Atmos seems to be the only thing engineers are talking about these days. It’s something I’ve talked about for years but no one else seemed to care. I’m glad it’s finally becoming mainstream.
I originally got into Atmos around 5 years ago. My work as a scoring mixer means that I’ve worked in surround (usually 5.1 or 7.1) ever since I started out. Atmos started to become popular around 6 - 7 years ago in the film industry and theatres slowly implemented it. At that point, no DAW properly supported it and scoring mixers treated it as 7.1 plus height channels in terms of bussing.
At the same time there were also other 3D surround formats like Auro-3D, DTS:X, and the NHK 22.2. The research in this area has been done for many years and there are a number of AES papers discussing these formats as well as recording techniques for them. This includes the papers I’ve written about recording for these formats and active acoustics which you can find on my Products page. Because Atmos wasn’t very popular at the time, I originally set up an Auro-3D system in my studio but unfortunately had no way of playing any released Auro-3D so it was mostly for my own experimentation. The Auro-3D system itself was very closed off which is largely why it never caught on. Atmos seemed a little more approachable and Steinberg released support for it in Nuendo at the same time with Nuendo 7. At that point I tore down my Auro-3D system and replaced it with 7.1.2 for Atmos.
After I invested into expanding my system into Atmos, it was essentially dead and only used for a few high budget films. Atmos music was nonexistent so my system sat unused for close to 5 years. I wanted to work in Atmos as well as binaural specific mixes but there was no market for it and no distribution services supported it. The closest option was to release a separate headphone-only version of an album mixed binaurally but this brings up an important issue that Atmos solves. One of the beauties of Atmos is that, when implemented correctly, it’ll be rendered perfectly on any system. That includes in a theater, in 7.1.4, in stereo, and on headphones. No need for a second release of an album that’s only intended to be listened to on headphones. A single Atmos file does it all.
With the Apple support of Atmos, other distributors also supporting it, and the push from Dolby, Atmos is now coming to the attention of artists, engineers, and consumers. It is now possible for anyone to easily create and release Atmos content. This can either be done with Pro Tools Ultimate and the Dolby Atmos Production Suite or with Nuendo and the built in renderer. Both of these options be can done on a single machine without any extra hardware. More and more DAWs support Atmos.
While everything surrounding Atmos is great and I think there’s potential for it to be the new norm, there are still some major issues which many people are overlooking. The first issue is that Dolby is not providing content creators (studios) a way to play back released Atmos material unless you have the Atmos files of a release which are almost impossible to get (I don’t know of anywhere that you can buy Atmos files). Yes, you can listen to it on headphones but anyone who has a 7.1.4 or 9.1.6 speaker setup is going to want to be able to listen to music or watch films on their system.
With 5.1, there are a number of platforms like the Windows Netflix app which will play 5.1 out of your system with no issue. For other platforms like Prime or DIsney, it’s possible to get a Dolby rack unit which takes the digital out from a device like an Apple TV and gives you digital outs that can be plugged into your system. The problem is that this doesn’t really exist for Atmos.
The solution that most studios (including me) use is an Atmos receiver with preamp outs. This typically costs around $2000 since only higher end receivers have preamp outs. The obvious problem with this is that there’s going to be the DA on the receiver then going to an AD to get into your system so there’s redundant (and not great quality) conversion. The other thing to consider is that studios don’t typically want all of the other features that a receiver has (DSP like Dirac, bass management, various I/O, amps, etc) so you’re paying a premium for something you’re not going to use.
So how do you feed your system digitally to avoid the redundant conversion of using preamp outs? Unfortunately the cheapest solutions start at $5000. This includes the unit from Arvus (AES outs) or the JBL Synthesis receiver (Dante out). Beyond that, it’s only receivers costing more than $10,000. A studio should not have to spend $5000 just to properly play material from Apple Music on their speakers or watch some Netflix. Dolby should either offer their own affordable decoder box with AES (and/or MADI) or offer a way for Atmos to be decoded natively in Windows so that something like Apple Music can play directly out on any interface. While this is a major oversight on Dolby’s part, this isn’t the biggest problem for Atmos.
The great thing about Atmos is it’s ability to play back on any device. Music in 5.1 is never going to catch on because no one has a 5.1 system. The majority of people listen to music on headphones and Atmos is able to beautifully translate 3D audio into headphones. Once everything is properly implemented, the end user doesn’t have to even know what Atmos is. The problem is that it’s not as simple as playing the binaural version of a mix. This simply doesn’t work.
Binaural audio is essentially a way of tricking our ears into thinking that we’re hearing spatial cues around us when it’s only coming from 2 speakers (1 on each ear). Because we all have different ears and hearing systems, the processing that simulates this can’t be applied to everyone. Doing so is like wearing someone else’s glasses. Most of the time you’re not going to have a good time and might run into a wall. Personally I haven’t heard any Atmos mix on headphones that didn’t sound awful and practically made me nauseous. It’s because what I’m hearing wasn’t made for my ears and so it’s like I’m hearing the world through someone else’s ears. My brain can’t understand what’s going on. For anyone interested in digging more into this, it’s all about the head related transfer functions (HRTFs). Research in this field has been done for decades (including by Dolby).
There is a simple solution for this. Most of the data needed to create a binaural render for a specific individual is based around their ear and head shape. With a few measurements or pictures, those parameters can be set and stored to make any binaural mix play properly for that individual. If there isn’t a system implemented that factors this in to binaural Atmos playback, I think that Atmos will die out because the content won’t sound very good for most people.
How do most people consume music? On their phones. Isn’t it convenient that our phones have cameras? All it takes is moving your phone around your head for it to scan you. There are already systems that do this but unfortunately they aren’t industry wide systems. There needs to be an industry accepted way of doing this. Our phones already scan our fingerprints and use facial recognition. It’s just one simple additional step to add when you get a new phone.
The other factor that needs to be considered is the type of headphones or earbuds being used as that also affects how a binaural render is perceived by a listener. With wireless options being increasingly used, it’s easy for our phones to detect what we’re listening on and adjust accordingly. Otherwise there can be a setting where you specify what you’re using.
Once the integration of HRTFs takes place, I think Atmos will become the norm for the recording industry and consumers don’t even have to know what it is but when they listen to a stereo project, they’ll feel that something is lacking.
You might be wondering what my role in Atmos is as a mastering engineer and that’s still something being defined. On one hand I can take someone else’s Atmos mix and make any necessary tweaks to get it ready for release as is the case for my usually mastering work. On the other hand, for Atmos, I can take on more of a mixing approach. This would be either taking stereo files and exploding it out into Atmos or taking a mix done binaurally and making sure that it works well on a full Atmos speaker system.
Because I’m taking on this finalizing approach in the Atmos environment, my new studio being built in the next little while will have one of the best Atmos systems possible to make it truly an Atmos mastering room. This includes custom speakers with complex DSP built in (similar to the Kii Three’s or D&D 8C) for phase and timing correction across all speakers. Fully digital monitoring means that the only stage of conversion happens in the speakers after the DSP processing with the DACs going directly to the amps which go directly to the drivers. In addition, all speakers will far outperform the Dolby specs. The system will initially be 7.1.4 and then will be expanded to 9.1.6 depending on how successful this format is. As for the system in my current studio, it is 7.1.4.
The beauty of Atmos is that anyone can do it with a pair of headphones. You don’t need a 7.1.4 speaker system to work with it. All you need are the software tools. Having said that, it’s important to check on a speaker system before release which is where I fit in and am able to provide that last step. If anyone has questions about this workflow, get in touch. I encourage everyone to start working in this format as it can lead to some great material even if your focus is on releasing a stereo mix.
I predict that in 3 - 5 years Atmos will either be gone and forgotten or the HRTF systems will be implemented and it’ll be everywhere. I don’t see any other option that works sustainably. Let’s see if I’m wrong!