Expectations, expectations, expectations. What devilishly tricky little things they can be — and nowhere is that more evident than in a custom home theater! Big home theater projects are a tough sell due to the high price tag, even if the client is inherently interested and enthusiastic. As a result, we tend to over-promise in order to get the client to sign on the dotted line, then under-deliver months or even years later. It’s not like we’re trying to deceive the client — maybe we’re even fooling ourselves a little bit about what we can achieve given the constraints of budget, space, etc. Managing expectations is one of the hardest parts of any project, and it’s where I see so many projects go off the rails. And one of the big stumbling points today can be in the content source material.
Not All Source Material Is Created Equal
Numero uno, for your sanity and mine, brief your clients on the fact of life that not all sound is designed and mixed with equal care and skill. As they are flipping around between providers, your clients will be faced with the entire spectrum of sonic offerings — unfiltered and unprotected. They will not all sound the same, and they will not all sound good. The number of clients who have the knowledge or motivation to seek out well-produced material is vanishingly small, so clients inevitably end up hearing inferior-sounding material and wondering where all their money went. If you brief clients about this on the front end, you may put them off a home theater altogether, but if you don’t tell them, they’ll blame you after the fact when their expectations aren’t met.
Surviving Streaming
The next hurdle is the delivery process. Most streaming services do not prioritize high-fidelity sound as part of their business model. In fact, it could be argued that an expensive sound system is not justified at all — perhaps even detrimental, as it reveals the flaws in the content more obviously.
In my experience, the best way to address this issue is to tell the client during the initial discovery and planning phase that you are going to provide them with several different options for procuring and playing content. Some of these options will be focused on convenience, and some will be focused on quality. It’s like the difference between making a quick sandwich when you’re in a hurry and just need to get some food in your system vs. going out for a gourmet meal in an elegant dining environment. Sometimes your clients will want to watch something just to get the job done; other times, they will want the full immersive entertainment experience. They can’t do the latter if they don’t have a nice system. No one eats sandwiches all the time, right?
Home Theater Week: Sony’s Cinema-at-Home Vision
On the back end, you must deliver a system that includes, at a minimum, one streaming platform (e.g., Apple TV 4K, Shield TV, Roku, etc.) and one “premium” platform. To my knowledge, there are only two (outside of a DCI-type setup) that qualify for the latter: 4K Ultra HD Blu-ray and Kaleidescape. If you don’t include one or both of these, then you’re going to have a problem; the client will have no way to get top-tier material into the system — you’re limiting them to sandwiches all the time.
So, what is so problematic about streaming audio? I’ve discussed this before, but as a refresher, in a nutshell, the issue is bitrate. Streaming services give consumers the expectation of quality and consistency by throwing around catchphrases like “Dolby Digital Plus,” “5.1,” and “Dolby Atmos.” The subtle implication is that all 5.1 and Atmos are the same regardless of what streaming service you’re using. The truth is far from that. Dolby Digital Plus, which delivers the vast majority of 5.1 and immersive content from streaming services, is an absolutely fantastic codec. It’s versatile, efficient, and capable of delivering many forms of high-fidelity audio at low bitrates — but it can’t work miracles.
The common bitrates used for streaming are 192 kbps (kilobits per second), 256 kbps, 384 kbps, 448 kbps, 575 kbps, 640 kbps, and 768 kbps. Of these, Dolby Atmos requires at least 384 kbps (there are some limitations at 384 kbps that go away at higher bitrates), while 5.1 tops out at 640 kbps (at least that we’ve seen). Based on our internal, subjective evaluations of various bitrates, we consider the minimum for “acceptable” quality on a high-end system to be 384 kbps for 5.1 channels and 575 kbps for Atmos.
Unfortunately, not all streaming services measure up. Even more disappointingly, some services handle Atmos with excellent quality but skimp on 5.1. For example, we’ve been looking at a bunch of stuff on Disney+ recently. They use 768 kbps for Atmos, which is outstanding. Unfortunately, they drop all the way back to 256 kbps for 5.1 — and we hear the compression artifacts. Very disappointing. Contrast this to Netflix, which also uses 768 kbps for Atmos but only drops back to 640 kbps for 5.1. The worst part is that this information is not clearly communicated to consumers. At most, they will see little Dolby, 5.1, and/or Atmos badges on the content. The choice of what bitrate to use is left up to each service, with some guidelines from Dolby. There is little to no accountability to the streaming public.
Home Theater Week: Convenience, Comfort, and Community
I’ve also mentioned before that there are ways for you (and your clients, theoretically) to see the bitrate. I highly recommend that you familiarize yourself with the bitrates used by various services in order to steer your clients toward material that will sound good on their high-end systems. The Apple TV 4K has a feature called the Playback HUD that is available from the Developer Options menu. This menu is normally hidden but can be shown by syncing the Apple TV 4K to an Apple computer running Xcode. (This is currently the only way to do it, which in our opinion is an inexcusable measure that Apple simply must remedy.) The Playback HUD displays the codec, number of channels, and the audio bitrate (along with a lot of other useful things) in yellow text superimposed over the picture while content is playing. Just remember to turn it off when you’re finished, because the Developer Options menu has a nasty habit of hiding itself again (also inexcusable), and the Playback HUD will persist with no way to get rid of it (except to reconnect Xcode, of course)!
Fortunately, the Fire TV platform offers a far more accessible Developer Options app that can be downloaded directly from the store. It’s free (contributions encouraged) and doesn’t require any unnecessary hoop-jumping with a computer — you simply turn the overlay on/off in the app and then navigate to another app to whatever material you want to check. The overlay will give you the audio bitrate inside most apps, but that’s about it (at least on the audio side; video is much more detailed). You have to figure out the codec and channel count from other indicators. That’s pretty easy, really. The codec is going to be Dolby (unless the program is 2.0, in which case it is probably AAC), and you should be able to tell whether the channel count is 5.1 or Atmos by looking at the input format on the surround processor (sometimes 2.0 channel mixes are carried inside a 5.1 “container” format, though).
Allocentric vs. Egocentric Rendering
This is a severely under-reported topic in my opinion, and it can quickly escalate into some heavy engineering discussions (which is probably why it’s under-reported), but I’ll try to simplify it. For that reason, engineering-types please cut me some slack if you’re already well-versed; I’m covering broad strokes.
Basically, Atmos uses an allocentric rendering system that assigns XYZ coordinates (width, length, height) to audio objects (which include channels/beds) and speaker outputs within a 3D space. That 3D space is defined by the outer perimeter of the speaker system, not the walls of the room. The renderer then maps the audio object coordinates to the speaker coordinates. It does not take into account where any of the speakers are actually positioned in the room, nor where listeners are located relative to the speakers. Egocentric rendering, on the other hand, references the audio objects and the speaker outputs to points in 3D space relative to a central listener (or group of listeners) using azimuths and elevations.
Home Theater Week: The Ultimate Dual-Purpose Home Theater
I’m not advocating for one type of rendering or the other — in the end, it is all in how you manage the details — but those details are where I want to point out a couple of cases where Atmos may deliver a result that is not necessarily what you or your clients expect. Before I go further, I want to include the disclaimer that these observations are based on the specific performance of our reference equipment in our reference rooms and may not apply to all equipment in all Atmos installations. I’m providing these as examples of the type of thing you should look for in the render — it’s up to you to make sure your systems are behaving in the manner you expect!
One of the things we’ve noticed is that audio objects intended for the traditional Surround positions (i.e., slightly behind the listener) get rendered to both the Front and Surround speakers when the base layer consists of only five speakers (i.e., 5.1.X). I won’t go into why this happens, but the effect is that the speakers are asked to phantom image Surround to a point between the Front and Surround speakers — rather than coming solely from the Surround speakers as in flat 5.1. Apart from potentially shifting the soundfield farther forward than expected, two speakers playing each Surround can create phase-related anomalies that cause effects to land within the soundfield in the wrong place! This is even true for a single listener, unless the speakers are very carefully matched in terms of amplitude, phase, and dispersion characteristics (which rarely happens). As such, we strongly recommend that Atmos installations include a minimum of seven speakers on the base layer, specifically including Side and Back speakers (Dolby calls these Surround and Rear Surround).
Another detail with Atmos is an apparent prioritization of the “Z” coordinate relative to the “X” and “Y” coordinates during the rendering of full-Z objects (meaning they are at the very top of the soundfield). If only one pair of Top speakers is present, the renderer will use them even in extreme cases where the Front Height or Top Front speakers are at the front of the room and the object is supposed to be at the back of the room. This issue can be improved by using one pair of Top Middle speakers, but the potential for spatial disconnects still exists due to the X/Y-axis separation between the base layer and overhead layer speakers. Thus, we recommend a minimum of two pairs of Top speakers — one generally in front of the listeners and one behind — in order to minimize or eliminate the potential for spatial disconnects related to the rendering of positive-Z objects.
Home Theater Week: Demos That Are Designed to Dazzle
It should be noted that our recommendations affirm Dolby’s own. They also recommend 7.1.4 as the minimum Atmos layout for mixing and monitoring. Like I said above regarding Dolby Digital Plus and bitrate, Atmos is an awesome renderer; just don’t expect it to work miracles if you don’t give it enough speakers!
For homework, I suggest that you research the XYZ coordinates that the Atmos renderer uses for the common speaker positions. What you find may surprise you!
Bass Quality
Everybody loves great bass, but great bass is not always as simple as properly designing and calibrating the room and sound system. Due to the wide range of people and facilities these days that mix and monitor material, coupled with the technology package available in most consumer processors to handle speaker crossovers and bass summing, there are ways for bass to get mangled even if it looks like you’ve done everything right on paper. This is a topic all its own, so let this just serve as a teaser. We will get to the bottom of it in the next column. Stay tuned!
I want to leave you with the following thought. Before you hand over the keys, so to speak, to your client, demonstrate the system using specifically curated material that exemplifies how each aspect of the system is supposed to sound — dialog quality, fidelity, sound panning and positioning, bass performance, etc. Then, leave the material easily accessible to the client. Instruct the client to revisit that material any time there is a question about the system’s performance. Can they not understand the dialog on a show? Go back to the reference dialog passage. Does it sound OK? Is there no bass on that new music track? Go back to the bass reference and see if the bass is still there.
In this way, you can reassure the client that all is well and also train the client how to tell the difference between good material and bad — without having to “roll the trucks” to the site because the system is “not working.” Please use 4K Ultra HD Blu-ray or Kaleidescape for this, not streaming. Please. Even if the client literally never watches Blu-ray, give them a player and give them a disc or two of reference material. It’s well worth the extra hundred dollars or so to provide ongoing evidence that you did your job, and the system really is worth what the client paid for it!
Anthony Grimani is co-founder of Grimani Systems loudspeakers (www.grimanisystems.com), and president of PMI Engineering (www.pmiltd.com) and MSR Acoustics (www.msr-inc.com). Chase Walton contributed to this column.
