Brief Summary

In this article, Roger Guerin breaks down what you need to know if you are moving to Dolby Atmos, from hardware through to adjusting your mixing template, through to deliverables.

Going Deeper

You might be aware by now that Apple Music is improving its offer by 10% for Spatial Audio music content. This is a welcome shot in immersive’s credibility arm. At the time of writing, immersive content has increase by 5000% since its debut in 2021. And this number will grow exponentially in the months to come. Adding to this great news, Avid has finally integrated the Dolby ATMOS Renderer in Pro Tools. The planets are all aligned for you to climb aboard this celestial band wagon, and this article will help you with just that.

The following is based on a conference presentation I gave at the last AES Convention in NYC. It will give you hints and ideas on how to upgrade your set-up and your workflow. Pro tools based, its intent is targeted for the musician as well as the re-recording mixer.

A Bit Of Technological Background

The promise of Dolby ATMOS is that it can be consumed on any type of device. Better yet you only have to mix once and occasionally double-check to make sure your intention translates well! At the beginning, it might be overwhelming with all these new numbers and expressions, but like on your very first console, the buttons and layout are extremely Cartesian.

Start With The Numbers

The usual suspects are: 5.1.2, 7.1.4, 9.1.6, with variations like 7.1.14 or 9.1.12 possible. The first number is the number of loudspeakers you have at ear level (Left, Right, Center, Surround Left, Surround Right, etc.). The .1 refers to the LFE and represents 10% of the human hearing range (hence the .1). The last number identifies the top loudspeakers, introducing height.

Two Flavours Of Dolby ATMOS

The one available in most DAWs is the Home Entertainment version (HE) with a maximum of 22 speaker channels (ie. in our example 9.1.6, 9+1+6 makes use of 16 speaker channels). The other version involves a bit more hardware (including an external specific PC configuration). Aimed and developed for the cinema, it has a maximum of 64 speaker channels. Both have the same limitation of 128 audio channels. Of course, the main question is: how much is it going to cost? The answer is “It depends”.

Three Formats To Consider

If you are already in 5.1 surround sound, chances are you already have an eight output interface. Adding two overhead loudspeakers gives you a 5.1.2 layout, a good start to dip your toes in the immersive format.
But if you are coming from a stereo setup, you might already have an eight output interface. Loudspeaker wise, the thing to watch out for is the center loudspeaker which has to be the same model as your Left and Right speakers. The surround and the top speakers must match between themselves. Additionally, it would be nice that they were all from the same manufacturer as your front loudspeakers. It’s a question of loudspeaker design philosophy, which makes for fewer surprises when you pan things around.

The next step up is 7.1.4 or 9.1.6 and if you must change your interface here are some things you might consider. A minimum 16 outputs is the norm, which is good up to 9.1.6. Connectivity has to be considered, making sure your current and future computers will be able to handle the interconnections. To future proof yourself, you might want to consider Dante. Built-in loudspeaker equalization (SPQ) is nice to have. Instant monitor recall when you are mixing for streaming, CDs, films, etc. You’ll thank me later!
An excellent article by our own Paul Maunder can give you a heads up. Don’t be intimidated by the title, the selection applies to all.
Loudspeaker wise the same logic applies: you will have to add matching ear level and ceiling-mounted loudspeakers.

As mentioned earlier, Dolby ATMOS is integrated in most DAWs and although many users will only ever use the integrated Dolby ATMOS Renderer. Here are some of the reasons why you still may require the external Dolby ATMOS Renderer.

Only available in the external Dolby ATMOS Renderer

Not available in Dolby ATMOS within Pro Tools

Although some features aren’t available compared to using the external renderer, there are many things that are made easier using the Integrated renderer within Pro Tools, and many more to come, I’m sure. To name a few: first you don’t have to configure both the external renderer and the I/O within Pro Tools. Big time and frustration saver!
And since everything is within Pro Tools, there’s no need for Time Code, a minor but significant irritant.


Making Use Of Your Existing Template

Once you are settled on your mixing format, and have installed, and calibrated the extra hardware, it’s time to upgrade your template. If you’ve spent years fine-tuning it to a point of near perfection and don’t want to compromise it, make a copy. Then you’re safe!


Simply put, you subdivide your tracks into “families” (stems), each feeding a subgroup with its processing then outputting to a stereo track. In our example, all the drums’ tracks are feeding a subgroup (sKit), split to the main mix (sMix - pMix) then to the audio track (pKit). The same for the others giving you all your stems and the final mix.
When working for screen, you are already mixing in stems, and have to deliver in 5.1, stereo, and maybe some monos.

The above might look complicated, just imagine the rest of it, the ambiances, the SFX, the sound design, the music, etc. But it isn’t really. You just need to be organized. Let’s take your session and bump it up to the next level to accommodate immersive mixing.
As you probably know, the Dolby ATMOS Renderer is split between beds and objects, not wanting to spend tool long on this, let’s say that beds are like your stems, and that objects are the components you will be flying around the soundfield.
In keeping with the subgroups = beds metaphor, the subgroups no longer feed the print tracks but go out to its assigned bed. Furthermore, we don’t even have to spend time in configuring the different deliverables, they will be done afterwards in one fell swoop.

You will notice we kept the integrity of the template, but now our subgroups only feed their corresponding bed. I would argue that it’s less of a stress on the computer and accompanying hard drives. The shocker, as you might have noticed, is the absence of a Master Fader.
You might think that this changes everything: that the plugin(s) on the Master Fader glue all your mix together. Perhaps. But don’t forget you’re not in stereo anymore: different medium, different listening experience. Furthermore, you have a lot more loudspeakers to express yourself, and the processing on your subgroups is always there to keep things in check.

Not being one for presets, it always surprises me when the default status of a plugin or a workflow is so far from usable (e.g. a compressor with a threshold at -24 dB).
Unfortunately, all the Dolby ATMOS Renderers I have seen default to one bed, 7.1.2 wide and all the objects enabled. Do yourself a favor and avoid that default setting.

Here are some configurations starting points. Of course your mileage will vary.

To be noted: the first ten (10) “audio channels” have to be a bed, but it doesn’t have to be one whole bed, it can be several beds that add up to ten (10). As explained later, my bed count is much more than that.

Some Observations

The Dialogue Bed is only 5.0 wide and no Objects?

Yes, because I don’t intend sending any LFE from Dialogue, and I don’t plan to fly Dialogue around, but there’s always the rare exception.


The Ambience Bed has no provision for the LFE?

Usually any LFE content for a scene will be included either in the SFX or the Sound Design tracks.

The SFX and the Music stems are not assigned to a 7.1.2 Bed?

I prefer to control what goes where. The typical Upmixer can upmix to and beyond 7.1.2 but while what goes up in the .2 is consistent, it makes for a very static mix.

You will also notice that some Objects are attributed to some Stems. They are not etched in stone and your use and inspiration will greatly influence these proposed numbers. They are intended to have some elements “fly”. Elements like: the overhead hawk in the BGs, the infamous helicopter pass-by or a synth going in and out of focus.

The six Objects attributed to the Reverb Top will be explained later.

Before I forget, only a Bed can be assigned to the LFE, even though it’s against any logical explanation, Objects cannot.

Observations

The same observations apply:

The Vox Bed is only 5.0 wide with some Objects?

Yes, because I don’t intend sending any LFE from the Voice, and I don’t plan to fly the lead Voice around. Back up Vx and their processing, on the other hand, are great candidates for Objects.


The Guitar Bed has no provision for the LFE?

The Guitar range is too high to be able to tickle the LFE.


The Keys, Drums and the Percs are not assigned to a 7.1.2 Bed?

The same logic applies: I prefer to control what goes where. The typical Upmixer can upmix to and beyond 7.1.2 but while what goes up in the .2 is consistent it makes for a very static mix.


You will also notice that some Objects are attributed to most Stems. They are not etched in stone and your use and inspiration will greatly influence these proposed numbers. They are intended to have some elements “fly”. The overhead Back up singers, the dramatic back-to-front strings (or brass) or the doubling of a synth line an octave higher going in and out of focus.


You might be wondering what’s with the six Objects attributed to the Reverb Top?

With the advent of immersive mixing, the reverb manufacturers are able to fan out additional outputs via their plugin’s insert. The trick is to assign these added outputs to objects and have them available to the upper tier. As seen in the next screenshot, three stereo auxiliaries were created with their inputs coming from three different inserts returns from the same plugin (A). In this case Cinematic Rooms Pro by LiquidSonics. They are then assigned to three “stereo” pairs (B). (Note: stereo objects do not exist, you have to gang two adjacent mono objects to be able to use them as a “stereo” source. Some DAWs even do that for you).

To create some excitement, I like to use a plugin that creates some kind of randomness (C), in this case, the Waves PS22 Split (D), but I am always on the lookout for something else. I will keep you posted. Then, to add to my signature, I play with the object’s size (E), just to have more excitement and unpredictability, keeping the listener engaged and… dare I say it… immersed.


Preparing To Output

Now comes the time to output all of this and as mentioned earlier it’s really fast. As with everything that’s efficient, some preparation is involved.

As seen in this combined screenshot, you have to specify the nature of the Bed. By default, you will have a long list of options. Get rid of the excess suggestions and customise to your taste. In our example bDial is part of the Dialogue Group (A). 
If you are considering having a Binaural version or you want your audience to have a Binaural experience, you must specify the characteristic of the distance (B), from Near, Mid or Far depending on the expected position of the sound. Or Off to deactivate that feature.


Difference Between Stereo And Binaural

Binaural is quite a buzz word these days and it’s a far cry of what it meant originally. Initially attributed to recordings typically made with two microphones inserted in a dummy head’s ears, the recordings were only intended to be played back with headphones. Admittedly, the impression of 3D was convincing, I imagine that’s why the word stuck. But today the math is way more sophisticated. 
Simply put, the use of an average Head Related Transfer Function (HRTF) algorithm helps give the impression of the immersive experience over headphones. But doing so, it introduces some degree of reverb and equalization.
Noticeable on solo instruments or in very quiet scenes.
Use it wisely.


Back To Output, Or In The Dolby World, Re-Render

To output the Dolby Atmos Master (ADM) needed by Apple Music or the Online people for video content, we proceed like we always did with a beautiful twist.

When you are ready to Bounce your Mix you will be greeted by this familiar menu and when you choose WAV (Dolby Atmos) (C) the Re-Render submenu opens (D). Then it’s up to you to Configure (E) what you or your clients need. Of course, they will need the complete Mix (G), and maybe some Stems (H), in the formats of your choice.
In our example (H), after hitting the custom button, you must choose what is to be included in the Dialogue stem in 7.1 with Beds & Objects.

It might seem like a lot of work, but here is the payback - In our example with just one click you get to output:


• The 7.1 Mix

• The 7.1 Music stem

• The 5.1 Mix

• The 5.1 SFX stem

• The stereo Mix

• The stereo Ambs & SFX combined

• The Binaural Mix

• The Binaural Music stem

Not bad for just one click!

So there you have it, what used to be a 5.1 subgroup is now a 7.1 bed. What used to be panned from left to right to the back is now so much more! I encourage you to experiment with this format because it’s here to stay. Besides, it’s very inspiring, you will be excited again in finding new tricks that you never thought possible in stereo or in 5.1 for that matter.


Right now, we are still in the infancy of immersive mixing, and if you give it half a chance you will be rewarded by self-gratification.

Spread the good word and share your thoughts below.