Issue link: https://digital.copcomm.com/i/1525212
Asher Pink. 51 F A L L Q 3 I S S U E T E C H the edits, and so on. It takes a human to have that judgment and that taste. AI isn't going to replace film crews. We work for real directors and real producers who have creative intent and creative tastes. That's not something a computer can judge." However, the committee does have con- cerns about how emerging technology will impact entry-level editors and their level of skill. Pink noted that some elementary tasks — ones delegated to apprentices or editors beginning their careers — could be handled with automation. So instead of having an entry-level editor cutting footsteps in sync to picture, that task could be automated. "That doesn't mean the work is going to go away. There is still going to be work. But that work is going to filter up to some of the higher function tasks, meaning there's going to be a larger skill gap that an en- try-level editor will need to surpass to make it up to professional. And this goes across the board — for sound and for picture. One of our concerns is that there will be the high skill/knowledge tasks but there won't nec- essarily be a ladder to get there," said Pink. Consider that those elementary-level tasks, like cutting in footsteps, teach a sound editor how to sync sounds to picture, how to attenuate a sound, use panning, and add processing to sell the idea that a sound is happening and moving within a space. Is proficiency in these basics necessary to achieve mastery later on? How will the use of new technologies change an editor's skill set? Miller said, "Our committee researches and advises the board of directors of the Editors Guild on what is coming and what impact that is going to have on our industry. We try to evaluate 'exposure.'" They explained that "exposure" is the sum of displacement and augmentation. It's the degree to which an occupation's tasks are susceptible to technological change. If a profession is highly "exposed," it means there's a new tool that can reduce the time it takes to do a task within that profession. In some cases, high exposure means that the job is in decline. Miller continued, "These professionals need to learn an additional skill set — learn how to use that new tool — to make them- selves marketable. For example, we found that transcribers are going to be highly impacted by AI; they have high exposure. So instead of transcribing, maybe they're organizing sound and running it through programs that are very good at doing tran- scripts. On the opposite end, picture editors for scripted projects have low exposure, because when you are answering to another creative, you have to understand what that creative person is looking for and a comput- er can't do that." Pink added, "We did find that the sound professions have a high rate of exposure, especially the closer they are to dialogue, just because there's this new tool that can generate very realistic sounding dialogue. We found that sound is definitely easier to create models off of than moving picture." STRICTLY PROCEDURAL Another development for film sound is the use of procedural audio. Pink and Miller see this shift in sound creation as more plausible than AI "text to sound effects" generators. Pink said, "Although generative sound is getting better, anyone who works in sound will attest to the fact that sound editing is not just cutting a sound to picture. It's choosing the right sound to deliver the right character. That is not something generative audio and a fully automated workflow can do." So what is procedural audio? In game sound, procedural audio is an audio system capable of synthesizing sounds in real-time according to a set of programmatic rules, algorithms, and mathematical models such as waveforms, noise, filters, or envelopes to control how the sound changes over t i m e — l i ke c h a n ge s i n p i tc h , vo l u m e, timbre, or rhythm. Historically, proce- dural audio does not use recorded audio material (it generates the sound), whereas procedural sound design requires an audio file that is then manipulated. (For more detail, check out Dara Crawford's concise blog post "What is Procedural Audio?" — https://daracrawford.com/audio-blog/ what-is-procedural-audio). Games now employ a combination of both procedural audio (generated sounds) and procedural sound design (pre-existing sounds) to great effect. Looking at "The Leg- end of Zelda: Tears of the Kingdom," during a 2024 GDC talk (Global Digital Cinema is an international company providing cinema solutions and services to distribution and exhibition companies), lead sound engineer Junya Osada explained how the sound team designed "a physics system for sound," as game director Hidemaro Fujibayashi called it. To very briefly summarize Osada's expla- nation, the sound team's approach mirrored the game physics team's approach in that both used a rule-based system to control how objects (and sounds) behave within the 3D space in-game. For example, one sound rule defined how loud a sound could be based on attenuation curves that incorporated acoustic phenom- ena like sound absorption, deflection, and occlusion. The audio system and the physics system worked together; information from the physics system fed the audio system, which could then automatically calculate changes in volume, and all the necessary processing (filters, reverb, and echo) as defined by the set of rules that all sounds in-game must follow. Additionally, sounds were attached to objects, and when the ob- jects were combined by the player in-game,