If it’s AI-powered audio tools you’re after, then I hope you have a NVIDIA RTX GPU

NVIDIA
(Image credit: Future / Paul Hatton)

When I first began watching short films made by amateur videographers and directors, it wasn’t the camera angles or colour grading that elevated their production quality. It wasn’t even the resolution of the image or whether there was that much coveted film grain that became so popular in the 2010s.

In fact, it wasn’t anything to do with the video at all. Instead, it was the quality of the audio. It was clean, free from noise, and appropriately enhanced. It sounded like it had been recorded in a specialist studio. I remember searching for this holy grail, setting up DIY audio booths, and delving deep into audio editing. I had some success but never reached the bar I set for myself.

Thankfully, there've been a lot of advancements in the area of audio processing in recent years, and yes, it is courtesy of AI. But far from replacing, or heaven forbid, removing creativity, this new breed of tools simply offers an easy and affordable way for creatives to deliver studio-quality sound.

In the following, we unpack three common audio problems that creatives face all the time. We then point you in the direction of a tool that can help you overcome those problems so you don’t have to do a reshoot. Each of these tools benefits from NVIDIA RTX acceleration, and two of the three tools discussed are 100% free.

Broadcast

(Image credit: Future / Paul Hatton)

1. My audio is full of background noise

If you’ve ever live streamed or been on a video shoot, then you’ll understand first-hand that it can be very difficult to control what noise enters the space. It might be a dog barking in a house down the road. It could be an aeroplane flying overhead. Or it might be a room that is more echoey than you expected it to be.

Some of these challenges can be overcome with physical solutions. For example, you could bring some soft furnishings into the room to reduce echo or wait until the aeroplane has passed. Or, as many people are doing now, you can use software to remove that noise on the fly.

One of the best tools for achieving this is NVIDIA Broadcast. Not only is it free, but it’s also a master at noise removal and room echo cancellation. All you have to do is hook in your video feed (webcam, phone, or camera) and toggle the two microphone effects on. It’s as simple as that. There are no complicated settings, dials, or workflows. Turn it on and the software’s AI will take care of the rest.

DaVInci Resolve

(Image credit: Future / Paul Hatton)

2. My dialogue track is crowded with other audio

But what about if you’re working with pre-recorded video and audio rather than a live stream? It is possible to use NVIDIA Broadcast for this as well, but it’s not native and requires some workarounds. Instead, I’d recommend the RTX-accelerated AI effects in DaVinci Resolve.

My favourite AI effect in DaVinci Resolve at the moment is called ‘AI Voice Isolation’. It does what it says on the tin. You’ll find it in the Inspector panel of the Fairlight page, and all you have to do is toggle it on and choose how aggressively (between 0 and 100%) it removes background noise from the voice.

Rather than performing a traditional noise gate, AI Voice Isolation uses an AI model trained on a huge number of audio clips. As well as being remarkably good and consistent, it also works in real-time so you can apply the effect and use it throughout the general editing process.

Premiere Pro

(Image credit: Future / Paul Hatton)

3. My speech recording was poorly recorded

I’m yet to meet someone who enjoys cleaning up speech audio. It’s time-consuming and tedious, and it often leaves you wondering why the speech wasn’t recorded better in the first place. Also, usually when a sound editor receives speech to edit, it’s usually too late to perform a retake.

That’s why sound editors and video editors are now making more use of the AI audio tools found in Adobe Premiere Pro. One specific example is the Enhance Speech tool, which has been generally available for over a year now.

NVIDIA says this about the tool: “Accelerated by NVIDIA RTX, the new feature removes unwanted noise and improves the quality of dialogue clips so they sound professionally recorded. It’s 75% faster on a GeForce RTX 4090 laptop GPU compared with an RTX 3080 Ti.”

Using the tool is as simple as navigating to the ‘Essential Sound’ tab, toggling on the effect, and adjusting the Mix amount with a slider that goes from 0 to 100%. Then, when you hit 'Enhance', Premiere Pro will work its magic and deliver, after a short while, your drastically improved speech audio.

NVIDIA RTX is the way forward for audio editing

All of the software packages and tools mentioned so far are accelerated by NVIDIA RTX technology, so if you want the very best performance and experience, then you know which GPU lineup you should turn to.

Paul Hatton
Writer

Paul is a digital expert. In the 20 years since he graduated with a first-class honours degree in Computer Science, Paul has been actively involved in a variety of different tech and creative industries that make him the go-to guy for reviews, opinion pieces, and featured articles. With a particular love of all things visual, including photography, videography, and 3D visualisation Paul is never far from a camera or other piece of tech that gets his creative juices going. You'll also find his writing in other places, including Creative Bloq, Digital Camera World, and 3D World Magazine.