It's no secret that Automated Dialogue Replacement (ADR) is not the most revered part of post-production. ADR mixer Greg Crawford jokes that "there are three things people don't want to do…number one, go to the dentist. Number two, go to the DMV. Number three, do ADR."
Despite being often shunned or feared due to the perceived complexity of ADR (not to mention the extra work involved), it is a great way to improve audio quality and clean up less-than-perfect dialogue captured during production. Nowadays, the available tools and technologies make it a more viable, easier option than ever before.
First up, let's recap on what exactly ADR is. Automated Dialogue Replacement refers to the process of re-recording audio in post-production – typically in a more controlled, quieter setting.
There are a few main reasons why you may want to utilize ADR. These include:
While the technology has advanced, the principle of ADR is the same. In general, you've got two different ways to carry out ADR. First, there's visual ADR – the process by which an actor listens to a line of dialogue and then aims to match that line of dialogue while watching the performance on screen (with no sound).
Alternatively, you have audio ADR, where the actor will listen to the line of dialogue and recite that dialogue over and over, with the audio recording playing in their headset. It's agreed that audio ADR may give a slightly better, more accurate result, but these different processes come down to the preference of actor and director.
Because ADR is recorded in a "clean" environment with no interference, it will most likely feel very jarring and completely wrong if you edit it straight into a scene. Audiences will know almost immediately that this audio was not recorded on set, and it will make for awkward viewing.
Expert ADR mixer Greg Crawford (responsible for ADR on blockbusters including Avengers Infinity War and Spiderman) gave a great interview, shedding light on the process of achieving great ADR.
Ultimately, he always aims to keep as much of the original production performance as possible. Sometimes, his ADR is as nuanced as just pulling a syllable or one word to cover for a moment where there was a rustle of clothing on set. He calls this "surgical ADR."
However, there are other scenarios where full-blown ADR is called for. Greg discusses the "three Ps" of pitch, performance and placement. He's constantly looking to match the actor's ADR pitch and performance with the pitch and performance they gave during production. Alongside this, he's always looking to match the placement of the microphones on set.
To achieve this:
The first three-pointers here are self-explanatory, but the fourth requires a little more work and begins on set.
Reverb is a small detail, but it can make the difference between good ADR and great ADR. Considering you're already recording room tone anyway, this isn't too much extra hassle, as the folks at The Film Look expertly demonstrate on YouTube.
Naturally, when it comes to recording ADR, you'll need to decide which software you will use for recording, editing and syncing the new dialogue. While the example of Adobe Audition is used above, there are plenty of others to consider:
Alongside software, there's hardware. Greg Crawford touches on it briefly in his interview, confirming what you should already know. If you can, use the same microphones as used in production. It will make your job that much easier. When it comes to microphones, here are a few great choices to consider:
Meanwhile, the quality and compatibility of your hardware will undoubtedly impact the quality of your ADR. You may consider looking at intuitive controllers that adjust sound levels, sync and playback during your ADR session. Here are a few:
Of course, a great pair of noise-canceling headphones will be needed, too. Popular throughout the industry are:
Post-covid (and in general, in modern times), it's not always feasible to get everyone together in the same place at the same time. This is one of the most challenging aspects of ADR – coordinating and communicating with multiple people in different parts of the world and different time zones. Fortunately, there are some great tools out there that can enable you to carry out ADR remotely.
In the past few years, AI has been a hot topic of discussion, advancing at a blistering pace and infiltrating just about every part of the filmmaking process. ADR is no exception. What has typically been a long, painstaking process of refinement, requiring many takes and long hours, could soon get a whole lot easier and shorter.
Check out this astonishing example of TrueSync in the movie "Fall" (2022).
Using AI, TrueSync analyses and tracks the performance of Virginia Gardener, recreates her face with a 3D model, removes her original dialogue, and then replaces it flawlessly with language that can take the scene from R-rated to PG13. You just watched the future of ADR.
This example demonstrates a new and exciting way of conducting ADR – one that could save hours of time, as well as money and resources. Lengthy processes in software such as Adobe Audition, where you have to sync up multiple tracks and experiment with impulse response recording files, could be rendered unnecessary. Things could be achieved with a few clicks.
While we're not quite there yet, it's clear that lip-synced dubs with custom-made AI-recreated voices could become commonplace in the next five to ten years. TrueSync is not the only example of how AI is starting to make its mark on the ADR space.
Lovo.AI, for example, now offers the ability to clone any voice and create high-quality custom voice content. Similarly, Murf.AI offers the tools to "clone once. Use forever" – a foreboding phrase that would make any SAG-AFTRA member rightly shudder. When you listen to the examples on these sites, they're not quite there yet. But they are pretty impressive and, in some cases, definitely usable.
Other notable AI technologies doing a similar thing include Respeecher, which uses deep learning and neural networks to transform one voice into another (again, without requiring the original actors). And then there's Lyrebird, which creates a digital copy of your voice using a few minutes of audio samples. Then, you can generate a totally new dialogue or alter the existing dialogue with different emotions, tones and effects.
The writing is on the wall here. This technology is only going to advance, and as we've already seen, the rate of that advance is extremely fast.
So, that's the rundown on the state of ADR today and where it's headed tomorrow. It's no secret that ADR has often been a laborious, tedious process (remember Greg Crawford's joke about ADR being one of the only three things people hate to do). However, there are many great tools out there that ensure you can get the job done to the high-quality standards we expect from all big blockbusters.
As we move forward into a new post-production landscape in which AI increasingly dominates, we can expect to see drastic changes to how ADR is conducted in the future. If you told the folks experimenting with ADR in early 1930s Hollywood that one day, ADR may be an almost irrelevant, minuscule aspect of the post-production process, they may have fallen off their chairs.
MediaSilo allows for easy management of your media files, seamless collaboration for critical feedback, and out-of-the-box synchronization with your timeline for efficient changes. See how MediaSilo is powering modern post-production workflows with a 14-day free trial .