Dolby Atmos mixing setups typically consist of anywhere from 8 to 16 speakers, allowing the mixer to seamlessly pan sound around the entire room. However, to ensure pans sound natural and provide a stable image, it is important that the time for the sound to travel from each speaker to the mixer is exactly matched. As the speakers are likely to be at slightly different distances from the mixer, speaker delays are employed to achieve this.

This article gives an overview of the options to calculate and insert delays, from quick and easy to more accurate methods, including how to find professional assistance. The ear-level and overhead speakers are covered first, then LFE and bass management subs as they require some additional attention to ensure a good summation with other speakers.

The first step when implementing delays is to consider which ear-level and overhead speaker(s) you will not apply a delay to.  This will be the speakers which are the furthest away, which are commonly but not always, the rear surrounds. The goal therefore is to delay all the other speakers to match these speakers.

There are several ways to calculate the delay values:

Tape measure

This is the simplest method but has inherent issues. In this method, you note down the difference in distance between the furthest away speakers and every other speaker and then convert them to milliseconds to give your delay values.

This method, however, only considers the time-of-flight delays and not digital delays which can vary due to factors such as different loudspeaker software firmware versions, or filter latencies.

To use this method all you need is a tape measure and the Delay Calculator in the Commissioning sheet of the Dolby Atmos Room Design Tool. To access this sheet, choose either the ‘Advanced’ or ‘Music Advanced’ user mode in the ‘Room Information’ tab, then check the ‘Enable Room Commissioning’ checkbox in the ‘Design Creation’ sheet, and then click over to the ‘Commissioning’ sheet.

 

Recording impulses from each speaker

In this method you playback impulses from your DAW through each of your speakers, record them back into your DAW, and then zoom into the waveforms and measure the difference between the source and recorded versions. As with the tape measure method, you need to normalize the delays measured. That is, calculate the delay between playback and recorded impulse of the furthest away speaker(s), and then subtract each of the delays measured in your DAW for each of the other speakers from that value to give the differences.

Nathaniel Reichman has an excellent article detailing a full procedure complete with audio files on Production Expert.

 

Monitor controller, auto-calibration, or speaker management software

Many popular speaker management and/or auto-calibration software can perform all the measurements and implement the delay for you. For most, this is by far the easier solution and recommended where possible. Check the user documentation of your chosen device to explore this option further.

 

Audio Analyses software

Another method is to use analysis software such as Rational Acoustics’ SMAART software or Room EQ Wizard to determine the delay of each speaker using functions such as a ‘delay finder’ or similar. As above, you will need to normalize the delay per speaker to determine by how much to delay all but the furthest speakers by.

 

Subwoofers

When determining the correct delay for the LFE sub, a simple measure of distance or time may not give the best result.

The LFE channel commonly carries audio which is also present in other speakers. Ideally, therefore, it would be in-phase and sums well with as many other speakers as possible. In practical terms, however, it is only possible to achieve phase alignment with a single speaker due to the different distances from each speaker to the mix position, and the center speaker is the most common choice.

The location of your LFE sub is likely to have been determined by where it gives the most uniform output across its operating frequency range at the mix position. To find the best summation with the center is therefore achieved by altering the delay to the LFE channel by small increments.

To do this, first determine the delay using one of the above methods. Next, route a pink noise signal (band limited to 120Hz if possible) to both the LFE and the Center. Now incrementing the delay to the LFE sub by 0.1 ms from approximately 0.5 ms below the calculated value to 0.5 ms above it, while looking how the two speakers are summing on an frequency spectrum analyzer (i.e., a real-time analyzer). When you find the best summation across the frequency range of the LFE sub, note the delay value. If you do not have a frequency spectrum analyzer available, you can perform the procedure by inverting the phase of the LFE sub and listening for when you get the strongest cancellation at the mix position. Remember to flip the phase back afterward.

Bass management subs are normally time aligned to the nearest ear-level speaker they are managing which is commonly one of the Rear Surround speakers, and using the same procedure as above.

 

Where to insert the delays

After you have calculated the delay values, you need to determine where to insert them. Where possible, the best solution is to use your dedicated monitor controller unit/software as this path is common to all replay sources. Other options include in the Dolby Atmos Renderer application if utilized or the monitor section of your DAW if available. When using your DAW however, ensure delays are inserted in the monitoring chain and not the record path.

 

Certified Service Partner and Dolby Atmos Specialist

As mentioned, the options listed range from quick and easy, to more accurate and complex. Dolby partners with professionals around the globe to provide room commissioning and calibration services.  Further details can be found on our Professional Partners webpage.

 

For related information on studio design see our Home Entertainment Technical Guidelines and Best Practices for Dolby Atmos Music Studios.