, the only way to get something into the LFE is to use the Bed. In my example, I used an aux send for the LFE and it returns into the LFE channel of the Bed. So in that scenario the Bed was not completely empty. It just had one the one LFE channel. Thanks again!
Thanks! I'm glad you like the videos. In terms of sonic difference, I don't find there to be a sonic difference between Objects and Beds on my speakers or in any fold down situations. There shouldn't be any sonic difference between the bed and an object bed for fold downs so scaling down shouldn't be an issue which is good, I think, since the more likely scenario for music is that mixes will be scaled down.
Thanks so much for another outstanding video, Dave! One thing I'm not clear about: in the context of arrays and "coverage," how/why does the Bed interpret fixed pan positions differently than the same fixed pan positions in an OBED? This would seem to be at odds with the accepted language about the Bed, which is that its LCR, etc. channels are really just like objects that have fixed, unalterable panning data. I hope this makes sense, and sorry if I missed something. Thanks again for sharing your experience!
Well, the bed channels are like objects in one sense, but the bed is also more like a traditional multi-channel format like 5.1 and 7.1. I think this doesn't get talked about much when it comes to music production because it's generally not as applicable as it is in the audio post world. I'm actually working on a follow-up video that will hopefully demonstrate all of this a little clearer.
Thanks for this! Could you please do a video on how the different track widths are compatible with each other? How do I make my Atmos mix (done in my 7.1.4 environment) utilize all the speakers in a larger, e.g. 9.1.6 environment? Cheers! 👍
I'm not quite sure what you're asking about track widths and compatibility, but this video might explain a bit about utilizing speakers in a larger environment than the one you are mixing in: th-cam.com/video/eDGCjoet5Kk/w-d-xo.html
@@goingto11 Thanks! Sorry for being unclear, I've watched the linked video earlier and if I understand you correctly, you set up a 9.1.6 track width bus which is then routed to your 7.1.4 Atmos outputs? And the Wide speakers are panned exactly 68 % in the F/R relationship? What makes this a bit confusing is two-fold. 1) I don't use Pro Tools so names of things in the DAW don't always match up and 2) I don't have access to a 9.1.6 room to verify that it actually works. :) But cool, I'll implement it accordingly in Cubase. Cheers!
Well, maybe one thing I should clear up is I'm not feeding an Atmos output--in some ways that doesn't exist. What I'm doing is feeding into the Atmos renderer which then feeds my monitors based on the configuration of the monitors which, in my case, is currently 7.1.4. That aside, I'm not familiar with Cubase, but there are probably a couple of approaches you could use to hit the wides. The first would be to use a 9.x.x bus, and then figure out the panning to hit those specific channels. If there's a meter for that bus, that should help you figure out the values. Another way some of us have done it in the past is to just use a stereo bus that is assigned to a pair of objects that are panned to the wide positions. Then when you want to hit the wides you just send to that bus. You'd still have to figure out the panning value for Cubase, though.
INTERESTING! So there must be some voodoo going on in that cinema or other commercial Atmos cinemas because that is NOT the way things mixed in the BED are reproduced in any home cinema, even the largest and most insane ones that use a Trinnov or Storm. It seems weird that the renderer is showing the balls with halo around them as if the “size” parameter has been engaged on everything. I’m gonna get a discussion going in our Atmos mixing group about this because this is a very interesting topic. Some things to discuss here: A) The Trinnov placement guide seems a bit misleading because they list the multiple side surrounds (LS, LS1, LS2) as an array. As you know an array is the same information duplicated through multiple speakers. In the Trinnov you can set up of those 3 left surround speakers as EITHER an array or discrete speakers which can then only be used if an object pans through. Something panned to Left Surround in the Bed will only come out of LS. If you want something panned to LS in the Bed to come out of LS, LS1, LS2, you’d need to switch to a different speaker preset in the Trinnov that has them set as an array. It will attenuate each speaker 3db and play that same sound out of all 3 speakers but then you lose the ability of an object panning along the wall to transition smoothly from speaker to speaker like it would in the regular preset where they are discrete speakers. Something very weird is going on in that theater you went to (side note, that’s amazing you got to play back some of your mixes in that type of space!) but from your pictures it looks like it can instantaneously decide when something is a bed and duplicate that sound in an array fashion or if it see’s it’s an object it places it discreetly to the one speaker. Also weird that in your pic when LS, LS1, LS2 are all showing on the meter, they are different level. You would think if it was array they would all be getting equal volume. When I was visiting LA last month, I was lucky enough to get to tour the Sony lot with Marc Fishman, and they showed us their Atmos mixing stages there which are basically small to medium size commercial cinemas. They have so many speakers, I counted 14 height speakers in one of the rooms! I asked if they are using any arrays and he said no, they are all set as discrete speakers. So they are not mixing with arrays. So again I’m wondering what the hell is happening in that theater!!! In any case it shows that it’s very limited to that theater or other commercial theaters but no device that would be used to process the home theater mix will behave that way so it’s a very limited thing for people to get music mixes played in a theater. Bigger point is it has me wondering what’s going on with decoding in theatrical? Even though the mixing stages have discrete speakers, is it doing the same thing when they’re mixing where it has some sort of ability to “on the fly” place a bed sound in the entire array and keep an object locked to discrete speakers? I’m definitely gonna follow up with Marc about that! B) Object Beds that you mention in this video seems like there is no winner other than for workflow but let’s talk about mixing in the Bed vs Object specifically for home theaters. Movie mixes like Ready Player One which are mixed as Bed only are stuck in 7.1.2. In every AVR or processor if you have 4 overhead or heights speakers, it takes the left top and duplicates that sound to Left Top Front and Left Top Rear. Same for Top right. So that way Bed height information will come out of all 4 of your top speakers but you still have discrete movement or placement for objects. If you have 6 overhead speakers though, Bed only height info will only come out of top middle. Most people find this frustrating and make a separate preset for movies like this that removes their top middles and uses only 4 height speakers so it makes that bed info cover a larger area. Also, anything mixed in Bed only will NOT use anything beyond the 7 ear layer speakers. So if you have front wides, extra surrounds, etc, BED information will NOT play out of those speakers. You would need to do something similar like with the heights where you make a separate preset where those extra speakers are setup as an array. Only the Trinnov and Storm can do that since in most AVR’s and processors arrays have to be setup manually and by sending that channel to an external device to duplicate it and time align and set the volume. So for home theaters that have let’s say an 11.1.6 setup with extra side surrounds and front wides, they really miss out on using all their speakers on mixes that are bed only. Thankfully, most mixes mixes use a combination of both where things that aren’t moving will be Bed but anything that has to move around in any way should be encoded as an object so all those speakers will get used. So to bring it back to your video, on an 11.1.6 home setup, something mixed in bed only or using object bed will sound identical. In fact the object bed will probably work better for the heights since that will have 4 discrete height locations rather than 2. Also, as I proof read this whole thing, it made realize that if every AVR can recognize BED height info and duplicate it on the fly, which couldn’t that be done with the surrounds or wides? That fact that even the Trinnov or Storm can’t do it on the fly tells you it must take some serious processing power. So maybe only in theatrical Atmos and mixing stages do they have processors that have the ability to see if something is bed or object and send it to the speakers as an “array” or discrete object.
Question on this discussion. Say I have a 5.1.4 setup in my room, and I make an Atmos mix using a 7.1.2 bed. I know with objects they will just get positioned in the room as best they can using the 5.1.4, but my question is about if you pan things to the side of the BED. Will audio panned to the sides in the 7.1.2 bed only come out of the Surrounds? Or will the renderer attempt to make a "phantom" side/rear and render thigns panned to the sides in a 7.1.2 bed between the fronts and surrounds, and then only things panned to the surrounds play 100% from the sides? Or, in this configuration, will audio panned to the sides of a 7.1.2 bed sound identical to audo placed on the surrounds? Thanks!
In theory, the trim/downmix settings should adjust how the side content is handled when you're monitoring a 5.1.4 downmix. "Direct Render" will create a phantom center between the front and surrounds while "Direct Render with Room Balance" will put side information into the surrounds. Personally, I prefer Direct Render with Room Balance but your mileage may vary.
I’m not a mixer but I’m a home theater enthusiast. Objects are very useful when listening in a home theater. It makes use of all speakers in that plane no matter if you have a 5,7, or 9 speaker base/bed layer. Same of height speakers. Objects allow the decoder to pan to all speakers in the object path.
So objects for sounds that actually move is really important imo.
great video Dave ! honestly thx for putting out content like this, I feel like the world needs more clarifications on dolby in general. So , my take on this concept of the OBEDS is that I use them not because of any sonic differences but because of flexibility and ease. So let's take an example of purely using beds. Beds unfortunately as of now are limited to 7.1.2 (hopefully this changes in the near future :D). So that being said, if we were to bus elements to a 7.1.2 AUX in PT then to the bed or, directly to the bed in dolby, its ,limited to 2 height speakers. This becomes a problem as the output from protools is only sending out information for top left and top right, so even when you monitor in 7.1.4 (we are still inputting 7.1.2) , you will notice that the renderer is just duplicating the same signal for both the top front and top rear because it has no way of knowing the discrete information since its only outputting 2 heights. So , thats why most mixers would have to reach for objects when panning information on the heights especially if there are more than just 2 height speakers. So long story short is that I'd rather have an Object BED so I have full access to panning as much elements as I need to and have discrete spatialising possibilities of all 4 or 6 heights. This example can also be applied when we start to mix in formats that got beyond bed capabilities (7.1.2). Instead of sacrificing objects which can be very limited especially when working on films , it's more beneficial to have OBEDS set up to represent the format and mix through that. I hope this makes sense, but I also love your explanation on it and I totally agree with you in the sense that there is absolutely no sonic differences.
Thanks! I'm glad you liked the video.
One of the things I forgot to mention about the height in the bed is, height speakers are also part of an array in a larger venue which is why it is .2 in the bed and will probably always be .2. I use a quad set of objects for hitting the corners so I guess I do use a little bit of a OBED. :)
@@goingto11 I don't get that part. If you have a speaker array, doesn't that array get the same signal anyway the way it is connected to and configured in the processor? That's why it is setup like an array? Otherwise, you could just set it up as individual speaker channels and it would work perfectly fine with Atmos objects (being able to pan across the individual speakers)? As I commented in a separate comment, 7.1.2 mixes are detrimental to all larger systems.
Thanks Dave, I feel like yours are the most practical Atmos videos I watch, all the time.
Wow, thanks!
Next up kids will be making: Fart sounds in Dolby layer cake.. er Sound Bed.. 🤣🤣🤣
really good points. I use Object Beds For time based EFX. and when I want to pinpoint a sound/ EFX in space. An Example is I can set up a 4-point object bed and insert a quad reverb. Then place that 4-point bed anywhere in the space, Beds also Blead into other speakers, But like you said making them very important also.
Thanks Dave for sharing !!! So valuable!! Have a good weekend !
Thankyou Sir... 🙏🙏🙏
Excellent Video... 👌👌👌❤❤❤
I was going to make a comment about separating the FX and objects, due to compatibility with certain plug-ins only being up 7.1.2 or lower. In my current workflow, I used OBEDs for "dry" instruments, typically without reverb or delays, and then used the FX on the Dolby Bed. Especially being able to hear the before and after just at the click of muting the objects and default bed within the Atmos renderer is oddly super fun to me!! It will definitely be a consideration in my own mixes from here on out.
I don't disagree with anything within this video, but I wonder how far we have to go to make sure how our mix translates to these mediums and still respecting the artists intent. This is where I think the idea of being a "mix engineer" in 2024 REALLY needs to be stressed to even bedroom producers. Because a speaker array system having single instruments coming from multiple speakers on the same side with the same content seems like it could create a different intent than what the artist had when making the song originally. A listening party of their song for the first time in a theater compared to in headphones or in a smaller Atmos studio could lead the artists to ask questions, such as "why does it sound like that??" And maybe this is just me, but if I go to see a movie in theaters, I really try to get in the center of the action as much as possible. If you're sitting where stuff is "hard panned" to the left or right side, you're still potentially missing some information that would otherwise be different than if you were sitting in the "sweet-spot", eventually giving you a different experience to everyone in the theater regardless. Hopefully the artist would be the one in the midst of the action hearing their album to the highest quality and intent as both them and mixing engineer had. But like you mentioned, could be a REALLY different experience for other sections of the room depending on how this mix was approached. I'm not an expert on speaker arrays or theater sound setup, this is all coming from my experiences of being a very non-religious movie goer. So please correct me if I am wrong on these ideas about public theaters and speaker arrays (would be a great video!!)
But I am also making comparison as if headphones and a decent 9.1.4 setup are anything close to being similar (as of 2024). There is no comparison how good one is over the other, but we still have to consider it in translation. We haven't even scratched the surface when it comes to the quality and size of movie theaters that aren't 100% up to the Dolby Spec, but are the only ones available for the artist. I would pay good money to see/hear this demonstration and comparison in person, I feel like this would be a great educational experience for newcomers into Dolby Atmos!! Great video Dave.
Excellent video! When mixing music I use the 7.1.2 bed with a few objects as well. Sometimes I do find in Apple playback panning objects the relationship between the bed and the objects shifts depending on panning, which can be annoying.
Great video, sir! Object beds are pretty fascinating- but I’m with you. Utilizing the beds makes better use of the arrays. If future-proofing a mix is the goal, the trend is that engineers and listeners will have more speakers in the future rather than fewer.
I hope they have more speakers, but unfortunately I think a lot of people would disagree with us.
I really love your vids David thank so much for taking the time to share your knowledge. I’m from Costa Rica and highly appreciate all the info you bring to the table jejeje
In regarding the usage of OBEDS I would like to say that I started to use them about 2 months from now and I quit to them because I was having the problem that ok I mix using the obeds but later I won’t have option to manage the binaural settings of those instrument but just for the bed which IMO limits a lot of the creative stuff you can get in binaural using different distances…. Is funny what you mentioned about the adm Size cause indeed that was one of the first things that came up to visibility all of the sudden my files increased the size and sometimes I created 2 obeds but for some reason actually used 1 obedient for the whole mix so I was creating adm with a phantom OBED doing NOTHING jajajaja so yeah I agree that the upload of those files should be charged based on the size otherwise people won’t care and like I my case I sent 1 or 2 files like that I then realized it was a waste of space so fixed it for the rest of files I created ever since….
Thanks Dave - your videos are always superb references! BTW: The bed still is 7.1.2 , what is the best way to access the 4 top channels individually? Im not on the latest PT version. Thanks!
Thanks!
I use a quad set of objects to hit the 4 height positions. So, I guess I do use a bit of an object bed when I'm mixing. :)
@@goingto11you can access all 6 ceiling positions with an object bed. You can also access the wide speakers by placing a stereo object about 25% back from the front toward the side.
I mix into a full 9.1.6 object bed for stationary instruments even though my actual array is 5.1.2 I then use objects for anything that moves and the bed to access the LFE channel.
Thanks for the great video! I have a question regarding the difference between a bed and object-based audio in Dolby Atmos, especially when considering playback on AirPods and Apple Music’s Spatial Audio. While I understand that using a bed might be better for theaters because more energy is coming from the speakers - I’m curious about the potential impact on headphones. When panning between speakers or placing sound within the virtual space, does using a bed introduce more phase issues or «smearing» compared to using object-based audio? I imagine that objects, with their precise control over spatial placement, might reduce phase artifacts in a headphone environment, whereas a bed might lead to less accurate spatial imaging. What is your take on that?
Good question. If I am panning between 7.1.4 speaker locations, I will use an object because you can have phase issues in binaural if you have the same information in multiple bed channels. Objects sound smoother to me in headphones when panning in between speaker locations. But for static panning to a speaker position, it's sonically the same if you use an object or a bed channel.
Hi Dave, great channel. A coworker recently turned me on to it and this is the best immersive mixing content out there. Keep up the great work! I recently had the opportunity to hear a mix I did using a 9.6 Object Bed in a 15.8 room. I immediately noticed the same thing you express in this video - some speakers (the ones that didn't have Objects snapped to them) were not engaged in the playback. I don't know if I would have noticed a difference if I wasn't staring straight at the Renderer (and its unused output meters), but the mix felt thin to me in that space. So in terms of scaling, I sort of came to the same conclusion - that the Object Bed method is inferior for scaling UP.
You mentioned in the video that there is no sonic difference between using the Bed and an Object Bed. I assume you are referring to when you A/B on your speakers. Is there any discernible difference in the downmix? Do you get different results using the Bed vs. Object Bed if you create a 5.1, stereo, or binaural re-render? We've established that Object Bed mixes may not scale UP as well, but what about the more likely scenario of scaling DOWN?
Another aspect to consider is the LFE. Since the LFE is not directly addressable via an Object, the only way to get something into the LFE is to use the Bed. In my example, I used an aux send for the LFE and it returns into the LFE channel of the Bed. So in that scenario the Bed was not completely empty. It just had one the one LFE channel. Thanks again!
Thanks! I'm glad you like the videos.
In terms of sonic difference, I don't find there to be a sonic difference between Objects and Beds on my speakers or in any fold down situations. There shouldn't be any sonic difference between the bed and an object bed for fold downs so scaling down shouldn't be an issue which is good, I think, since the more likely scenario for music is that mixes will be scaled down.
Thanks so much for another outstanding video, Dave! One thing I'm not clear about: in the context of arrays and "coverage," how/why does the Bed interpret fixed pan positions differently than the same fixed pan positions in an OBED? This would seem to be at odds with the accepted language about the Bed, which is that its LCR, etc. channels are really just like objects that have fixed, unalterable panning data. I hope this makes sense, and sorry if I missed something. Thanks again for sharing your experience!
Well, the bed channels are like objects in one sense, but the bed is also more like a traditional multi-channel format like 5.1 and 7.1. I think this doesn't get talked about much when it comes to music production because it's generally not as applicable as it is in the audio post world. I'm actually working on a follow-up video that will hopefully demonstrate all of this a little clearer.
Excellent - thanks, Dave! Appreciate the reply
Thanks for this! Could you please do a video on how the different track widths are compatible with each other? How do I make my Atmos mix (done in my 7.1.4 environment) utilize all the speakers in a larger, e.g. 9.1.6 environment? Cheers! 👍
I'm not quite sure what you're asking about track widths and compatibility, but this video might explain a bit about utilizing speakers in a larger environment than the one you are mixing in: th-cam.com/video/eDGCjoet5Kk/w-d-xo.html
@@goingto11 Thanks! Sorry for being unclear, I've watched the linked video earlier and if I understand you correctly, you set up a 9.1.6 track width bus which is then routed to your 7.1.4 Atmos outputs? And the Wide speakers are panned exactly 68 % in the F/R relationship? What makes this a bit confusing is two-fold. 1) I don't use Pro Tools so names of things in the DAW don't always match up and 2) I don't have access to a 9.1.6 room to verify that it actually works. :) But cool, I'll implement it accordingly in Cubase. Cheers!
Well, maybe one thing I should clear up is I'm not feeding an Atmos output--in some ways that doesn't exist. What I'm doing is feeding into the Atmos renderer which then feeds my monitors based on the configuration of the monitors which, in my case, is currently 7.1.4.
That aside, I'm not familiar with Cubase, but there are probably a couple of approaches you could use to hit the wides. The first would be to use a 9.x.x bus, and then figure out the panning to hit those specific channels. If there's a meter for that bus, that should help you figure out the values. Another way some of us have done it in the past is to just use a stereo bus that is assigned to a pair of objects that are panned to the wide positions. Then when you want to hit the wides you just send to that bus. You'd still have to figure out the panning value for Cubase, though.
INTERESTING! So there must be some voodoo going on in that cinema or other commercial Atmos cinemas because that is NOT the way things mixed in the BED are reproduced in any home cinema, even the largest and most insane ones that use a Trinnov or Storm. It seems weird that the renderer is showing the balls with halo around them as if the “size” parameter has been engaged on everything. I’m gonna get a discussion going in our Atmos mixing group about this because this is a very interesting topic. Some things to discuss here:
A) The Trinnov placement guide seems a bit misleading because they list the multiple side surrounds (LS, LS1, LS2) as an array. As you know an array is the same information duplicated through multiple speakers. In the Trinnov you can set up of those 3 left surround speakers as EITHER an array or discrete speakers which can then only be used if an object pans through. Something panned to Left Surround in the Bed will only come out of LS. If you want something panned to LS in the Bed to come out of LS, LS1, LS2, you’d need to switch to a different speaker preset in the Trinnov that has them set as an array. It will attenuate each speaker 3db and play that same sound out of all 3 speakers but then you lose the ability of an object panning along the wall to transition smoothly from speaker to speaker like it would in the regular preset where they are discrete speakers. Something very weird is going on in that theater you went to (side note, that’s amazing you got to play back some of your mixes in that type of space!) but from your pictures it looks like it can instantaneously decide when something is a bed and duplicate that sound in an array fashion or if it see’s it’s an object it places it discreetly to the one speaker. Also weird that in your pic when LS, LS1, LS2 are all showing on the meter, they are different level. You would think if it was array they would all be getting equal volume. When I was visiting LA last month, I was lucky enough to get to tour the Sony lot with Marc Fishman, and they showed us their Atmos mixing stages there which are basically small to medium size commercial cinemas. They have so many speakers, I counted 14 height speakers in one of the rooms! I asked if they are using any arrays and he said no, they are all set as discrete speakers. So they are not mixing with arrays. So again I’m wondering what the hell is happening in that theater!!! In any case it shows that it’s very limited to that theater or other commercial theaters but no device that would be used to process the home theater mix will behave that way so it’s a very limited thing for people to get music mixes played in a theater. Bigger point is it has me wondering what’s going on with decoding in theatrical? Even though the mixing stages have discrete speakers, is it doing the same thing when they’re mixing where it has some sort of ability to “on the fly” place a bed sound in the entire array and keep an object locked to discrete speakers? I’m definitely gonna follow up with Marc about that!
B) Object Beds that you mention in this video seems like there is no winner other than for workflow but let’s talk about mixing in the Bed vs Object specifically for home theaters. Movie mixes like Ready Player One which are mixed as Bed only are stuck in 7.1.2. In every AVR or processor if you have 4 overhead or heights speakers, it takes the left top and duplicates that sound to Left Top Front and Left Top Rear. Same for Top right. So that way Bed height information will come out of all 4 of your top speakers but you still have discrete movement or placement for objects. If you have 6 overhead speakers though, Bed only height info will only come out of top middle. Most people find this frustrating and make a separate preset for movies like this that removes their top middles and uses only 4 height speakers so it makes that bed info cover a larger area. Also, anything mixed in Bed only will NOT use anything beyond the 7 ear layer speakers. So if you have front wides, extra surrounds, etc, BED information will NOT play out of those speakers. You would need to do something similar like with the heights where you make a separate preset where those extra speakers are setup as an array. Only the Trinnov and Storm can do that since in most AVR’s and processors arrays have to be setup manually and by sending that channel to an external device to duplicate it and time align and set the volume. So for home theaters that have let’s say an 11.1.6 setup with extra side surrounds and front wides, they really miss out on using all their speakers on mixes that are bed only. Thankfully, most mixes mixes use a combination of both where things that aren’t moving will be Bed but anything that has to move around in any way should be encoded as an object so all those speakers will get used.
So to bring it back to your video, on an 11.1.6 home setup, something mixed in bed only or using object bed will sound identical. In fact the object bed will probably work better for the heights since that will have 4 discrete height locations rather than 2.
Also, as I proof read this whole thing, it made realize that if every AVR can recognize BED height info and duplicate it on the fly, which couldn’t that be done with the surrounds or wides? That fact that even the Trinnov or Storm can’t do it on the fly tells you it must take some serious processing power. So maybe only in theatrical Atmos and mixing stages do they have processors that have the ability to see if something is bed or object and send it to the speakers as an “array” or discrete object.
Question on this discussion. Say I have a 5.1.4 setup in my room, and I make an Atmos mix using a 7.1.2 bed. I know with objects they will just get positioned in the room as best they can using the 5.1.4, but my question is about if you pan things to the side of the BED. Will audio panned to the sides in the 7.1.2 bed only come out of the Surrounds? Or will the renderer attempt to make a "phantom" side/rear and render thigns panned to the sides in a 7.1.2 bed between the fronts and surrounds, and then only things panned to the surrounds play 100% from the sides? Or, in this configuration, will audio panned to the sides of a 7.1.2 bed sound identical to audo placed on the surrounds? Thanks!
In theory, the trim/downmix settings should adjust how the side content is handled when you're monitoring a 5.1.4 downmix. "Direct Render" will create a phantom center between the front and surrounds while "Direct Render with Room Balance" will put side information into the surrounds. Personally, I prefer Direct Render with Room Balance but your mileage may vary.
I also don't use, it's basically making an object based technology speaker agnostic. It's for people that don't understand the bed, or how to use it.
L A Y E R C A K E ;)