The No.1 Website for Pro Audio
Mixing for Immersive Audio
Old 4 weeks ago
  #1
Gear Maniac
 

Mixing for Immersive Audio

I just read the following in an article detailing a Skywalker immersive audio recording project https://www.mixonline.com/recording/...erican-soldier

"The project was originally planned for 7.1.4 release in Dolby Atmos, but as some consumer Atmos-capable receivers downsample 7.1.4 to 5.1.4 automatically without a native way to play the side speakers, Merceruio and Jones decided to play it safe and mix in 5.1.4."

I really don't like generic folddowns in consumer audio systems - whether it is 7.1.4 to 5.1.4, or more significantly folddown to 2ch and to binaural. With movie soundtracks, they can probably fend for themselves and the rules/expectations are different anyway because of how they do dialogue, atmosphere, music and effects. With music I am a lot less than happy.

Judging by what I have heard on domestic Atmos systems without upper satellite speakers, we are safest mixing to 5.1 - if only to avoid generic folddown messing with what we have set up in a recording. Flavour of the month seems to be processing mixes for binaural and much of what I have hear has been phasy and peculiar - with strange effects from folddown of anything in height-channels.

Anyone else experiencing immersive? Nearly everything I have heard so far in a domestic environment has been via soundbars or the Amazon Echo Studio which seem to work better with movies than with music.
Old 4 weeks ago
  #2
Lives for gear
I suppose immersive by definition puts the listener in the centre of the action, or in this case upon the conductor's podium, as shown in that pencil sketch. Given the precision of that sketch I'd imagine it wouldn't be too difficult to replicate both horizontal location and height placements of instruments around the conductor with careful mic choice, orientation and mixing.

I'd have imagined a Soundfield type of array might have been a typical approach in this situation...but clearly it isn't.

At least in this situation the conductor is truly in the middle of the ensemble, rather than occupying the boundary between audience and musicians as they do in traditional concert setups. With that in mind, if the camera work matches the immersive miking, there's a reasonable chance the audio and visual could act in complimentary fashion to one another to make for a congruous cinematic experience (at least in the movie theatre, if not at home)

How well it might fold down to 2 channel stereo is anybody's guess...a dedicated stereo mix (and allocation of same to a dedicated layer on a BluRay or DVD) would seem the best approach.

This approach could work, due to the conductor-centric miking, where other surround recordings of typical concerts fail...by trying to put the listener in the centre of the orchestra, rather than offering an enhanced audience perspective recording.
Old 4 weeks ago
  #3
Quote:
Originally Posted by TonyF View Post
I just read the following in an article detailing a Skywalker immersive audio recording project https://www.mixonline.com/recording/...erican-soldier

"The project was originally planned for 7.1.4 release in Dolby Atmos, but as some consumer Atmos-capable receivers downsample 7.1.4 to 5.1.4 automatically without a native way to play the side speakers, Merceruio and Jones decided to play it safe and mix in 5.1.4."

I really don't like generic folddowns in consumer audio systems - whether it is 7.1.4 to 5.1.4, or more significantly folddown to 2ch and to binaural. With movie soundtracks, they can probably fend for themselves and the rules/expectations are different anyway because of how they do dialogue, atmosphere, music and effects. With music I am a lot less than happy.

Judging by what I have heard on domestic Atmos systems without upper satellite speakers, we are safest mixing to 5.1 - if only to avoid generic folddown messing with what we have set up in a recording. Flavour of the month seems to be processing mixes for binaural and much of what I have hear has been phasy and peculiar - with strange effects from folddown of anything in height-channels.

Anyone else experiencing immersive? Nearly everything I have heard so far in a domestic environment has been via soundbars or the Amazon Echo Studio which seem to work better with movies than with music.


There is a second half to that quote though - โ€œThat changed the mix flow slightly because we had recorded in and were listening to it with side speakers,โ€ says Jones. โ€œWhen we mixed it, we didnโ€™t have the side speakers, so I had to create that phantom image.โ€

That doesn't make a whole lot of sense to me, as a Dolby Atmos master doesn't contain any speaker channel information. There is no 5.1.4 deliverable. A 7.1.4 system will still interpret the side information from the master, and that 'phantom side' will no longer be phantom at all. It really only ramps up from there too, as Atmos can support much, much bigger speaker configurations than 7.1.4.

Could that possibly contribute to the differences you are hearing between film and music? Larger budget Atmos films fully embrace the methodology. Regardless of whether the film might be viewed on headphones, or stereo, or 5.1 surround, it is pretty common for them to be mixed on dub stages with speaker configurations that are much larger than the 7.1.4 minimum certification requirement. Sometimes as many as 64 speaker channels.

It is hard to imagine that mixing with such a higher degree of spatial resolution, rather than relying on spatialisation through phantom imagining, wouldn't result in a more phase coherent mix and master across multiple formats. Could the biggest difference between film and music in Atmos simply be the degree of trepidation in trusting the platform to do what it is designed to do? For the most part, it seems like a lot of film rerecording mixes don't even bother doing manual 5.1 or stereo fold downs anymore. They just leave it to the Atmos renderer to take care of.
Old 4 weeks ago
  #4
Lives for gear
 
JCBigler's Avatar
When I was out on tour with the Blue Man Group North American tour this past year, before the Coronapocalypse shut everything down, we were running an L-Acoustics L-ISA immersive audio system.

A bit different than what I think you are looking for, since it is a live PA system rather than a studio or film application. But when I did my L-ISA training in Westlake, CA one of our class projects was the Ibert Flute Concerto. Using a system like this for live orchestra is a very ear-opening experience and gives you a level of realism in the playback of the system that L-R or LCR can't approach.

On Blue Man, we had a main system of 5 arrays for the frontal system which output the immersive information, plus a dual cardioid subwoofer hang at center. We also had front fills, outfills and delay outputs when needed which were matrixed on our console.

One of the main benefits to the immersive system the way L-Acoustics implements it is that it puts a larger percentage of the audience inside the immersive zone, as it spreads the sweet spot out over a larger area.

Our system was built around a frontal imaging system with 5 line arrays (L-Acoustics Kara), but you can do as many as 64 channels of outputs on an L-ISA processor, including over head/height speakers. And it will allow up to 96 channels of inputs. We had 20 channels of input from Q-Lab and 76 channels from the console that were feed into the the L-ISA processor. You can adjust, in real time and preprogrammed, individual channels (or objects as they are called in the immersive audio parlance) which control both level as well as distance and height. So you can move an object left-right, as well as closer and further away, and over head if you have those speakers installed. If you can a computer with a touch screen monitor you can do all of this in real time with you fingers live during a show. Everything was synced to SMPTE time code and or triggered via MIDI.

They have a demo room their headquarters in Westlake that has, iirc, 24 surround speakers, plus 5 over heads and a subwoofer.

Quote:
Originally Posted by studer58 View Post
How well it might fold down to 2 channel stereo is anybody's guess...a dedicated stereo mix (and allocation of same to a dedicated layer on a BluRay or DVD) would seem the best approach.
In the L-ISA system, the downmix is handled by the L-ISA processor. It automatically outputs a LCR, Stereo, and mono downmix of the immersive audio information. This is particularly useful when loading a show into a smaller venue that doesn't have the height or width necessary to accommodate the full PA system, and to send a down mix to recording, or house program feeds, etc... With the L-ISA system, that's all built in and you can't change how it handles the downmix.

The entire immersive system relies on the incorporation of a phantom image between each speaker source. The more speakers you have, the more phantom images you will have. That's particularly useful when panning from left to right, or front to back as it give a much more granular and smoother movement from one side to the other or front to back. So with 5 speaker sources, you get 4 phantom images between each speaker, plus a sort of outer phantom image as you can pan outside of the main speakers. When we had to load in a smaller PA, the LCR was doable, but the L-R system was painful to mix on once you get used to hearing the show in a 5 wide array system.

Here's a pic of our main PA system, and the demo room at the L-Acoustics headquarters in Westlake, CA.
Attached Thumbnails
Mixing for Immersive Audio-82568552_10157964215074324_7949864433444978688_o-2.jpg   Mixing for Immersive Audio-img_20191219_152248312.jpg  
Old 4 weeks ago
  #5
Lives for gear
 
huub's Avatar
You can always mix in 5.1.4 by using a 5.1 bed and 4 objects.
Just saying.

Huub
Old 4 weeks ago
  #6
Lives for gear
 

Quote:
One of the main benefits to the immersive system the way L-Acoustics implements it is that it puts a larger percentage of the audience inside the immersive zone, as it spreads the sweet spot out over a larger area.

very well said! - and as such immersive live formats (pls note that the term 'immersive' gets used in different ways for use in live sound, at the cinema and also for listening music via home hifi - goals imo are vastly different!) have a clear advantage over most conventional surround and 3d systems: one doesn't need to sit in the sweet spot to experience, well, 'immersive' sound...

d&b has a somewhat similar system called soundscape and there are a couple of other manufacturers (such as astro audio) which allow you to use their processing regardless of dsp/amps/speakers.

i wish i'd get to use any of the systems more often; requirements and restrictions setup time and costs however are mostly prohibitive...

___


for listening to classical music at home, i find immersive/3d mostly pointless, at least the aspect of flying objects around the soundspace: imo this is highly distracting and does neither serve the music nor the listening experience.

___


however, the addition of elevated speakers of course affects the soundfield and i personally like it a lot outside the realms of home hifi: anyone been to pink floyd's exhibition at victoria & albert? brilliant!

i dont buy into dolby's ponzi scheme and use whater format the exhibition or venue calls for: whether that's ambisonic, ambeo etc.

on the recording side, i either use a soundfield or a double m/s for mains (and de-matrix to 5.1) and a double-ortf for the heights, often flown far above the main mic - i don't like the height channels stemming from (close to) coincident capsules.

on the mixing side, the 5.1 mix remains the basis and the .4 channels i keep static, the latter getting fed from the tracks of the double-ortf and occasionally with some 2in-4out 'quad' reverb mixed in.

by comparison to 5.1, the 5.1.4 allows for mixing efx differently: i'm using less efx than when mixing in 5.1 as the height speakers and the room seem to take care about that - downsides are that almost no one (at least who i know) has a capable surround playback system at home and even less an immersive setup; also, the room heavily affects results.

bumping up to 7.1.4 and beyond (or any width in ambisonic) yields the usual benefits but imo is only necessary for large rooms; same for more than 4 highs.

i'm mixing on a system consisting of 5 tannoy mains, 4 smaller genelec for heights and a kh sub; i can go up to 7.1.8 (with the sub consisting of 3 subwoofers - wouldn't wanna call it 7.3.8 though as the subs are not getting separate mixes; they are arrayed as a csb).

oh, and downmixing of any sort is evil!
Old 4 weeks ago
  #7
Gear Maniac
 

You can always mix in 5.1.4 by using a 5.1 bed and 4 objects.
Just saying.

Huub


You are right, for sure. Consumer domestic Dolby ATMOS is nominally 7.1.2 plus objects. So if you want to create 7.1.4 immersive music the easiest way is to have a 7.1.0 bed and four objects, one object in each uppermost corner.

The issue for me is that of folddown in playback systems which are not able to cope directly with 7.1.4 channels ending up everywhere they were mixed to go.

If I stick my Soundfield microphone up in the Royal Albert Hall, convert its output to 7.1.4 with front, side, rear and four channels of height, and encode it to Atmos, what happens when someone plays the recording in his/her listening room on an Atmos 5.1.4 or 5.1.2 system, or through an Amazon Echo Studio, or in 2ch stereo when streamed from Amazon, or on headphones if Atmos binauralises it?
Old 3 weeks ago
  #8
Lives for gear
 
huub's Avatar
I have to say that in my experiments, the atmos downmixes work quite well.
The safest bet would be of course to listen mainly to the stereo downmix and make sure that the full immersive mix is "impressive" enough.
That's the way I mix in 5.1 really.

Huub
Old 3 weeks ago
  #9
Lives for gear
 
Bruce Watson's Avatar
 

Quote:
Originally Posted by TonyF View Post
Anyone else experiencing immersive?
Morton Lyndberg of 2L in Norway has made quite a few surround sound encoded Blu-rays. Many of these come with SACD (both stereo and 5.1) discs as well as 7.1.4 variants. I think he's more partial to Auro 3D rather than Dolby Atmos, but I haven't been paying that much attention.

I've got a handful of the 2L releases. I have to say it's interesting. The immersive mixes are, well, immersive. Enveloping in ways very similar to what I'm used to in concert halls. Basically they try to override your room acoustics with the recorded concert hall acoustics. It works (for me anyway) better than that description makes it seem like it should.

The mix on the SACD of the same material, but mixed specifically for stereo, is a completely different animal. Considerably cleaner sound, easier to locate instruments. But nowhere near as enveloping.

To be honest, I like them both about equally. First class recordings, first class mixes. I think the immersive version blurs things too much. But the immersion is very seductive. I'd like to hear more of this; I suspect the more recordists work in this medium, the better they'll get at it. The "truth" is probably somewhere between the two mixes I'm hearing in my listening room. But it's early days yet. We'll see.

Some pictures of the nine mic array Lyndberg uses on some of his work run down the right hand side of this page. If anyone wants to see what unlimited funds can do with Grace Design bars and parts, it's interesting. I personally wonder about the distances involved -- In particular, I wonder if this (particularly the front/back distances) cause some of the blurring I'm hearing in the recordings.
Old 3 weeks ago
  #10
Quote:
Originally Posted by Bruce Watson View Post
'd like to hear more of this; I suspect the more recordists work in this medium, the better they'll get at it. The "truth" is probably somewhere between the two mixes I'm hearing in my listening room. But it's early days yet. We'll see.

... I personally wonder about the distances involved -- In particular, I wonder if this (particularly the front/back distances) cause some of the blurring I'm hearing in the recordings.
I share your feelings, Bruce.

Immersive recordings using this approach are essentially just a-bunch-of-omnis-in-the-middle-of-a-room. When you listen to the stereo, you hear that mush and smearing... and it sounds like that even in a dedicated immersive room. If you don't have a truly stellar stereo recording with a beautiful 3D perspective, adding more omnis certainly isn't going to make it better, and having the players in a circle surrounding the array is going to make matters even worse. My advice to ensembles is to avoid smooth-talking salesmen who want to use them as play-purses and guinea pigs.

Personally, I'm continuing to add dimensions and resolution, but the mix is still everything.
Old 3 weeks ago
  #11
Lives for gear
 
Bruce Watson's Avatar
 

Quote:
Originally Posted by NorseHorse View Post
Immersive recordings using this approach are essentially just a-bunch-of-omnis-in-the-middle-of-a-room. When you listen to the stereo, you hear that mush and smearing...
So... anyone have an example of immersive recording that works well in playback? Something on blu-ray, 5.1, 7.1, atmos, whatever? IOW, something that a reasonable disc player and AVR can decode.

I'd like to hear it because I think the idea of immersive is a good one. I'm just not sure what is actually possible, and what is not. Said another way, just how close can we get to the "you are there at the concert hall" level of immersive?
Old 3 weeks ago
  #12
Gear Maniac
 
MandoBastardo's Avatar
Just ensure every release includes a complimentary Smyth Realizer for those listeners wanting a 24 speaker array inside their head.
Old 3 weeks ago
  #13
Gear Maniac
 
elpillo's Avatar
 

Quote:
Originally Posted by Bruce Watson View Post
So... anyone have an example of immersive recording that works well in playback?
Hey, Bruce. It all depends on what you mean by this. Looks like the L2 people have developed a system that clearly works quite well for them and for other people, including some colleagues (look at 30+ Grammy nominations Morten has)

On the other hand, according to the Prof. Hyunkook Lee's research, at Huddersfield University, the difference between having height microphones elevated from the lower array, and coincident ones, isn't much in terms "immersive feeling". In his research people actually preferred the sound of coincident height better. Schoeps implemented the ORTF-3D based on his research (similar to his ESMA-3D array)

In this conference he talks about it:
https://www.youtube.com/watch?v=5SsQdfKAhVA

Still a lot to play and experiment with. I've been doing my own recordings for experimentation only and with the gear I have and is a fun process. I don't own 11 of the same microphones but I do it with what I have.
Old 3 weeks ago
  #14
Lives for gear
 

Quote:
Originally Posted by MandoBastardo View Post
Just ensure every release includes a complimentary Smyth Realizer for those listeners wanting a 24 speaker array inside their head.
certainly an interesting product, thx - the thing is though that one does neither need to record nor mix (and deliver) music in multichannel formats to get impressive results: check out klangfabrik's headphone mix processor which needs but a stereo input...

https://www.klang.com/en/products/klang_fabrik
Old 3 weeks ago
  #15
The biggest issue I have with the virtual speaker sets is how the down mixes from 7.1.4 and 7.1.6 to 5.1.2 and 5.1.4 happen. When going from 7 to 5, the side channels get spread evenly between the front and rear, causing some strange summing artifacts in the front L and R. To be honest, I think that the side channels are fine for theatrical, but become more of a burden for music.
I'm using a 5.1 bed and 4 objects for the height channels in my music mixes thus far, so I can, again, avoid the strange downmixes of the top front and wide into the main LCR when going from .4 to .2.
In the 5.1 to 7.1 conversion, Ls and Rs get spread to side and rear evenly and don't sound nearly as wonky as going the other way around.
Also, has anybody done anything at 96 in the newest render?
As always, YMMV.
All the best,
-mark
Old 3 weeks ago
  #16
Gear Maniac
 
MandoBastardo's Avatar
Quote:
Originally Posted by deedeeyeah View Post
certainly an interesting product, thx - the thing is though that one does neither need to record nor mix (and deliver) music in multichannel formats to get impressive results: check out klangfabrik's headphone mix processor which needs but a stereo input...

https://www.klang.com/en/products/klang_fabrik
Sure...it could sound great.

But... has a huge flaw. It's a communal processor which just doesn't work in our new socially distant era. Listening to headphones with 3 other people at the same time, in the same small stuffy room + the cats is an agoraphobic faux paw.

Although... if my inner audiophile was to spend $6k on its own selfish needs and disable the other outputs, could sit next to my Wavedream and EC Studio without too much disdain. ;-)
Old 3 weeks ago
  #17
Lives for gear
 

Quote:
Originally Posted by MandoBastardo View Post
Sure...it could sound great.

But... has a huge flaw. It's a communal processor which just doesn't work in our new socially distant era. Listening to headphones with 3 other people at the same time, in the same small stuffy room + the cats is an agoraphobic faux paw.

Although... if my inner audiophile was to spend $6k on its own selfish needs and disable the other outputs, could sit next to my Wavedream and EC Studio without too much disdain. ;-)
nah, playing together is picking up again, at least in those places where people in charge were smart enough to implent some restrictions early enough:

i'll be mixing live next weekend, albeit for a relatively small crowd (currently limited to 300 people, with limitations expected to get raised to 1000 within the next two months).

klang DOES sound great! download the app and give it a try...

...and on the input side: chech out austrian audio's ambicreator: much fun!
Old 3 weeks ago
  #18
Lives for gear
 
Bruce Watson's Avatar
 

Quote:
Originally Posted by mpdonahue View Post
The biggest issue I have with the virtual speaker sets is how the down mixes from 7.1.4 and 7.1.6 to 5.1.2 and 5.1.4 happen. When going from 7 to 5, the side channels get spread evenly between the front and rear, causing some strange summing artifacts in the front L and R. To be honest, I think that the side channels are fine for theatrical, but become more of a burden for music.
I hear ya. I can see how that might work, and how it might not. It's a shame really since I personally don't get much out of the difference between 5.1 and 7.1 theater mixes. But I thought those rear channels (as opposed to the side channels) would become useful for music (eventually) -- supplying the reverb from the back of the hall maybe. IDK; just me being hopeful I guess.

Here's a question though: is there no flag that can be set to tell the decoder to just toss excess data? That is, when you're sending a 7.1 mix to a 5.1 system, just toss out the rear surround channels rather than attempt to "redistribute" that data across the 5.1 speakers "randomly"? Seems like such a flag would be useful so I doubt there is one.

Quote:
Originally Posted by mpdonahue View Post
I'm using a 5.1 bed and 4 objects for the height channels in my music mixes thus far, so I can, again, avoid the strange downmixes of the top front and wide into the main LCR when going from .4 to .2.

In the 5.1 to 7.1 conversion, Ls and Rs get spread to side and rear evenly and don't sound nearly as wonky as going the other way around.
Well, have you got any blu-rays (EDIT: I mean: can you point me to titles and retailers so I can buy them) you can recommend to demonstrate your 5.1 technique? or your 7.1? Or 7.1.4? I'd like to hear whatever different techniques I can find. I'm bound to like some more than others, which will of course point me in a direction to explore maybe.
Old 3 weeks ago
  #19
From my experiments, Atmos doesn't just shuffle stuff randomly forward. It still very much adheres to existing fold down techniques of previous channel based formats. It is one of the reasons why I don't quite understand the mention of 'phantom sides'. 7.1 is 5.1 plus rear speakers. 7.1.4 is 5.1.4 plus rear speakers. The side channels always exist! The only difference is where the rear channels are folded into them or not... but you should never be hearing your side channels coming through the fronts.
Old 2 weeks ago
  #20
can you tell me how this sounds on your system? https://soundcloud.com/corey-spencer-673305839/esc
๐Ÿ“ Reply
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
๐Ÿ–จ๏ธ Show Printable Version
โœ‰๏ธ Email this Page
๐Ÿ” Search thread
โ™พ๏ธ Similar Threads
๐ŸŽ™๏ธ View mentioned gear