With the increasing use of Dolby Atmos distribution formats, more and more music creators, studio operators, labels and engineers are asking themselves whether the format will become a new music standard and what demands this will place on music makers in the future. Is it worthwhile to deal with Atmos? Will stereo soon be obsolete? What equipment do I need for 3D audio? Do I have to change my software and retrain?
Dolby Atmos is an object-based audio format that, unlike channel-based formats such as stereo or 5.1, can reproduce a three-dimensional world independent of the speaker configuration. As a result, studio and consumer no longer need to use the same speaker setup.
Object-based mixing requires some rethinking and a well-sorted workflow that breaks with some traditional ways of working that music creators have become accustomed to over the past few decades. But once it is comprehended and set up, the 3D world should be seen more as a creative addition to what we already know – so Atmos and mono/stereo don’t contradict each other.
How does mixing work for Dolby Atmos?
First things first: for object-based mixing in Dolby Atmos, sound engineers have to integrate the handling of significantly more channels into their workflow. We usually worked with a two-channel master bus, now we have 128 channels! But don’t worry: in practice, of course, these channels codon’t have to be routed manually every time, but are set up once and saved for future projects.
The Dolby Atmos format consists of two different types of signals: “Beds” and “Objects”.
“Beds” are conventional, channel-based surround signals that lack metadata (e.g., soundscapes). They are placed on a fixed 7.1.2 surround bus and don’t move in space. If you want to present signals in classic mono, stereo or channel-based surround, you can simply use this bus.
“Objects” are mono signals that can be given spatial coordinates using a 3D panning plug-in and can thus be placed freely. Of the total 128 channels, 10 are used for the “Bed” bus, so there are up to 118 channels available for independently placeable objects.
For mixing in Atmos, two software components are necessary:
- DAW with object support. In the professional context, these are currently mainly Avid ProTools Ultimate and Steinberg Nuendo, but mixing Atmos is also already feasible in Logic Pro and Ableton Live. It is to be expected that more DAWs will integrate native support for object-based mixing in the future. The assignment of object coordinates is handled by a corresponding 3D panner.
- Dolby Atmos Renderer. The renderer has the task of turning the delivered individual signals into an audible event using the supplied metadata. Listening to your Atmos mixes is made possible with the renderer, and the various target formats are generated by the renderer as well. Nuendo has its own Atmos renderer, alternatively the Dolby Atmos Production Suite renderer is used. The renderer can either be running on the same or on a dedicated rendering computer. Hardware devices (Dolby RMU) can also be leased for the larger Atmos mixing stages.
In addition, of course, a suitable monitoring setup is mandatory. A professional mixing environment for Dolby Atmos starts in 7.1.4 (7 speakers at ear level + 1 LFE + 4 ceiling speakers), so 12 speakers in total. However, the Atmos renderer also allows rendering for smaller monitoring systems and even for headphones, so it is also possible to work from the home studio or create a rough mix on the road. Anyone who wants to offer Atmos mixes professionally should of course adhere to the recommended, technical standards for the studio setup. Dolby engineers can be called in for consultation, training and calibration, but a certification process is not necessary to deliver music mixes in Atmos.
One tends to deal with (many) more channels in 3D audio production than in classic stereo. So in addition to a well-structured workflow, a powerful studio computer is a must. Then there is the Atmos renderer itself, which of course also devours quite a bit of CPU resources.
The target formats
Since Atmos technology was primarily developed with cinema and film in mind, it works in 48, 96 or 192 kHz rather than 44.1 kHz which is commonly used in the CD market. Atmos masters are output as DAMF (Dolby Atmos Master File). In this format, audio data, object coordinates and project information are split into three separate files.
For the distribution of Atmos mixes, a simplified format is used that requires only one file: ADM BWF. Here, all the above information and signals are combined in one single WAV file.
The Atmos renderer also outputs channel-based WAV files in the desired format. Pretty much everything from a binaural headphone mix to theater setups are possible. So, while multiple mixing sessions were previously required for multiple sound formats (e.g., for DVD and cinema release), Atmos allows multiple formats to be played out simultaneously from the same mix and even re-rendered later from the Atmos master if needed. So up/down mix sessions for the different distribution formats are no longer necessary.
If stems are required, there is the option in the Atmos Renderer to add objects to groups and output them in multiple files during re-rendering.
Mastering Dolby Atmos
Technically, you are dealing with 128 individual channels;so the requirements in the mastering process for Dolby are rather complex. Loudness control is built into the Dolby Atmos renderer. It can be measured in stereo and Atmos and gives the possibility to comply with loudness specifications during export.
However, if you want to implement, for example, sum compression in an Atmos mix, it can quickly become complicated because there is no sum. Sidechain routings for such cases require diligence, but allow for a lot of flexibility. It’s all possible, just a little different than usual.
Obviously, channel-based re-renders for the various distribution formats can still be mastered in the classic way.
Remastering for Atmos
The Beatles in Dolby Atmos? How could that possibly be? Well, the magic word is “remastering.” If possible, the original tapes are brought out of storage and a completely new mix is created based on the original individual signals. If the signals are of acceptable quality, this is basically a lossless process. However, if the tapes are damaged or the signals are only available in summed form, remastering can be a very compromised affair and is more like a technical restoration than an artistic activity. So everything depends on the quality of the original backups. In this respect, remastering for 3D audio is no different from remastering any other type of format.
Is good old stereo now obsolete?
Hardly. It certainly looks like object-based mixing is here to stay in the music market, but that doesn’t mean that mono, stereo and the like have had their days. After all, the channel-based acquaintances are of course still the basis both for recordings and the monitoring situation. So it will continue to depend on the respective project, the target market and, of course, the budget whether object-oriented work is the thing to do or not.
If you listen to an Atmos mix on headphones, you will usually hear a binauralized mix. This means that a multiple of data is transferred to the end device, only to turn it into two channels again at the end. The same headphone experience as with Dolby Atmos can therefore also be produced with simpler means (and without special demands on the production environment) and distributed via conventional stereo distribution channels. However, the binaural mix is of course limited to headphones and only works to a limited extent in the three-dimensional speaker arrangement. 3D for headphones is not a new topic and can be implemented with ease using binaural stereo tools such as the Binaural IR Pack for the HOFA IQ-Reverb V2. I fully expect that binaural stereo mixing will continue to grow in popularity and importance together with the growth of Dolby Atmos.
I would advise aspiring audio engineers to at least take a fundamental look at the subject of 3D audio. Whether it’s Atmos, Ambisonic or binaural is secondary for now. Simply download 1 or 2 demos of 3D-capable plug-ins, start playfully and gather your first experiences. You’ll also find a growing amount of 3D audio content in the HOFA-College courses. It would be a shame to have to turn down a potential job just because of being afraid of new formats. But above all, new things should be fun! Have the courage to experiment! 3D is now and you can join in! And you’ve already made a good start by reading this article. Have fun!
*Update (as of October 20th): Native Dolby Atmos support in Logic Pro X.
For Logic users interested in immersive audio there is very pleasant news: With version Logic Pro 10.7 (free update), Apple has integrated native support for Dolby Atmos into its popular Mac-only DAW! What does this mean in concrete terms? Object-based mixing in 3D becomes a breeze! There is no need for complex routing efforts, manual clocking connections or additional rendering hard- or software, which not only saves money, but above all saves nerves and should make it much easier for many sound creators to enter the “object world”. Of course, it is also possible to export fully distributable ADM BWF files to release on Amazon, Apple Music, etc.
The workflow is actually quite simple: When creating the project, select the Atmos 3D option and a corresponding Dolby Atmos plugin automatically appears on the master channel. This is not coincidentally very similar to the Dolby Atmos Renderer, but is still somewhat simpler in structure. Here you only have to select the appropriate monitoring format (i.e. whether you want to listen in stereo, 5.1, 7.1.4 or binaural with headphones) and you’re ready to go! On the right side a list of the bed channels and objects can be displayed, and the perceived spatial distance can be set in three stages and the proven room view shows the position of active objects. Both the list and the room view are practical but optional, so you can just choose your monitoring format and never have a look at your renderer again. It’s that simple!
A channel can be “objectified” simply by right-clicking the pan pot, selecting the “3D Object Panner” and placing the signal as desired. For each channel, a choice can be made between a conventional pan control (or balance control in the case of stereo signals), a binaural panner and the new 3D Object Panner. This means that classic stereo and surround signals can easily coexist with binauralized elements and 3D objects in one session. By the way, the bed bus can be fed using the regular surround panner, if desired. The bus “Surround” is fed as the routing destination and the panner automatically switches to the correct surround mode.
In my opinion, the implementation in Logic Pro is very accessible and pleasantly solved, and in my first Atmos test runs in Logic, no technical problems were encountered. I especially like the mostly self-explanatory operation regarding routing, panning and so on; that is still not self-evident. Even though Apple took a little time with the native integration (Steinberg Nuendo, for example, was there much earlier with its own renderer), it was worth the wait. If you use Logic, you have no excuse not to try 3D audio, because it has never been so easy (and so cheap).
What is your experience with Dolby Atmos and object-based audio formats so far? Let us know in the comments!