Today, Blackmagic Design held their pre-NAB livestream and introduced DaVinci Resolve 20. Resolve 20 was just one of many, many hardware and software products that BMD CEO Grant Petty demoed during the two-hour stream. Resolve 20 is a big update and follows BMD’s usual pattern, the beta comes first and is available now. Remember that you can’t install the shipping version of Resolve alongside the beta. A database upgrade is required when you’re updating to a new version of Resolve, so plan accordingly (and update your money-making production jobs at your own risk).

5-8 update: Talking to BMD at NAB they told me a database update isn’t required, it’s just a project update.

What’s new in DaVinci Resolve 20? First and foremost, it is AI. This update includes tons of new AI workflow tools. There was a mention of how much behind-the-scenes work the engineers have had to put in over the last few versions, rewriting old code, optimizing the application and supporting new hardware. Now that that work has been done, they can spend more time working on features. And we get many of those in this update. Here are the bullet points and some of my commentary!

AI IntelliScript generates a timeline based on your script. Transcribe your clips and pair them with your script, and Resolve will generate a timeline based on the comparison of the transcription and the script text. You’ll see a selected take on video track one, and then other takes will be stacked on tracks above with the clips disabled. It was mentioned that the takes don’t have to be exact, which is probably where the AI really comes into play. So variations on the take can still be viewed together. Often we think about AI as doing the work of an assistant, and this is definitely something that an assistant could do or could have done until AI started doing it.

New caption tools include animated subtitles which of course is all the rage on social media. There are so many third-party tools to emulate what CapCut has been doing so it will be nice to be able to keep some of that within the application.

AI Music Editor will make music edits to your specific, desired length. You can set a specific time or use the Live Trim option to trim music to a specific length. There’s an analysis process when you first enable this on the clip, but once it’s done, you can go to work. Like there’s an option to save different versions because rarely is the first one exactly right. When you’re done, decompose the AI music edit right in the timeline for your original clip music edited clip. There’s also a new beat analysis tool as well.

AI Audio Assistant will basically do an audio mix for you based on specific “standards.” They can do things like organize tracks and level out dialogue – a lot of that stuff that an audio engineer would do. And a lot of that stuff that editors who have to do their own mixes have to do.

A much-needed Keyframes panel will help with animation. And while it looks like you still have to go to the inspector to actually adjust the parameters, this new keyframes panel is a nice start to what can perhaps be more updates to keyframing in the edit page in general. You can also turn on keyframes at the bottom of the timeline as well.

AI Dialog Matcher will match dialog from one environment and or one mic to another. This is such a common task, it’s a type of place where I feel like AI can really shine. There are times when the non-audio engineer just cannot get dialogue to match, even when there’s sometimes shot. But in the same place with the same gear. As the environment changes, the presence and the sound of the voice change. This will be fun to try.

AI Mulitcam SmartSwitch will switch your camera angles based on who is talking. It’s not just based on whatever my castle is presence, but also the live stream mentioned it’s analyzing lip movement as well.

A really nice little update will let you open a timeline in a new Source Viewer. This replaces the previously clunky option where you had to swap the Source viewer timeline and the Timeline viewer timeline when you had loaded a timeline into the Source viewer. It kinda worked, so this new feature looks much better.

I also like that when you create a new project, you can assign where to place any new media that Resolve has to create. They no longer end up in the mysterious CacheClip file you find later, which is many upon many GBs (or TBs) large.

There was a mention that in the future they might have to start charging for DaVinci Resolve Studio upgrades. BMD has never charged for an upgrade, even when those were big upgrades so this will be a statement worth watching.

Undoubtedly, adding AI features to our legacy editing tools is the wave of the future. I like that BMD seems to be keeping these AI tools useful here, And integrating them in mostly into specific submenus feels like the right way to do it. When you need an AI assistant, then you go to a specific place to find what you need. And you know you’re bringing AI tools into the mix. Just a quick play with a lot of these things in the beta. They all seem quite speedy, though sometimes you’ll have to download a model before you can use the feature.

The public beta of Resolve 20 is now available, but just remember, we’ve got to uninstall the current version to install the beta.

• Import Blackmagic Cloud Shared folders to media pool.
• New optimized UI layouts option for vertical videos on cut, edit and color pages.
• Voiceover palette with cue, voice tools and teleprompter.
• Improved keyframing with dedicated curve view and timeline drawer.
• Text+ paragraph, line wrapping and bounding tools.
• MultiText tool with layers and easy transform, clip and wrap controls.
• Live overwrite now supports edit keys and search dial.
• Mouse drag to live overwrite multi source and sync bin.
• Trim with safe edit avoids overwriting adjacent clips.
• Full featured audio mixer added to cut page.
• Voiceover tool with record and monitor options in edit page.
• Chroma Color Warper grading.
• Display viewer overlays for remote monitoring.
• Stream and monitor H.265 4:2:2 on supported hardware.
• Deep image compositing toolset.
• Multi layer pipelining for OpenEXR, PSD and stereoscopic 3D.
• Vector warping toolset for image patching and cleanup.
• Fusion support for 180 VR.
• 3D Scene Dome Light.
• View color page grade in media out node on Fusion page.
• Clip EQ now features 6 bands.
• EQ and Level Matcher processes match tone and clip levels.
• EQ and Gain now also available as Fairlight FX plugins.
• Fairlight Chain FX to create and restore frequently chained plugins.
• Resolve FX AI Set Extender.
• AI IntelliScript creates timelines with a user provided script.
• AI Dialogue Matcher matches clip tone, level and reverberance.
• AI Music Editor automatically edits music to desired length.
• AI animated subtitles highlights or animates spoken words.
• AI Multicam SmartSwitch for automatic angle switching.
• AI Voice Convert with built-in and user trainable models.
• AI SuperScale now includes 3x and 4x enhanced upscaling.
• AI Magic Mask v2 with paint brush.
• AI Resolve FX Depth Map v2.
• AI IntelliCut to remove silences.
• AI IntelliCut to checkerboard dialogue by speaker.
• AI IntelliCut to generate ADR Cues using transcription speaker info.
• AI Audio Assistant automatically creates a finished mix.
• AI Detect Music Beats displays and allows beat snapping.

If you want to dig in deeper, give this video a watch. This hits all the good features.