I've been using it for many years, and it's grown in power and functionality with every release. BMD has offered free upgrades to Pro users (with no subscription) for many years, which is refreshing given the current state of software licensing.
For those upset with the new agent integration, keep in mind that one of the big knocks on DaVinci Resolve is that it's hard for beginners. Powerful tools like color grading and nodes take years to truly master (and that won't change), but making it easier for users to collaborate with the agents they already use to help learn and perform basic tasks should help with the learning curve.
Ultimately, professional video editing is an art. An agent might save time with asset management, project/timeline setup, etc., but it won't replace the real experience and taste required for larger projects. I would rather spend the extra time on the nuanced/deeper creative work.
Not only are the upgrades free, the full license is still only $295 and can be freely migrated between macOS, Linux, and Windows (unless something has changed, serial numbers can be activated on two systems at a time and transferred freely so long as you have Internet access).
I'm no expert, but I never found Resolve materially more difficult to learn or use than Final Cut Pro or Premiere. And if you're doing anything moderately complex, it's potentially easier to learn because things that require external tools in FCPX and Premiere are built in to Resolve Studio.
I suspect the agent integration is primarily workflow/UI related, but regardless, I don't see it as any more anti-creative than incorporating precanned effects and presets into a project, and potentially a very useful learning tool since the Resolve UI is quite good at letting you see how the sausage is made.
$295 one time permanent license with free upgrades and can be migrated between different computers?
I don't have a single reason to do any form of video editing, let alone professional video editing, but that kind of pricing model makes me want to buy a license just to make sure that companies like this stick around.
Black Magic supports all three platforms and they’re not exactly a large company. They show every day that there is no excuse/reason not to support multiple platforms. Anything else is just an excuse by a lazy developer. The developer for Blender is also the same. A user of both programs.
Black magic also make a program that support/ties into the camera hardware on smartphones and they support both Android and iOS.
> $295 one time permanent license with free upgrades and can be migrated between different computers?
I've been using Resolve since version 14, still getting upgrades to it! My license does only work on one "computer" at a time though, so when I switch between Linux and Windows, I need to re-enter and activate the license, as when I do that it deactivates the previous (other OS) one. You have seemingly unlimited amount of times to switch though, as I've probably done it on a bi-weekly basis for quite some time.
I don't know if something has changed, but if you buy resolve now you can have it activated on two computers at the same time. I've got my resolve studio license activated on my mac laptop and linux workstation at the same time.
Having a powerful video editor you can grab any time you need to is useful in a way few other bits of software are. Sure, for trimming and even some basic editing or watermarks, ffmpeg works great, but there are times when having full editor ui is beneficial
Taking away some of the more monotonous and "uncreative" tasks in any job can be taking away some of the variety in the work. Variety is nice. Variety affords us downtime. As the job gets stripped down to its core function, while a machine keeps the pipeline full so that the work can continue uninterrupted, the downtime slips away.
And great for beginners, now they get to pay some chatbot company to cover the gaps left by detereorating software design. Lucky they're already using agents... are they?
I have to disagree. You are already many levels away from the actual pixels. This is what software is for. Drawing a line for you personally is dismissive of the rest of the user base. Just because there is a so called agent added to the mix doesn't remove the human. It includes far more humans and enables them to use the tool. This is design if done right. Also you seem to be critiqueing the concept not the implementation.
Hard dissagree, it's just the pot boiling to harvest data and render future users helpless without the chatbot. Your stance is anti-human, and the robots and ceos will not appreciate your efforts in giving capital more power over creativity.
I don't sense that you have thought this out in depth. You are cycling talking points even if you don't mean too - that's what I hear.
I have a hard time seeing your way. I build software. Tools for humans to do stuff. Whatever they want. Then it gets complex. There are methods to make it more accessible to more people including so called agents. I want to go to there I want to enable the humans.
I feel you bro that you see human getting squeezed out. I do not see it that way. The human is still the nexus.
CEOs and tech bros and hype are all a distraction and a fog that will dissipate with time. We need to focus on the functionality and utility. We are engineers
You are a valuable person in the conversation and we all need you to participate with eyes wide open. I assume you are young. That's not discounting it is welcoming.
[edit: cleaned stuff up to hopefully make it more direct and non-confrontational]
This feels inverted to me: downtime affords us variety. If I have little downtime then I'll strongly prefer known, efficient methods of doing something, rather than experimenting with new ones.
> As the job gets stripped down to its core function, while a machine keeps the pipeline full so that the work can continue uninterrupted, the downtime slips away.
Editing to add my own salient tl;dr: shifting the "easy" tasks onto AI agents leaves the human ones to shoulder the cognitively demanding loads, resulting in burnout.
There are nuances, of course, but if you imagine that while once a human might have been responsible to move 500lbs a day, broken up into primarily small loads, now a robot is moving anything under, say, 20lbs and the human is left to only carry the heavier loads across the same time period so is now hefting 50 or 100lbs at a time and required to do across the same day, so now they're moving 1,500lbs a day.
I've been thinking a lot about this across many different mediums, not just filmmaking, and I think it's one worth digging deeper on.*
What is lost when losing the (understandably annoying) process of the grind? By letting LLMs do the busy work, do you lose the opportunity to stumble into randomness? Do you lose the understanding of what is possible/interesting with the medium?
And do you end up in this derivative future where decisions are made by a system that was trained on all the previous uses of it?
Obviously that's fairly hyperbolic, but I have absolutely felt the pull of it all. When it decides to take a couple extra steps I didn't ask for creatively and I think "eh, that's fine."
*EDIT: I wrote this and looked back and this SOUNDS like an LLM wrote it. See what I mean??????
yes to agents! how i had wished i could just point an agent at my footage and instruct "do something sensible with color grading and audio cleanup". i just want something presentable for a hobby video podcast. it doens't have to look a million bucks, and i have no time or desire to learn the intricacies involved. just something that would get me 70% (or 50%!) the way there would be more than fine.
Resolve is not any harder to use than any other NLE.
Although... BMD has actually made it harder, by adding bloat like the Cut page. I have attempted to find value in it a couple of times, and didn't. Now I disable it.
BMD needs to stop throwing crap into Resolve and buckle down and do two things:
1. Fix bugs, including UI defects. The Deliver page is a mess. The clip selection between the Edit timeline and the Fusion page makes no sense. And lots more.
2. Unify the node views. Resolve is a bunch of stuff BMD bought and threw together on different pages. The "integration" is janky and can, especially for the beginner, seriously degrade your work. Once example is how Fusion compositions inexplicably downgrade everything to timeline resolution. So you can bring 4K footage into a Fusion comp, but if your output timeline of the moment happens to be merely HD, the entire composition (including keys) will be degraded to work at that resolution. I mean... really? And beyond that, the relationship between Fusion comps and the timeline events they wind up in is extremely fragile and breaks arbitrarily and unrecoverably all the time.
A single nodeview for all operations would be a long-overdue game-changer for editing software.
Have you ever considered that you aren't the target audience for the cut page? It's supposed to be paired with a speed editor and is mostly intended for assembling rough cuts. You probably aren't cutting features or television or multicam so it's usefulness is probably out of your range of use.
No offense whatsoever. But this whole line "but it won't replace the real experience and taste required for larger projects" is not how I see it. I wish people are more humble at this juncture of AI progress. This is not to say humans won't play a part but saying xyz cannot be replaced is a big stretch. Saying its not there currently is okay but saying it won't replace I don't see where that's coming from.
The way I see it, it’s all about how you use it. Let people have tools; their output stands on its own legs.
As pro software, the UI has a learning curve. Many YouTube videos are out of date, or unnecessarily drawn out.
Additionally, as a hobbyist, project settings, color spaces, clipping, etc can be tricky to miss.
I’ve already been using Claude to (1) help me with UI, (2) QA my delivery, and (3) occasional fixes for audio repair; waveform sync (can’t afford time codes).
I make all creative decisions - including most menial ones - myself, I wouldn’t have it any other way; I enjoy the act of creating. I welcome this integration.
I’m sure some users are gonna agentically grade or whatnot. I don’t like it, but I’m also don’t believe in telling other people what they should do with their own work and creative process.
Because there has to be demand for it. There are hundreds of tasks that could be automated right now, and people still choose to do them by hand. Just because something is technically replaceable doesn’t mean it will be replaced and editing is a deeply personal part of the process for almost every director. It will change, but for serious work, it will never be a sure shot job handled entirely by AI. Sure, for low effort marketing videos you’d outsource to Fiverr, it’ll happen sooner or later but that’s not what we’re talking about here.
> ...but for serious work, it will never be a sure shot job handled entirely by AI.
Maybe not entirely handed by AI but when AI starts doing it better than most and in a fraction of the time, for a fraction of the cost, why do you think it wont be at least partly delegated to AI?
fair but that's also a matter of time, in my opinion.
1) software doesn't commit suicide like some pilots did.
2) software has superhuman reflexes, processing power and are not affected by emotion.
3) software also will eliminate a whole class of failure modes that caused many crashes in the past like: "I won't seek nearest and safer airport because that will be expensive in fuel costs"
> I won't seek nearest and safer airport because that will be expensive in fuel costs
If, and that's a big if, a pilot did choose a less safe airport because of monetary cost to his airline, it would be because those are the guidelines of the airline. And those guidelines would obviously be in the software too.
There's a fuzzy gray line here. Whether formalized or not, the way I see it, pilots will de facto be penalized for making decisions that cost an airline a ton of money, especially if it turns out to not have been an actual emergency.
I have no idea what legislation is in place regarding this topic and protecting pilots. But even with really good protections, there's no way that it won't affect your social standing with the higher ups if you divert and cost the company exorbitantly.
So pilots do have to make judgment calls about how much they're willing to risk their lives and the lives of their passengers versus a perceived risk, and worse, this judgment call exists in a weird fuzzy area where they're "allowed" to make the safe bet, but in practice, they will be punished for it.
Software doesn't (generally) have this luxury. It decides whether to land or continue flying based on some defined set of conditions.
It can still decide to risk the lives of the passengers. But that decision has to be approved by the airline, and can't live as an informal "you will lose brownie points with management" sword over the necks of the flight crew.
Unless we start to incorporate LLMs and probabilistic models into autopilots... but no-one would ever do that, right? ....
Because there is no demand for it to be fully handled by AI. Nobody is denying that parts will be handled by AI but that’s also you moving the goal post.
Coraline is a claymation movie made by hand, and Wildwood is coming out. But it’s because a cofounder of Nike loves stop motion animation. Despite critical acclaim these movies don’t make money. Just like we don’t have masterwork marble sculpture makers anymore. I have serious doubts anyone will surpass the works of Michaelangelo in the next hundred years; maybe a thousand.
The process is the point for any passionate creative. And yes, for every passionate creative you have a burned out creative who's trying to make ends meet who will use more AI centric tools because both that creative and the people paying them likely do not care how things look when they're completed, only that they're completed, on-time, and under budget.
However, again, an actual creative wouldn't. Imagine you and I are climbing a mountain: I'm climbing the mountain because I love climbing mountains. Climbing mountains is what I live for. I work full time and hard as hell so I can afford trips to climb mountains. You, on the other hand, have a terminal illness and you know that at the top of this mountain is a plant you can eat that will cure you with no side effects. So we're both doing this thing, but for WILDLY different reasons.
If as we're climbing a helicopter flies up and hovers behind us and the pilot says "hey I'll take you to the top of the mountain if you want!" we're going to have very different responses: you don't care about climbing mountains, and even if you do, you're not on this particular climb to enjoy it; you're here because you need this plant to cure your illness. You'd be crazy to not jump in that helicopter and get what you need to live. However, I'd be equally crazy to do the exact same thing, because climbing the mountain is why I'm here, I live for climbing mountains, and I want to climb this one.
This is what AI people do not get about creatives and why we talk past each other. AI people think creative people make things to have finished works, and that's certainly a part of it, and basically a requirement if you'd like to monetize your creativity. But like, I work in software as my day job and I'm bloody good at it, but I use AI to generate code sometimes because the process of developing software doesn't do nearly the same thing for me as creating art. I love making art, and I'd do it even if it made me no money. And I know that, because I've been doing it my whole life and it's never made me money. I feel depressed if I go too long between making works of art. I get ideas building up inside me and they itch and scratch at my being until I sit down and spend several hours turning said idea into something I can see, but like, if I just put a prompt into an AI image generator, I feel absolutely nothing doing that.
All of that to say: I genuinely do not believe AI will completely replace all of this, for no other reason than only a fraction (sizable though it may be) of creatives will ever want it, let alone use it. The quality of it and arguing about if it can "match" a human creative isn't the point, the humanity is the point. I don't want to look at AI generated art, and so I won't, and I especially do not want to replace my creative output with that of Midjourney. Ever. I don't care how good or cheap it is which is why those arguments just bounce off of me. I'm here to climb a mountain and you're offering me a staircase.
For me, the point of creativity is play, exploration, and experimentation. I've always jumped on new technology and new media because they're the best place to find those.
From what I've seen, a lot of the "it's a process" people are more interested in craft than art.
They're not creating anything unique, novel, or personal. They 'create' like an engineering problem - master this skill, copy this style, reproduce some acceptable and recognisable version of this popular imagery/content/genre.
There's a huge gap between competent craft and work that has a voice and is surprising - no matter how it's made.
People with that original voice are rare, and many aren't afraid of AI.
That's not to say that AI doesn't it make even easier to make uninteresting, unoriginal content. But the 'process' can also be fighting the default output quality to transcend that.
I agree. I enjoy creative play and learning how to emulate what the masters can do in film, music, art, engineering and computer science. When it's so easy to take so much for granted, I find it an excellent if somewhat time-consuming way to deepen my appreciation and admiration of their work.
> But this whole line "but it won't replace the real experience and taste required for larger projects" is not how I see it.
Why not? Personally (although I'm not saying others have to see it this way too), I'd wait until I can see what AI can replace, before believing so, rather than the opposite.
So far, AI isn't able to "set direction" at all in any sensible way, especially for a creative project like TV or movies, the results are just too bad (today). Although it's excellent at "following directions", which is why these new APIs are great and exciting, it lets people who are good at setting direction to leverage these tools to have the direction concretely followed by something, without having to create a group of collaborators.
Until we've seen proven that AI can indeed "set directions" ("have taste"), I think it's fair to assume they're unable to do so in a way that it produces something that humans generally enjoy. And very far from the kind of masterpieces a human can create.
The person you are replying to seems to have a depth of experience with editing over many years.
It seems pretty clear that AI capabilities are jagged. There are good reasons to believe based on research and the subjective evaluation of experts who are quite friendly to AI that there are some tasks creative tasks that AI has been really quite good at (totally disrupting stock illustrations) some where it does so much better than typical person, but not nearly as well as well as an expert (song creation) and some where, so far, it pretty much fails, and sees almost no progress even as models get better and better. (long form creative writting). Indeed, writting style sometimes seems to fall backwards with recent releases.
Also, it's subjectively, but commonly held to be true that most of the content created when AI can execute well lack taste, unless the human creating with AI has it. The term "slop" exists and in legible for a reason.
I mean, maybe this is just a question of different people speaking about different time horizons.
But it seems perfectly sensible for someone to say that there is a lot of automation that can happen in the near term, and that they see no reasonable evidence that AI can do interesting or even coherent work this space.
I don't know that shallow accelerationist boosterism (or panicked doomer despair ) that handwaves away the distinctive variations AI performnce is that enlightening.
I understand what you're saying and I agree, ostensibly, but on the caveat that it's hinged on the "we don't know what we don't know" aspect of reality. If we restrict our domain to what's forseeable, it's still apparently the case. Astra's demos have me scratching my head, diminishing returns are kicking in harder and faster than anticipated. It seems we've entered a phase to where frontier companies have completely abandoned improving the substrate because they can't justify it. It seems like it's all just harnesses and reasoning orchestration at this point. Hell, everybody but Google seems to have given up even on projection layer improvements (and Gemini is unusably bad in other ways).
Watching this next generation closely, because there are just a shitload of red flags starting to crop up that don't seem to be going away.
I would gladly enjoy to collaborate with an agent if I can learn more about video editing.
I edit my video only for passion and to learn something new though, so I guess that professionals may see this in a different way.
But I don’t think that it’s much different from how Claude code revolutionised the software scene.
Same. I edit lots of stuff in DaVinci, but am quite bad at it. As in, I spend forever figuring out how to do the stuff I want to do. I don't need the AI to invent stuff for me or take any decisions, just help me achieve my vision. Often it's quite simple stuff, where I'm forced to watch a too long video tutorial to figure out how to do something. But if I instead can ask it "help me to configure it so that the subtitles get baked into the video" or whatever it would be nice.
I really wish the Linux version supported h264 video and AAC audio. These are extremely disk space efficient codecs and even on 12 year old hardware you can edit 1080p videos without any CPU bound lag. Having to re-encode your source material and re-encode your exports is too tedious to do in practice.
On Windows I edited 100+ podcasts with Resolve and it's so good. Video editing on Linux for the last 9 months with Kdenlive has been killing me (vs Camtasia or Resolve on Windows).
And to bring up a likely reason: my impression is that H.264 and AAC still have active patents that apply to the major implementations, and those are not licensed gratis for all uses, and that applies more heavily to the encoding side. Raspberry Pi used to (and might still, I don't know) offer add-on license keys to unlock certain video codec support for what I assume was a similar reason. A lot of individual Linux desktop users just do their small-scale FFmpeg runs under the radar, but commercial software delivery is different.
With proprietary operating systems, the commercial entity behind the OS can negotiate for licensing. Some commercially-associated Linux distributions have efforts in similar directions, like Fedora for instance has some special secondary-repo stuff around H.264 support, but that's at minimum much more fragmentary as a target for an application developer concerned with the legalities.
The problem with AV1 is it requires a massive amount of compute resources. This affects both recording (encoding) and playback (decoding) and certainly affects editing. My AMD RX 480 (8 GB) card doesn't even support AV1 encoding too.
Being able to edit without lag is really important, also keeping your CPU load reasonably low when recording is important to avoid dropped frames, or your recording streams becoming out of sync (ie. a webcam video with audio recorded from another input device while you record your desktop).
I noticed on Linux, if I do anything CPU or disk intensive while recording (even with an SSD), I often see these things get out of sync at the point of load. It's an easy fix usually by just cutting the video where it starts to get out of sync and shifting the audio track over ~200-300ms, but it's something that could happen many times so it becomes an issue in the end. I recorded thousands of videos on Windows and never had this problem but it happens every few videos on Linux just doing casual screenshare videos.
AAC vs Opus is similar, except audio related.
In general when recording videos I usually use x264 as mkv which is CPU bound because it produces a much better looking video vs using h264. That's using a bit rate and other settings that tries to balance quality with low disk space.
On a low power machine you should use proxies. Davinci Resolve free version only supports patented codecs if the OS supports them. Microsoft and Apple pay licensing fees to use these codecs.
So? File size matters for delivery formats. But not for intermediate formats straight out of camera. For editing, bitrate and decode performance are often way more important.
Hardware decode support for AV1 and H265 is still pretty bad. If you use those formats, a lot of computers will stutter when you move the play head around while editing. On my mirrorless cameras, I much prefer filming in medium to high bitrate h264. I fill up my storage cards faster, but my camera batteries last longer recording like that and the footage is much easier to work with. I have a 4090, which supports h265 decode. But it only works for 4:2:0 footage. If I film in h265 422, my computer falls back to a software decoder. It's easier to just use h264, then I don't have to think about it.
If you want to ship high quality deliverables, you don't want resolve to do the final encode anyway. Resolve tries to use NVENC for encode, which is a much worse encoder than x264 and friends in ffmpeg. From resolve, output prores or something. Then use ffmpeg to do the final encode. I think there might also be ffmpeg encoding plugins.
tldr; Codec matters, but there's a lot more nuance to getting a good image than just picking the newest video codecs.
I've edited well over a thousand hours of video on my (Debian) DaVinci Resolve workstation. It's a rock-solid video editor that I can spend a day in without having to worry about a crash.
But I desperately wish Blackmagic would bring support for VST3 plugins and JACK to Linux. Or, at the very least, allow me to use my MIDI control surface in Fairlight on Linux.
As it stands, I have to do all of my audio production in Reaper, and that's a bit cumbersome.
> DaVinci Resolve 21.1 now supports integration with AI assistants such as Claude, Claude Code and ChatGPT Codex, allowing customers to analyze projects, organize media, adjust settings and batch render using everyday conversational language. Customers can ask their AI assistant to create highlight edits from long form video, remove unwanted clips and render deliverables, so they can spend more time being creative rather than getting bogged down in repetitive tasks.
I have videographer friends, and this is exactly what they've been asking for. Their dream is to be able to say "sync the A and B camera footage, chunk the whole interview down into segments, and then pull together the highlights based on this outline."
Not every application of AI is "do the creativity for me."
I would not be surprised if engineers who laid the groundwork for this back then would be happy with this workflow.
I am reminded of an interview I saw about the Stranger Things intro (before mainstream audiences got weird about the show). The producers wanted a practical effects intro, and consulted with people who actually made 1980s intros the show was trying to replicate.
The old guard said the show’s producers were crazy to do it with practical effects and should just use CGI since this was what they would have done, if they had the technology.
Edit: oh god, there is a nearly identical HN post from 2024 from me that I do not remember making. I swear I am not an LLM. I am also hoping that I am not an old man telling his stories over and over either: https://news.ycombinator.com/item?id=40453073
Is it that surprising that a bunch of old guys don't have the same energy and excitement as the new ambitious up-and-comers?
George Lucas seemed pretty satisfied when filming the Star Wars prequels to plop himself down in a chair in a room with a giant bluescreen set where he didn't have to walk around and direct in a hot, sandy desert for hours like in the original films. But the movies looked a lot worse for it.
The irony being that a lot of older practical VFX artists aren't able to tell when CGI is terrible. They're often so enamored with what the tech enables, that they're the last to see how quickly it becomes shoddy. Not sure what the equivalent of the hedonic treadmill is for CG, but it ages very rapidly, versus the best practical effects which maintain a charming physicality - even when they're no longer convincing.
> a lot of older practical VFX artists aren’t able to tell when CGI is terrible.
Source? As a CGI VFX artist in a previous life, I disagree vehemently, and I think you’re spinning an untrue narrative. I’ve never met any VFX artists, whether practical or CG (and I’ve met many), who weren’t keenly aware of how bad effects were and how poorly they age. The few practical folks left are way less enamored of CGI than the CGI people, and they were always that way (by definition?), so it seems like you’re mixing them up.
The fact that shoddy effects get used speaks to directors and budgets and production studios and what audiences pay for, it has almost nothing to do with the artists. That said, you’re also comparing bad CGI with good practical effects, which undermines your point. There were plenty of practical effects that lost their charm as fast as any CG ever did, and today there is plenty of CGI that remains convincing; it depends entirely on which effect and which shot in which movie you’re talking about.
Sorry, ask Lucas what exactly? How would that support parent’s claim that practical VFX artists can’t tell when CG is bad? Lucas hasn’t directed anything for over 20 years, and he was a writer, producer, director, and studio exec. He had already been answering questions about practical vs CGI for decades before he retired. Yes he’s overseen some CGI effects that people complained about a long time ago, and plenty more that nobody complained about, and he was perfectly aware of what people think didn’t age well. More relevantly, I know the artists he employed were aware of what would age well before the films were released.
I just watched the entire cycle of Don Quixote films by/about Terry Gilliam. That is a master class in how to ensure that your magnum opus is the worst film you've ever made.
I'm not complaining either, I wouldn't call myself a video editor, but I do end up spending some of my time in Resolve, and anything to remove the dreadful monotonous tasks of editing interviews is welcome. Although Resolve already makes it pretty easy, with the "edit by moving the transcript around" feature which is amazing. Anyway, I'm happy about being able to sick Codex at Resolve now :)
If we can get the environmental impact of this stuff ameliorated (a big if), this is the real benefit of this kind of AI assistance imo- it transforms everyone into a director and frees them from being a surrogate for a director.
See also: 3D modeling. Autodesk Fusion's interface is truly awful and people put years into learning it to do complex things. The MCP server functionality they added, though still in its infancy with Claude and what it can do, vastly lowers the barrier to entry for people just looking to get into making basic things. They can focus on the actual creative process rather than fumbling with the awkward tooling.
A lot of professional skill nowadays is misdirected from ‘knowing how to do the thing’ into ‘learning how to use a computer to do the thing’.
The role of music producer includes a large chunk of ‘knowing how to program synths and samples and drums and effects’. The role of industrial designer includes a lot of ‘knowing how to model things in CAD’. Even roles like academic researchers require skills like ‘formatting LaTeX’.
I’m genuinely quite excited for a shift in human computer interaction which reduces that ‘knowing how to do things on the computer’ tax.
But there are a whole swathe of jobs where a lot of what people do is ‘know how to do the thing on a computer’. Some of that is really because the bureaucracy has been embodied in a computer system so for example the job of handling an insurance claim seems to involve a lot of ‘knowing how to operate the claims computer system’ - actually that person’s skill is ‘knowing how to navigate the claims bureaucracy’.
Completely agree. How many people over the past 20 years have been musically gifted, but found ProTools or Logic impenetrable? The AI unlock is about enabling that creativity, not replacing it.
AI is elevating one’s capability but not replacing it.
Then you have experience and skills.
Someone with zero skills in music will be able to make a 'song' but a savvy producer would use tools differently and could leverage them to actually make something creative.
Anyone can make an app those days (or a website or whatever) but mileage varies based on one's nature, also the gotchas of someone with zero knowledge of a field could result issues that someone with experience would've noticed and resolved much earlier or avoided completely.
Me with zero knowledge in architecture can "design" my house but I wouldn't trust my plan without someone looking at it.
AI and AGI provides great lengths in empowering but it might also introduce the false feeling of being capable.
Creativity, capability and motivation still takes important place even with AI.
Exactly. Sometimes I just need a 3D printed part to finish a DIY project, I don't want to waste hours designing it. I had my fair share of time playing around with SketchUp when I was a teenager, but that's it.
For example: the guys on youtube using AI to make new Star Wars movies [1] still did tons of work putting it all together, including the 3d renderings and editing, and they are leaning on old scriptwriting/dialogue from a 1991 Star Wars book. AI is just helping them automate the process and generate some visuals/voice acting, but it still takes a skilled artist with lots of technical knowledge to guide it in the right direction and properly utilize their existing tools/workflows... and most importantly good taste.
It's already possible, but in my opinion it will never be popular. In some ways this has already been proven through other media; ChatGPT can easily generate a novel, but how many people are reading pure-AI novels?
Replace “novels” with music and go check YouTube for slop playlists and people happily eating it up.
The problem of AI “art” has never been that it cannot pass for the real thing. There are plenty of people buying slop ebooks on Amazon today. People will gladly generate a new episode of TNG.
And it will be popular because the discerning ones are the minority. Netflix is a multibillion dollar company serving a lot of mediocre stuff that people happily pay for just to turn off their brain for 2 hours every night.
Hence why I don’t care to justify why I avoid AI generated content at all costs, no matter how good it might look, with most people. It might sound snobbish to say, but they just wouldn’t get it.
These 'integrations' everywhere just mean we get more APIs that can be accessed programmatically, which I think is good for users, even if you don't plan to use to use AI. It's like the optimistic web 2.0 APIs everywhere days
No no, if you've ever spent serious time editing, it's a big deal -- if they pull this off correctly, it will be a game changer.
Editing has so much grunt work, where you just have to spending time organizing, collecting files, renaming things, finding missing files... Just tedious project management stuff that is excellent for an AI to do.
I'm not an experienced video editor, but I don't think the bulk of video editing involves generating new content on the fly except for narrow cases like adding titles, which is where an AI would easily be able to lean on scripting the existing tooling. There's a ton of grunt work involved in fixing up cuts and figuring out where everything is and so forth.
It’s because those codecs require SDK’s that are expensive to license. That’s not a Linux specific issue - it’s always been that way for the non-studio version of Resolve (with some exceptions on MacOS since the ProRes codecs are already part of the OS).
This is one of the the cases where “the language model won't necessarily respect IP law” is really relevant. Individual users may be able to get away with this because they're too scattershot to sue, and users in some jurisdictions may actually be fine regardless, but broadly, the patent licensing around things like H.264 is not something BMD can afford to ignore. AFAIK the biggest reason Google was pushing WebM in the first place was because they, a huge corporation, were balking at being under the thumb of the MPEG LA!
I appreciate that it can be confusing for users that patents can still apply to techniques where implementations exist that are both libre and gratis from a copyright perspective, but this has been an issue for Linux-adjacent software distribution for ages—like, see how Debian used to put certain stuff that was encumbered by patents in the US but not in other jurisdictions in special “non-US” repository sections, that sort of thing. It's not Black Magic's doing that the codec space is like this. I wish people wouldn't make so many mocking assumptions sometimes.
Haha yeah was about to post this very comment. I'm glad to see they have their priorities in order instead of checks notes supporting MP4 on Linux, the most common type of goddamn encoding. GoPro footage literally does not work with Resolve outside Windows.
The Claude connector is a nice idea, but there are lots of Python APIs missing in Resolve.
Even simple interactions like moving a clip 3 seconds in the timeline will not work with the MCP.
I just tried it, and got only the message:
"Resolve's scripting API also has no way to reposition an existing timeline item".
I've had Claude Code drive DaVinci Resolve before in three ways:
- Generate .otio files that are imported. Unfortunately, .otio doesn't fully support everything Resolve can do.
- Resolve has a scripting engine. Claude can write Lua scripts, then execute the script (through UI Automation)
- Computer Use. It's quite token intense, but it works.
I didn't try it, but for the paid version, someone also built a MCP server based on the scripting engine.
I'm programmatically adding analysis to sim racing lap guides, therefore my flow is probably different from the usual editing flow, but I have to say I found Claude Code + Opus to be quite capable of making edits with computer use.
I spent a few tokens last year trying to create a Python lib for native (undocumented) Resolve files. I was making progress, but it was slow (I ended up scripting Resolve in different ways). I imagine that current models could reverse-engineer the native file format relatively quickly.
I had AI write a program to edit with resolve. It does transcription, looks at the transcript and decides where to make cuts. If does a decent job and saves me a bunch of time.
I helped my girlfriend use the MCP server as she had a project where all the media files had been moved to a new source location, and using Codex, it managed to hunt down all the files and stitch it all back together. Ended up saving quite a lot of manual boring labour, and was quite cool to see it work in real time:)
I used IntelliScript a lot. I give Claude a subtitle export, ask it to cut it and import the cut subtitle track with IntelliScript. Resolve then creates a video version that’s cut based on the subtitle changes
I'm all for AI features here. The software is so feature packed and conplex, that after years of usage I still need to search and look up tutorial videos every single time I edit something.
How to ease key frames, how to fix my HDR captured footage for SDR, how to export the best quality video, how to burn in subtitles with word highlights, how to fix my audio aaand so on.
Either the AI will show me the particular feature and setting or a YouTube tutorial will. I think the AI might be faster and easier in this scenario.
Would be SO sick! I feel like there is such a big appetite for it and people would do the free labor to support it. I wonder what keeps it in this horrible purgatory of install hell.
This aur package [1] works well for me. Unfortunately the new maintainer removed the automatic download and you need to manually download the zip via the form on Blackmagic's website (however not even the email is validated).
The free version of resolve is incredibly generous already. There are no watermarks, and most features are included. The studio version is a once-off cost, and honestly its pretty reasonable. If you're doing really serious scripting, maybe upgrade?
In the mean time, the free version of resolve 21.0.4 will keep working indefinitely.
Have anyone tried using this as a substitute for CaptureOne/Lightroom? Any experiences? I'm aware it's a quite new feature added to the suite, but would love to one day stop paying the pricey subscriptions the other two alternatives demand
I tried and as a individual (single/one at a time) image editor is it fantastic, beats Lightroom by a mile and the node editor gives me the ability to "emulate" panels from CaptureOne (and OFX plugin support means I can just LLM my own if I need).
Library management is where Lightroom (and honestly CaptureOne) still wins, I couldn't find a good way to manage a 100k photo library.
I created one project and tried importing the whole library, but it took more than 8 hours and didn't make meaningful progress (~10%?) so I stopped (compared to 15hr for the whole thing in Lightroom).
I also did a project for about 1K photos and while the import worked, the library management itself was lacking. I could not find ways to do image stacks for bursts, group by metadata, etc. The obvious filters are there, but I think it needs more work.
I spent a little while trying. My knowledge of lightroom didn't really carry over. I couldn't work out how to do things like masks. I'm sure if i spent longer and watched some tutorials I'd get further but I decided to just let the photo mode cook a little longer before I try to switch over.
I make motorcycle travel videos for YouTube, and the bulk of where my time goes in the preparation of a timeline: chronologically cutting the highlights, synchronising dual camera footage. If Claude + Davinci Resolve will enable me to cut this time slightly, even if it's 10%, it's a welcome change. Ever since I made the jump from Premiere Pro, I've been positively surprised with how good Resolve is. Excited to try this out!
Yes, correct, there’s a feature for that. Unfortunately one of the audio channels from the camera footage on the motorcycle is never good enough for this feature to work for me.
i hope blackmagic pays attention to what steam is doing with steamos.
steam has put so much into making gaming on linux so good, one of the last pieces remaining for a ton of people to make the switch is editing.
don’t get me wrong, i run davinci on my non-rocky linux desktop and it’s almost entirely painless, but, what steam is doing with their distro of linux is incredible.
if they were to consider migrating to steamos rather than rocky, between the masses switching to linux since gaming is good enough for so many, if davinci was mostly solid, i think blackmagic could pull millions away from the adobe hell they’re stuck in on windows.
particularly since davinci’s new photo tab is fucking incredible and will so soon be a perfect replacement for lightroom.
As an anecdote, I recently needed to colour-grade some short footage for personal purposes. My workflow at the time was to essentially ask an LLM for guidance on how to use DaVinci Resolve to perform the task, step by step. I would make an adjustment, take a screenshot of the controls and the result, and repeat that process a handful of times, eventually ending up with a pretty nice result.
I can definitely see the utility of something like this for cases where a media budget just wouldn't have existed in the first place.
I used Google's Agent Development Kit to make a super simple agent outside of Resolve that I can use to say "take a 10s clip from file1.mp4 starting at 00:01:25 and a 20s clip from file2.mp4 starting at 00:02:00" then tell it to preview and render. Speeds things up.
I’ve been using it in Linux for >2 years, including their special cutting keyboard (blanking on the name).
Works without a hitch, including multi monitor, gpu acceleration etc
I use it on Manjaro, installed via AUR.
That said, IIRC it only works with nvidia GPUs and requires X11, and is supposedly finicky to get the environment right for (but the AUR gets it right for me)
The good thing about DaVinci Resolve's Linux system requirements is that these are so outdated that basically every developer links their plugins with static libraries they build themselves; and Resolve itself bundles lots of .so files with the app itself. Also Rocky Linux 8 is a good baseline Linux distribution: simple system libraries interdependence, so trivial to export and bundle those.
In other words, binaries created for RockyLinux8 with portability in mind are easily portable to sane modern Linux distributions.
(That doesn't make developing on Rocky Linux 8 fun at all)
The biggest issues are the supports of video cards (nvidia is the king) and different popular codecs. The problem is partly solved by using paid version, but the gpu/codec support is still not even with Windows, so you have to use ffmpeg to re-encode from/into supported formats.
Runs flawlessly on my strix halo Ubuntu 24 machine after fixing the install process [1]. I have an ffmpeg script that batch converts all my files into the supported codec.
Tried to work with ut for 5 months on Ubuntu. Installation was painful. Have to use some tutorial. No AAC audio codec support as well, which is the code of my smartphone camera. Randomly decided to not work after Ubuntu updates. I'm on Windows now...
I don't know about recent versions, but I managed to install 18 on Linux Mint.
Making it run wasn't easy, and it can't decode MP3 and some video codecs because of licensing issues. It also can't decode WebM. This is kind of a problem because of a lot of free music you can download to add to videos and a lot of video clips you can download from Pexels for example are encoded with codecs it can't decode, so you have to re-encode it with FFMPEG first to a supported codec.
Seems like an unusually boring Resolve release. There's nothing there I particularly want to try out. Nothing wrong with that, I've just got used to every release having a ton of shiny new stuff in it.
I guess the agentic stuff is probably exciting to somebody.
Resolve is extremely RAM hungry. For 4K video I wouldn't want to be editing with less than 12GB of VRAM. I'm guessing that excludes most integrated graphics.
Source - I work as a professional editor (amongst other things). Please commenters, do not come at me with your five minute low bit rate projects that required only a LUT and some contrast and 'worked fine' in DaVinci on a potato.
They must be planning to be bought by one of these companies. Otherwise I dont see why they would not just continue with their strategy of integrating AI tools within their product. They just reduced DR to just another tool. This is like opening a five star restaurant in a food court at the mall.
For people like me who have no idea what this is, I looked it up:
DaVinci Resolve is a proprietary application for non-linear video editing, color correction, color grading, visual effects, and audio post-production. It is developed by the Australian company Blackmagic Design for macOS, Windows, iPadOS and Linux.
...
DaVinci Resolve has been used for the color grading and/or editing of feature films such as Alien: Covenant, Avatar, Best of Enemies, Deadpool 2, Jason Bourne, Kingsman: The Golden Circle, La La Land, Love & Mercy, the Pirates of the Caribbean series, Prometheus, Anora, Robin Hood, Spectre, Star Wars: The Last Jedi, and X-Men: Apocalypse.
My latest 'free' computer motherboard has an i7-6700K, with integrated graphics. I tried to run Resolve 20 but it complained about my not having a real GPU. I ended up getting an Intel Arc B580 for around $300, which I thought was a better value than getting a used GPU off eBay.
Going to look at this version, as I haven't gone through the tutorials yet, so the assistants would be quite helpful.
I don't see how companies like this can survive going forwards. DaVinci Resolve is effectively ~200 Astra/Fable prompts, maybe even fewer if you're willing to forgo polish.
I'd give you $100K no questions asked if you manage to get something even close to Resolve with 2K Astra/Fable prompts, polish included. You vastly underestimate how much work goes into "polish" when software with huge surface area is decades old.
I think there's a really interesting tension between what you're saying and what the parent post is saying, and it's something that's been discussed for a long time but LLMs are throwing a new wrench into it: a given user probably only uses 5-10% of the features of a piece of software, but across the whole body of users you get close to full coverage.
I use DaVinci Resolve for editing GoPro, Canon DSLR, iPhone, and Drone FPV footage. I import clips, use the razor to cut them and rearrange them, occasionally crop and zoom, and then encode them down to h264 or whatever. There are a bazillion features that I sort of know exist but haven't ever had any need to use. Could I make a tool using an LLM that covers my DaVinci workflow? Almost certainly, and I don't think it would take anywhere near 200 prompts. I've made significantly more complex bespoke data analysis pipelines for drone flight telemetry in like... a dozen prompts. But I have no illusions that what I would make as my "DaVinci Clone" would have anywhere near the functionality of DaVinci, just the tiny piece of that functionality that I use.
I'm pretty sure your editing workflow can be achieved in the Preview app on macOS. If your editing workflow is just basic chopping clips out of a longer video then yes, you could vibe code it.
> a given user probably only uses 5-10% of the features of a piece of software
Then vibe your own 5% editor, but saying that a software so mature and complete as Resolve can be done with some AI and time is something I expect coming from the mouth of a teenager, not an adult.
regardless of whether it’s possible, there is enormous value in the community, ecosystem, and learning materials centred around Reaolve as it is right now. the workflow has been refined over years. no one shot slop effort will have that network effect behind it, and if the workflow isn’t a drop-in replacement it’s dead in the water anyway.
You can say anything, why stop at claiming 200 prompts, why not say you think Astra/Fable can one shot Davinci Resolve! And you can EASILY make the BM cinema cameras and other hardware by prompting Fable to run a 3D printer. (≧▽≦)
The free tier is very generous and it’s only $300, no subscription. Those Fable credits would cost a lot more than that. You’d spend thousands of dollars to save $300.
Moreover, not everyone wants to have the hassle of building and maintaining their own version of a thing that isn’t compatible with anyone else’s version of the same thing. And if you hire anyone to help you, now you have to train them on how to use your app. If they leave, those skills are useless for their next job.
(I’m not even covering the idea that you could do it in 200 prompts. I got Claude to create a video player and that alone took over 300 prompts.)
Black Magic Design's primary business is selling cinematography hardware. Anyways, those ~200 prompts are going to cost you more money than a license once you factor in the cost of your own time. Value proposition of AI still ain't there, and it probably won't be until the hardware bubble pops, and it very well might never pop without market intervention. This is assuming your 200 prompts is a realistic figure, which I'm frankly skeptical of.
you do know this company also sells myriad of hardware , can't Fable those just yet. and I am sure majority of their income comes from that part of their business, selling equipment to studios etc.
I've been using it for many years, and it's grown in power and functionality with every release. BMD has offered free upgrades to Pro users (with no subscription) for many years, which is refreshing given the current state of software licensing.
For those upset with the new agent integration, keep in mind that one of the big knocks on DaVinci Resolve is that it's hard for beginners. Powerful tools like color grading and nodes take years to truly master (and that won't change), but making it easier for users to collaborate with the agents they already use to help learn and perform basic tasks should help with the learning curve.
Ultimately, professional video editing is an art. An agent might save time with asset management, project/timeline setup, etc., but it won't replace the real experience and taste required for larger projects. I would rather spend the extra time on the nuanced/deeper creative work.
Not only are the upgrades free, the full license is still only $295 and can be freely migrated between macOS, Linux, and Windows (unless something has changed, serial numbers can be activated on two systems at a time and transferred freely so long as you have Internet access).
I'm no expert, but I never found Resolve materially more difficult to learn or use than Final Cut Pro or Premiere. And if you're doing anything moderately complex, it's potentially easier to learn because things that require external tools in FCPX and Premiere are built in to Resolve Studio.
I suspect the agent integration is primarily workflow/UI related, but regardless, I don't see it as any more anti-creative than incorporating precanned effects and presets into a project, and potentially a very useful learning tool since the Resolve UI is quite good at letting you see how the sausage is made.
$295 one time permanent license with free upgrades and can be migrated between different computers?
I don't have a single reason to do any form of video editing, let alone professional video editing, but that kind of pricing model makes me want to buy a license just to make sure that companies like this stick around.
Black Magic supports all three platforms and they’re not exactly a large company. They show every day that there is no excuse/reason not to support multiple platforms. Anything else is just an excuse by a lazy developer. The developer for Blender is also the same. A user of both programs.
Black magic also make a program that support/ties into the camera hardware on smartphones and they support both Android and iOS.
> $295 one time permanent license with free upgrades and can be migrated between different computers?
I've been using Resolve since version 14, still getting upgrades to it! My license does only work on one "computer" at a time though, so when I switch between Linux and Windows, I need to re-enter and activate the license, as when I do that it deactivates the previous (other OS) one. You have seemingly unlimited amount of times to switch though, as I've probably done it on a bi-weekly basis for quite some time.
I don't know if something has changed, but if you buy resolve now you can have it activated on two computers at the same time. I've got my resolve studio license activated on my mac laptop and linux workstation at the same time.
Having a powerful video editor you can grab any time you need to is useful in a way few other bits of software are. Sure, for trimming and even some basic editing or watermarks, ffmpeg works great, but there are times when having full editor ui is beneficial
They make their money from hardware (and hopefully will continue to be able to do so).
This is commoditizing Adobe for them.
Highly recommend the studio version. The ability to edit by transcript is worth the price alone.
Or buy the ~$400 speed editor and get the license for free. That's what many people do.
Taking away some of the more monotonous and "uncreative" tasks in any job can be taking away some of the variety in the work. Variety is nice. Variety affords us downtime. As the job gets stripped down to its core function, while a machine keeps the pipeline full so that the work can continue uninterrupted, the downtime slips away.
And great for beginners, now they get to pay some chatbot company to cover the gaps left by detereorating software design. Lucky they're already using agents... are they?
I have to disagree. You are already many levels away from the actual pixels. This is what software is for. Drawing a line for you personally is dismissive of the rest of the user base. Just because there is a so called agent added to the mix doesn't remove the human. It includes far more humans and enables them to use the tool. This is design if done right. Also you seem to be critiqueing the concept not the implementation.
I don't mean to be harsh, only passionate.
Hard dissagree, it's just the pot boiling to harvest data and render future users helpless without the chatbot. Your stance is anti-human, and the robots and ceos will not appreciate your efforts in giving capital more power over creativity.
I don't sense that you have thought this out in depth. You are cycling talking points even if you don't mean too - that's what I hear.
I have a hard time seeing your way. I build software. Tools for humans to do stuff. Whatever they want. Then it gets complex. There are methods to make it more accessible to more people including so called agents. I want to go to there I want to enable the humans.
I feel you bro that you see human getting squeezed out. I do not see it that way. The human is still the nexus.
CEOs and tech bros and hype are all a distraction and a fog that will dissipate with time. We need to focus on the functionality and utility. We are engineers
You are a valuable person in the conversation and we all need you to participate with eyes wide open. I assume you are young. That's not discounting it is welcoming.
[edit: cleaned stuff up to hopefully make it more direct and non-confrontational]
> Variety affords us downtime
This feels inverted to me: downtime affords us variety. If I have little downtime then I'll strongly prefer known, efficient methods of doing something, rather than experimenting with new ones.
> As the job gets stripped down to its core function, while a machine keeps the pipeline full so that the work can continue uninterrupted, the downtime slips away.
This I agree with, though.
This concept of, say, "productive downtime" was the topic of an article I encountered/skimmed not long ago: "There are no easy calls anymore" https://www.customerexperiencedive.com/news/agents-are-overl...
Editing to add my own salient tl;dr: shifting the "easy" tasks onto AI agents leaves the human ones to shoulder the cognitively demanding loads, resulting in burnout.
There are nuances, of course, but if you imagine that while once a human might have been responsible to move 500lbs a day, broken up into primarily small loads, now a robot is moving anything under, say, 20lbs and the human is left to only carry the heavier loads across the same time period so is now hefting 50 or 100lbs at a time and required to do across the same day, so now they're moving 1,500lbs a day.
> Taking away some of the more monotonous and "uncreative" tasks in any job can be taking away some of the variety in the work. Variety is nice.
Ultimately, and this applies to so much of life, "to each, their own," right?
I've been thinking a lot about this across many different mediums, not just filmmaking, and I think it's one worth digging deeper on.*
What is lost when losing the (understandably annoying) process of the grind? By letting LLMs do the busy work, do you lose the opportunity to stumble into randomness? Do you lose the understanding of what is possible/interesting with the medium?
And do you end up in this derivative future where decisions are made by a system that was trained on all the previous uses of it?
Obviously that's fairly hyperbolic, but I have absolutely felt the pull of it all. When it decides to take a couple extra steps I didn't ask for creatively and I think "eh, that's fine."
*EDIT: I wrote this and looked back and this SOUNDS like an LLM wrote it. See what I mean??????
> Taking away some of the more monotonous and "uncreative" tasks in any job can be taking away some of the variety in the work.
Thank. God.
I never want to rotoscope again.
And frankly I'm not fond of editing either.
yes to agents! how i had wished i could just point an agent at my footage and instruct "do something sensible with color grading and audio cleanup". i just want something presentable for a hobby video podcast. it doens't have to look a million bucks, and i have no time or desire to learn the intricacies involved. just something that would get me 70% (or 50%!) the way there would be more than fine.
Resolve is not any harder to use than any other NLE.
Although... BMD has actually made it harder, by adding bloat like the Cut page. I have attempted to find value in it a couple of times, and didn't. Now I disable it.
BMD needs to stop throwing crap into Resolve and buckle down and do two things:
1. Fix bugs, including UI defects. The Deliver page is a mess. The clip selection between the Edit timeline and the Fusion page makes no sense. And lots more.
2. Unify the node views. Resolve is a bunch of stuff BMD bought and threw together on different pages. The "integration" is janky and can, especially for the beginner, seriously degrade your work. Once example is how Fusion compositions inexplicably downgrade everything to timeline resolution. So you can bring 4K footage into a Fusion comp, but if your output timeline of the moment happens to be merely HD, the entire composition (including keys) will be degraded to work at that resolution. I mean... really? And beyond that, the relationship between Fusion comps and the timeline events they wind up in is extremely fragile and breaks arbitrarily and unrecoverably all the time.
A single nodeview for all operations would be a long-overdue game-changer for editing software.
Have you ever considered that you aren't the target audience for the cut page? It's supposed to be paired with a speed editor and is mostly intended for assembling rough cuts. You probably aren't cutting features or television or multicam so it's usefulness is probably out of your range of use.
No offense whatsoever. But this whole line "but it won't replace the real experience and taste required for larger projects" is not how I see it. I wish people are more humble at this juncture of AI progress. This is not to say humans won't play a part but saying xyz cannot be replaced is a big stretch. Saying its not there currently is okay but saying it won't replace I don't see where that's coming from.
The way I see it, it’s all about how you use it. Let people have tools; their output stands on its own legs.
As pro software, the UI has a learning curve. Many YouTube videos are out of date, or unnecessarily drawn out.
Additionally, as a hobbyist, project settings, color spaces, clipping, etc can be tricky to miss.
I’ve already been using Claude to (1) help me with UI, (2) QA my delivery, and (3) occasional fixes for audio repair; waveform sync (can’t afford time codes).
I make all creative decisions - including most menial ones - myself, I wouldn’t have it any other way; I enjoy the act of creating. I welcome this integration.
I’m sure some users are gonna agentically grade or whatnot. I don’t like it, but I’m also don’t believe in telling other people what they should do with their own work and creative process.
Because there has to be demand for it. There are hundreds of tasks that could be automated right now, and people still choose to do them by hand. Just because something is technically replaceable doesn’t mean it will be replaced and editing is a deeply personal part of the process for almost every director. It will change, but for serious work, it will never be a sure shot job handled entirely by AI. Sure, for low effort marketing videos you’d outsource to Fiverr, it’ll happen sooner or later but that’s not what we’re talking about here.
> ...but for serious work, it will never be a sure shot job handled entirely by AI.
Maybe not entirely handed by AI but when AI starts doing it better than most and in a fraction of the time, for a fraction of the cost, why do you think it wont be at least partly delegated to AI?
the "never" is what gets me here.
Autopilots have been able to fly planes better than humans at almost zero cost for years yet no one has replaced pilots.
fair but that's also a matter of time, in my opinion.
1) software doesn't commit suicide like some pilots did.
2) software has superhuman reflexes, processing power and are not affected by emotion.
3) software also will eliminate a whole class of failure modes that caused many crashes in the past like: "I won't seek nearest and safer airport because that will be expensive in fuel costs"
4) software is already landing planes and saving lives in the process: https://youtu.be/XawXSLRCM3o
5) Most part of a comercial flight is already controlled by software anyway. Pilots mostly take off/land and input coordinates and altitudes.
6) push for fully automous aircraft during modern war is going to speed up development and test of the tech.
I think it's a matter of time here too.
> I won't seek nearest and safer airport because that will be expensive in fuel costs
If, and that's a big if, a pilot did choose a less safe airport because of monetary cost to his airline, it would be because those are the guidelines of the airline. And those guidelines would obviously be in the software too.
There's a fuzzy gray line here. Whether formalized or not, the way I see it, pilots will de facto be penalized for making decisions that cost an airline a ton of money, especially if it turns out to not have been an actual emergency.
I have no idea what legislation is in place regarding this topic and protecting pilots. But even with really good protections, there's no way that it won't affect your social standing with the higher ups if you divert and cost the company exorbitantly.
So pilots do have to make judgment calls about how much they're willing to risk their lives and the lives of their passengers versus a perceived risk, and worse, this judgment call exists in a weird fuzzy area where they're "allowed" to make the safe bet, but in practice, they will be punished for it.
Software doesn't (generally) have this luxury. It decides whether to land or continue flying based on some defined set of conditions.
It can still decide to risk the lives of the passengers. But that decision has to be approved by the airline, and can't live as an informal "you will lose brownie points with management" sword over the necks of the flight crew.
Unless we start to incorporate LLMs and probabilistic models into autopilots... but no-one would ever do that, right? ....
Nothing about modern airplanes prevents pilots from doing a Malaysian 370.
At this point that's a political problem, not a technical one.
Because there is no demand for it to be fully handled by AI. Nobody is denying that parts will be handled by AI but that’s also you moving the goal post.
Coraline is a claymation movie made by hand, and Wildwood is coming out. But it’s because a cofounder of Nike loves stop motion animation. Despite critical acclaim these movies don’t make money. Just like we don’t have masterwork marble sculpture makers anymore. I have serious doubts anyone will surpass the works of Michaelangelo in the next hundred years; maybe a thousand.
What makes you say that? You think people like paying professionals to edit their videos? Surely some do, but everyone?
you're right. ads are already infested with AI videos which means it works for some and those are no longer paying the price of fully manual editing.
Maybe read again.
> the "never" is what gets me here.
The process is the point for any passionate creative. And yes, for every passionate creative you have a burned out creative who's trying to make ends meet who will use more AI centric tools because both that creative and the people paying them likely do not care how things look when they're completed, only that they're completed, on-time, and under budget.
However, again, an actual creative wouldn't. Imagine you and I are climbing a mountain: I'm climbing the mountain because I love climbing mountains. Climbing mountains is what I live for. I work full time and hard as hell so I can afford trips to climb mountains. You, on the other hand, have a terminal illness and you know that at the top of this mountain is a plant you can eat that will cure you with no side effects. So we're both doing this thing, but for WILDLY different reasons.
If as we're climbing a helicopter flies up and hovers behind us and the pilot says "hey I'll take you to the top of the mountain if you want!" we're going to have very different responses: you don't care about climbing mountains, and even if you do, you're not on this particular climb to enjoy it; you're here because you need this plant to cure your illness. You'd be crazy to not jump in that helicopter and get what you need to live. However, I'd be equally crazy to do the exact same thing, because climbing the mountain is why I'm here, I live for climbing mountains, and I want to climb this one.
This is what AI people do not get about creatives and why we talk past each other. AI people think creative people make things to have finished works, and that's certainly a part of it, and basically a requirement if you'd like to monetize your creativity. But like, I work in software as my day job and I'm bloody good at it, but I use AI to generate code sometimes because the process of developing software doesn't do nearly the same thing for me as creating art. I love making art, and I'd do it even if it made me no money. And I know that, because I've been doing it my whole life and it's never made me money. I feel depressed if I go too long between making works of art. I get ideas building up inside me and they itch and scratch at my being until I sit down and spend several hours turning said idea into something I can see, but like, if I just put a prompt into an AI image generator, I feel absolutely nothing doing that.
All of that to say: I genuinely do not believe AI will completely replace all of this, for no other reason than only a fraction (sizable though it may be) of creatives will ever want it, let alone use it. The quality of it and arguing about if it can "match" a human creative isn't the point, the humanity is the point. I don't want to look at AI generated art, and so I won't, and I especially do not want to replace my creative output with that of Midjourney. Ever. I don't care how good or cheap it is which is why those arguments just bounce off of me. I'm here to climb a mountain and you're offering me a staircase.
No, the process is the point for you.
For me, the point of creativity is play, exploration, and experimentation. I've always jumped on new technology and new media because they're the best place to find those.
From what I've seen, a lot of the "it's a process" people are more interested in craft than art.
They're not creating anything unique, novel, or personal. They 'create' like an engineering problem - master this skill, copy this style, reproduce some acceptable and recognisable version of this popular imagery/content/genre.
There's a huge gap between competent craft and work that has a voice and is surprising - no matter how it's made.
People with that original voice are rare, and many aren't afraid of AI.
That's not to say that AI doesn't it make even easier to make uninteresting, unoriginal content. But the 'process' can also be fighting the default output quality to transcend that.
I agree. I enjoy creative play and learning how to emulate what the masters can do in film, music, art, engineering and computer science. When it's so easy to take so much for granted, I find it an excellent if somewhat time-consuming way to deepen my appreciation and admiration of their work.
For anything that can create better margins, and that is not outlawed, there will be demand.
> But this whole line "but it won't replace the real experience and taste required for larger projects" is not how I see it.
Why not? Personally (although I'm not saying others have to see it this way too), I'd wait until I can see what AI can replace, before believing so, rather than the opposite.
So far, AI isn't able to "set direction" at all in any sensible way, especially for a creative project like TV or movies, the results are just too bad (today). Although it's excellent at "following directions", which is why these new APIs are great and exciting, it lets people who are good at setting direction to leverage these tools to have the direction concretely followed by something, without having to create a group of collaborators.
Until we've seen proven that AI can indeed "set directions" ("have taste"), I think it's fair to assume they're unable to do so in a way that it produces something that humans generally enjoy. And very far from the kind of masterpieces a human can create.
Yes people really need to take the AI’s feelings into account before they say things like that.
The person you are replying to seems to have a depth of experience with editing over many years.
It seems pretty clear that AI capabilities are jagged. There are good reasons to believe based on research and the subjective evaluation of experts who are quite friendly to AI that there are some tasks creative tasks that AI has been really quite good at (totally disrupting stock illustrations) some where it does so much better than typical person, but not nearly as well as well as an expert (song creation) and some where, so far, it pretty much fails, and sees almost no progress even as models get better and better. (long form creative writting). Indeed, writting style sometimes seems to fall backwards with recent releases.
Also, it's subjectively, but commonly held to be true that most of the content created when AI can execute well lack taste, unless the human creating with AI has it. The term "slop" exists and in legible for a reason.
I mean, maybe this is just a question of different people speaking about different time horizons.
But it seems perfectly sensible for someone to say that there is a lot of automation that can happen in the near term, and that they see no reasonable evidence that AI can do interesting or even coherent work this space.
I don't know that shallow accelerationist boosterism (or panicked doomer despair ) that handwaves away the distinctive variations AI performnce is that enlightening.
I understand what you're saying and I agree, ostensibly, but on the caveat that it's hinged on the "we don't know what we don't know" aspect of reality. If we restrict our domain to what's forseeable, it's still apparently the case. Astra's demos have me scratching my head, diminishing returns are kicking in harder and faster than anticipated. It seems we've entered a phase to where frontier companies have completely abandoned improving the substrate because they can't justify it. It seems like it's all just harnesses and reasoning orchestration at this point. Hell, everybody but Google seems to have given up even on projection layer improvements (and Gemini is unusably bad in other ways).
Watching this next generation closely, because there are just a shitload of red flags starting to crop up that don't seem to be going away.
Taste is subjective - that's why it can't be replaced.
I would gladly enjoy to collaborate with an agent if I can learn more about video editing. I edit my video only for passion and to learn something new though, so I guess that professionals may see this in a different way. But I don’t think that it’s much different from how Claude code revolutionised the software scene.
Same. I edit lots of stuff in DaVinci, but am quite bad at it. As in, I spend forever figuring out how to do the stuff I want to do. I don't need the AI to invent stuff for me or take any decisions, just help me achieve my vision. Often it's quite simple stuff, where I'm forced to watch a too long video tutorial to figure out how to do something. But if I instead can ask it "help me to configure it so that the subtitles get baked into the video" or whatever it would be nice.
> it won't replace the real experience and taste required for larger projects
[Citation needed]
I really wish the Linux version supported h264 video and AAC audio. These are extremely disk space efficient codecs and even on 12 year old hardware you can edit 1080p videos without any CPU bound lag. Having to re-encode your source material and re-encode your exports is too tedious to do in practice.
On Windows I edited 100+ podcasts with Resolve and it's so good. Video editing on Linux for the last 9 months with Kdenlive has been killing me (vs Camtasia or Resolve on Windows).
A few weeks ago I used claude to reverse engineer the linux version of resolve to add support for AAC media. The patch is here, and it works great:
https://github.com/josephg/resolve-aacfix
AAC streams are decoded and sent to ffmpeg.
H.264 is supported on Linux in DR Studio, it’s just not in the free version.
And to bring up a likely reason: my impression is that H.264 and AAC still have active patents that apply to the major implementations, and those are not licensed gratis for all uses, and that applies more heavily to the encoding side. Raspberry Pi used to (and might still, I don't know) offer add-on license keys to unlock certain video codec support for what I assume was a similar reason. A lot of individual Linux desktop users just do their small-scale FFmpeg runs under the radar, but commercial software delivery is different.
With proprietary operating systems, the commercial entity behind the OS can negotiate for licensing. Some commercially-associated Linux distributions have efforts in similar directions, like Fedora for instance has some special secondary-repo stuff around H.264 support, but that's at minimum much more fragmentary as a target for an application developer concerned with the legalities.
H.264 isn't efficient though. AV1 (and even H.265) are anywhere from 20-50% more space efficient for the same visual quality.
AAC vs Opus is a little bit more debated, but on average should outperform AAC.
The problem with AV1 is it requires a massive amount of compute resources. This affects both recording (encoding) and playback (decoding) and certainly affects editing. My AMD RX 480 (8 GB) card doesn't even support AV1 encoding too.
Being able to edit without lag is really important, also keeping your CPU load reasonably low when recording is important to avoid dropped frames, or your recording streams becoming out of sync (ie. a webcam video with audio recorded from another input device while you record your desktop).
I noticed on Linux, if I do anything CPU or disk intensive while recording (even with an SSD), I often see these things get out of sync at the point of load. It's an easy fix usually by just cutting the video where it starts to get out of sync and shifting the audio track over ~200-300ms, but it's something that could happen many times so it becomes an issue in the end. I recorded thousands of videos on Windows and never had this problem but it happens every few videos on Linux just doing casual screenshare videos.
AAC vs Opus is similar, except audio related.
In general when recording videos I usually use x264 as mkv which is CPU bound because it produces a much better looking video vs using h264. That's using a bit rate and other settings that tries to balance quality with low disk space.
On a low power machine you should use proxies. Davinci Resolve free version only supports patented codecs if the OS supports them. Microsoft and Apple pay licensing fees to use these codecs.
So? File size matters for delivery formats. But not for intermediate formats straight out of camera. For editing, bitrate and decode performance are often way more important.
Hardware decode support for AV1 and H265 is still pretty bad. If you use those formats, a lot of computers will stutter when you move the play head around while editing. On my mirrorless cameras, I much prefer filming in medium to high bitrate h264. I fill up my storage cards faster, but my camera batteries last longer recording like that and the footage is much easier to work with. I have a 4090, which supports h265 decode. But it only works for 4:2:0 footage. If I film in h265 422, my computer falls back to a software decoder. It's easier to just use h264, then I don't have to think about it.
If you want to ship high quality deliverables, you don't want resolve to do the final encode anyway. Resolve tries to use NVENC for encode, which is a much worse encoder than x264 and friends in ffmpeg. From resolve, output prores or something. Then use ffmpeg to do the final encode. I think there might also be ffmpeg encoding plugins.
tldr; Codec matters, but there's a lot more nuance to getting a good image than just picking the newest video codecs.
I've edited well over a thousand hours of video on my (Debian) DaVinci Resolve workstation. It's a rock-solid video editor that I can spend a day in without having to worry about a crash.
But I desperately wish Blackmagic would bring support for VST3 plugins and JACK to Linux. Or, at the very least, allow me to use my MIDI control surface in Fairlight on Linux.
As it stands, I have to do all of my audio production in Reaper, and that's a bit cumbersome.
Question: does it allow easy speed ramp/curved time remapping and what is the effect on the audio when making gradual playback speed changes?
Not even BMD is safe from the agent apocalypse:
> DaVinci Resolve 21.1 now supports integration with AI assistants such as Claude, Claude Code and ChatGPT Codex, allowing customers to analyze projects, organize media, adjust settings and batch render using everyday conversational language. Customers can ask their AI assistant to create highlight edits from long form video, remove unwanted clips and render deliverables, so they can spend more time being creative rather than getting bogged down in repetitive tasks.
I have videographer friends, and this is exactly what they've been asking for. Their dream is to be able to say "sync the A and B camera footage, chunk the whole interview down into segments, and then pull together the highlights based on this outline."
Not every application of AI is "do the creativity for me."
I would not be surprised if engineers who laid the groundwork for this back then would be happy with this workflow.
I am reminded of an interview I saw about the Stranger Things intro (before mainstream audiences got weird about the show). The producers wanted a practical effects intro, and consulted with people who actually made 1980s intros the show was trying to replicate.
The old guard said the show’s producers were crazy to do it with practical effects and should just use CGI since this was what they would have done, if they had the technology.
Edit: oh god, there is a nearly identical HN post from 2024 from me that I do not remember making. I swear I am not an LLM. I am also hoping that I am not an old man telling his stories over and over either: https://news.ycombinator.com/item?id=40453073
Is it that surprising that a bunch of old guys don't have the same energy and excitement as the new ambitious up-and-comers?
George Lucas seemed pretty satisfied when filming the Star Wars prequels to plop himself down in a chair in a room with a giant bluescreen set where he didn't have to walk around and direct in a hot, sandy desert for hours like in the original films. But the movies looked a lot worse for it.
"guy with bad knees recommends we all take the gondola down"
The irony being that a lot of older practical VFX artists aren't able to tell when CGI is terrible. They're often so enamored with what the tech enables, that they're the last to see how quickly it becomes shoddy. Not sure what the equivalent of the hedonic treadmill is for CG, but it ages very rapidly, versus the best practical effects which maintain a charming physicality - even when they're no longer convincing.
> a lot of older practical VFX artists aren’t able to tell when CGI is terrible.
Source? As a CGI VFX artist in a previous life, I disagree vehemently, and I think you’re spinning an untrue narrative. I’ve never met any VFX artists, whether practical or CG (and I’ve met many), who weren’t keenly aware of how bad effects were and how poorly they age. The few practical folks left are way less enamored of CGI than the CGI people, and they were always that way (by definition?), so it seems like you’re mixing them up.
The fact that shoddy effects get used speaks to directors and budgets and production studios and what audiences pay for, it has almost nothing to do with the artists. That said, you’re also comparing bad CGI with good practical effects, which undermines your point. There were plenty of practical effects that lost their charm as fast as any CG ever did, and today there is plenty of CGI that remains convincing; it depends entirely on which effect and which shot in which movie you’re talking about.
Ask George Lucas.
Sorry, ask Lucas what exactly? How would that support parent’s claim that practical VFX artists can’t tell when CG is bad? Lucas hasn’t directed anything for over 20 years, and he was a writer, producer, director, and studio exec. He had already been answering questions about practical vs CGI for decades before he retired. Yes he’s overseen some CGI effects that people complained about a long time ago, and plenty more that nobody complained about, and he was perfectly aware of what people think didn’t age well. More relevantly, I know the artists he employed were aware of what would age well before the films were released.
I just watched the entire cycle of Don Quixote films by/about Terry Gilliam. That is a master class in how to ensure that your magnum opus is the worst film you've ever made.
>Not every application of AI is "do the creativity for me."
However it can lead to AI "doing the creative for them" and them out of a job in the near future.
Sure, in the future everything will have a healthy piss-tint yellow shine. =3
I'm not complaining either, I wouldn't call myself a video editor, but I do end up spending some of my time in Resolve, and anything to remove the dreadful monotonous tasks of editing interviews is welcome. Although Resolve already makes it pretty easy, with the "edit by moving the transcript around" feature which is amazing. Anyway, I'm happy about being able to sick Codex at Resolve now :)
If we can get the environmental impact of this stuff ameliorated (a big if), this is the real benefit of this kind of AI assistance imo- it transforms everyone into a director and frees them from being a surrogate for a director.
See also: 3D modeling. Autodesk Fusion's interface is truly awful and people put years into learning it to do complex things. The MCP server functionality they added, though still in its infancy with Claude and what it can do, vastly lowers the barrier to entry for people just looking to get into making basic things. They can focus on the actual creative process rather than fumbling with the awkward tooling.
A lot of professional skill nowadays is misdirected from ‘knowing how to do the thing’ into ‘learning how to use a computer to do the thing’.
The role of music producer includes a large chunk of ‘knowing how to program synths and samples and drums and effects’. The role of industrial designer includes a lot of ‘knowing how to model things in CAD’. Even roles like academic researchers require skills like ‘formatting LaTeX’.
I’m genuinely quite excited for a shift in human computer interaction which reduces that ‘knowing how to do things on the computer’ tax.
But there are a whole swathe of jobs where a lot of what people do is ‘know how to do the thing on a computer’. Some of that is really because the bureaucracy has been embodied in a computer system so for example the job of handling an insurance claim seems to involve a lot of ‘knowing how to operate the claims computer system’ - actually that person’s skill is ‘knowing how to navigate the claims bureaucracy’.
Completely agree. How many people over the past 20 years have been musically gifted, but found ProTools or Logic impenetrable? The AI unlock is about enabling that creativity, not replacing it.
Joe said it right in Halt and Catch Fire:
> Computers aren't the thing. They're the thing that gets us to the thing.
That was that character’s take on Jobs’s ‘bicycle for the mind’ concept.
Just finished watching the series. What a great show.
And now those people are crushed by a wave of AI slop music
AI is elevating one’s capability but not replacing it.
Then you have experience and skills.
Someone with zero skills in music will be able to make a 'song' but a savvy producer would use tools differently and could leverage them to actually make something creative.
Anyone can make an app those days (or a website or whatever) but mileage varies based on one's nature, also the gotchas of someone with zero knowledge of a field could result issues that someone with experience would've noticed and resolved much earlier or avoided completely.
Me with zero knowledge in architecture can "design" my house but I wouldn't trust my plan without someone looking at it.
AI and AGI provides great lengths in empowering but it might also introduce the false feeling of being capable.
Creativity, capability and motivation still takes important place even with AI.
Exactly. Sometimes I just need a 3D printed part to finish a DIY project, I don't want to waste hours designing it. I had my fair share of time playing around with SketchUp when I was a teenager, but that's it.
Blender with fable and Astra is doing pretty amazing things it seems..
Yeah, exactly. I cannot wait to use this to take care of the boring stuff.
That's why there are no assistant editor jobs anymore.
Pretty soon there will be no need for recording and editing .
IIRC House MD was one of the first shows to use digital (5D mkII).
Soon there will be AI generated shows. Perhaps entirely personalized.
Youtube, generate me a TNG episode.
> Youtube, generate me a TNG episode.
It will never be that simple.
For example: the guys on youtube using AI to make new Star Wars movies [1] still did tons of work putting it all together, including the 3d renderings and editing, and they are leaning on old scriptwriting/dialogue from a 1991 Star Wars book. AI is just helping them automate the process and generate some visuals/voice acting, but it still takes a skilled artist with lots of technical knowledge to guide it in the right direction and properly utilize their existing tools/workflows... and most importantly good taste.
[1] https://www.youtube.com/watch?v=t4gZR1BDzOw
We can always streamline creativity by allowing agents to post jobs on mechanical Turk
> We can always streamline creativity by allowing agents to post jobs on mechanical Turk
Which is maybe why Mechanical Turk will be shutdown in 3 weeks. No more people laundering AI.
Not anymore! [0]
[0]: https://www.mturk.com
It's already possible, but in my opinion it will never be popular. In some ways this has already been proven through other media; ChatGPT can easily generate a novel, but how many people are reading pure-AI novels?
Replace “novels” with music and go check YouTube for slop playlists and people happily eating it up.
The problem of AI “art” has never been that it cannot pass for the real thing. There are plenty of people buying slop ebooks on Amazon today. People will gladly generate a new episode of TNG.
And it will be popular because the discerning ones are the minority. Netflix is a multibillion dollar company serving a lot of mediocre stuff that people happily pay for just to turn off their brain for 2 hours every night.
Hence why I don’t care to justify why I avoid AI generated content at all costs, no matter how good it might look, with most people. It might sound snobbish to say, but they just wouldn’t get it.
https://x.com/ratlimit/status/2011250910471815650
… but set it in the Star Wars universe.
And full of product placements...
Resolve has had de-aging filters, and TTS voice synthesis for awhile.
If the chat-bot/spy-ware aren't integrated, than I don't see a problem with LLM plug-in opt-in support. =3
These 'integrations' everywhere just mean we get more APIs that can be accessed programmatically, which I think is good for users, even if you don't plan to use to use AI. It's like the optimistic web 2.0 APIs everywhere days
No no, if you've ever spent serious time editing, it's a big deal -- if they pull this off correctly, it will be a game changer. Editing has so much grunt work, where you just have to spending time organizing, collecting files, renaming things, finding missing files... Just tedious project management stuff that is excellent for an AI to do.
They've added it as people are opting for stuff like Palmier Pro, FableCut, and likes, which can edit videos using AI.
I can imagine the conversation with Studio lawyers about contaminating IP with uncopyrightable generative content liabilities. =3
I'm not an experienced video editor, but I don't think the bulk of video editing involves generating new content on the fly except for narrow cases like adding titles, which is where an AI would easily be able to lean on scripting the existing tooling. There's a ton of grunt work involved in fixing up cuts and figuring out where everything is and so forth.
So let them? A product that focuses on chasing the competition is effectively useless, because the competition will always be ahead.
LOVE this. Game changer in capability and it'd kill Resolve if they didn't do it.
Furthermore, can one use custom API endpoint or is it forced use of corpo AI?
They'd rather add agentic AI support instead of adding proper codec support on the free Linux version (which is already supported on Windows).
smh
It’s because those codecs require SDK’s that are expensive to license. That’s not a Linux specific issue - it’s always been that way for the non-studio version of Resolve (with some exceptions on MacOS since the ProRes codecs are already part of the OS).
Is it surprising that they’d rather make features that cater to a large portion of their user base instead of a tiny fraction of it?
Someone was posting recently how it cost him tens of dollars of LLM Agent time to reverse engineer Linux Davinci and add codec he wanted back in.
This is one of the the cases where “the language model won't necessarily respect IP law” is really relevant. Individual users may be able to get away with this because they're too scattershot to sue, and users in some jurisdictions may actually be fine regardless, but broadly, the patent licensing around things like H.264 is not something BMD can afford to ignore. AFAIK the biggest reason Google was pushing WebM in the first place was because they, a huge corporation, were balking at being under the thumb of the MPEG LA!
I appreciate that it can be confusing for users that patents can still apply to techniques where implementations exist that are both libre and gratis from a copyright perspective, but this has been an issue for Linux-adjacent software distribution for ages—like, see how Debian used to put certain stuff that was encumbered by patents in the US but not in other jurisdictions in special “non-US” repository sections, that sort of thing. It's not Black Magic's doing that the codec space is like this. I wish people wouldn't make so many mocking assumptions sometimes.
Haha yeah was about to post this very comment. I'm glad to see they have their priorities in order instead of checks notes supporting MP4 on Linux, the most common type of goddamn encoding. GoPro footage literally does not work with Resolve outside Windows.
The Claude connector is a nice idea, but there are lots of Python APIs missing in Resolve. Even simple interactions like moving a clip 3 seconds in the timeline will not work with the MCP. I just tried it, and got only the message: "Resolve's scripting API also has no way to reposition an existing timeline item".
I've had Claude Code drive DaVinci Resolve before in three ways: - Generate .otio files that are imported. Unfortunately, .otio doesn't fully support everything Resolve can do. - Resolve has a scripting engine. Claude can write Lua scripts, then execute the script (through UI Automation) - Computer Use. It's quite token intense, but it works.
I didn't try it, but for the paid version, someone also built a MCP server based on the scripting engine.
I'm programmatically adding analysis to sim racing lap guides, therefore my flow is probably different from the usual editing flow, but I have to say I found Claude Code + Opus to be quite capable of making edits with computer use.
I spent a few tokens last year trying to create a Python lib for native (undocumented) Resolve files. I was making progress, but it was slow (I ended up scripting Resolve in different ways). I imagine that current models could reverse-engineer the native file format relatively quickly.
I had AI write a program to edit with resolve. It does transcription, looks at the transcript and decides where to make cuts. If does a decent job and saves me a bunch of time.
I helped my girlfriend use the MCP server as she had a project where all the media files had been moved to a new source location, and using Codex, it managed to hunt down all the files and stitch it all back together. Ended up saving quite a lot of manual boring labour, and was quite cool to see it work in real time:)
I used IntelliScript a lot. I give Claude a subtitle export, ask it to cut it and import the cut subtitle track with IntelliScript. Resolve then creates a video version that’s cut based on the subtitle changes
Works pretty well!
> I'm programmatically adding analysis to sim racing lap guides
Yosh is that you?
No, I'm Chris :) But if you know Yosh well, would be happy to connect and compare notes!
I'm all for AI features here. The software is so feature packed and conplex, that after years of usage I still need to search and look up tutorial videos every single time I edit something.
How to ease key frames, how to fix my HDR captured footage for SDR, how to export the best quality video, how to burn in subtitles with word highlights, how to fix my audio aaand so on.
Either the AI will show me the particular feature and setting or a YouTube tutorial will. I think the AI might be faster and easier in this scenario.
I install, run and update via this container script [1] on Arch. It works very well.
[1] https://github.com/zelikos/davincibox
This aur package [1] works well for me. Unfortunately the new maintainer removed the automatic download and you need to manually download the zip via the form on Blackmagic's website (however not even the email is validated).
1: https://aur.archlinux.org/packages/davinci-resolve
did they remove it based on policy? seems.. less convenient.
This is a good place to share that MakeResolveDeb makes installing Davinci Resolve very easy on Ubuntu/Mint/Debian
Way more people should know about this solution.
https://www.danieltufvesson.com/makeresolvedeb
That's the experience on CachyOS. They package it up for their users and updates just work.
"Advanced scripting now requires DaVinci Resolve Studio" :thumbs_down
https://forum.blackmagicdesign.com/viewtopic.php?f=21&t=2398...
The free version of resolve is incredibly generous already. There are no watermarks, and most features are included. The studio version is a once-off cost, and honestly its pretty reasonable. If you're doing really serious scripting, maybe upgrade?
In the mean time, the free version of resolve 21.0.4 will keep working indefinitely.
Have anyone tried using this as a substitute for CaptureOne/Lightroom? Any experiences? I'm aware it's a quite new feature added to the suite, but would love to one day stop paying the pricey subscriptions the other two alternatives demand
I tried and as a individual (single/one at a time) image editor is it fantastic, beats Lightroom by a mile and the node editor gives me the ability to "emulate" panels from CaptureOne (and OFX plugin support means I can just LLM my own if I need).
Library management is where Lightroom (and honestly CaptureOne) still wins, I couldn't find a good way to manage a 100k photo library.
I created one project and tried importing the whole library, but it took more than 8 hours and didn't make meaningful progress (~10%?) so I stopped (compared to 15hr for the whole thing in Lightroom).
I also did a project for about 1K photos and while the import worked, the library management itself was lacking. I could not find ways to do image stacks for bursts, group by metadata, etc. The obvious filters are there, but I think it needs more work.
I spent a little while trying. My knowledge of lightroom didn't really carry over. I couldn't work out how to do things like masks. I'm sure if i spent longer and watched some tutorials I'd get further but I decided to just let the photo mode cook a little longer before I try to switch over.
I make motorcycle travel videos for YouTube, and the bulk of where my time goes in the preparation of a timeline: chronologically cutting the highlights, synchronising dual camera footage. If Claude + Davinci Resolve will enable me to cut this time slightly, even if it's 10%, it's a welcome change. Ever since I made the jump from Premiere Pro, I've been positively surprised with how good Resolve is. Excited to try this out!
I thought DaVinci Resolve could automatically sync camera footage for a few years now. I did it about 4 years ago.
Yes, correct, there’s a feature for that. Unfortunately one of the audio channels from the camera footage on the motorcycle is never good enough for this feature to work for me.
i hope blackmagic pays attention to what steam is doing with steamos.
steam has put so much into making gaming on linux so good, one of the last pieces remaining for a ton of people to make the switch is editing.
don’t get me wrong, i run davinci on my non-rocky linux desktop and it’s almost entirely painless, but, what steam is doing with their distro of linux is incredible.
if they were to consider migrating to steamos rather than rocky, between the masses switching to linux since gaming is good enough for so many, if davinci was mostly solid, i think blackmagic could pull millions away from the adobe hell they’re stuck in on windows.
particularly since davinci’s new photo tab is fucking incredible and will so soon be a perfect replacement for lightroom.
OpenAI's latest "GPT-Image-2.5 Flare and Sunburst" announcement video was edited by Astra in DaVinci Resolve, according to Brent Schooley https://x.com/chefbrent/status/2097405034220736659
As an anecdote, I recently needed to colour-grade some short footage for personal purposes. My workflow at the time was to essentially ask an LLM for guidance on how to use DaVinci Resolve to perform the task, step by step. I would make an adjustment, take a screenshot of the controls and the result, and repeat that process a handful of times, eventually ending up with a pretty nice result.
I can definitely see the utility of something like this for cases where a media budget just wouldn't have existed in the first place.
I used Google's Agent Development Kit to make a super simple agent outside of Resolve that I can use to say "take a 10s clip from file1.mp4 starting at 00:01:25 and a 20s clip from file2.mp4 starting at 00:02:00" then tell it to preview and render. Speeds things up.
Off topic, but can anyone share their experience with DaVinci Resolve on Linux? It seems that they only officially support Rocky.
I’ve been using it in Linux for >2 years, including their special cutting keyboard (blanking on the name).
Works without a hitch, including multi monitor, gpu acceleration etc I use it on Manjaro, installed via AUR.
That said, IIRC it only works with nvidia GPUs and requires X11, and is supposedly finicky to get the environment right for (but the AUR gets it right for me)
I'd be happy to buy if it they add proper linux support. I have an AMD card and I don't feel like fighting with the install process.
I'm not entirely sure I would put that on BMD. AMD support on linux when it comes to actual compute capabilities has always been hit or miss.
Only recently has AMD even become viable for LLM inference on consumer chips and that is the same compute codepaths Resolve would depend on.
So effectively AMD is the issue, AMD on linux is great, AMD on linux for a workstation was kind of hit or miss.
Major Resolve plugin developer here.
The good thing about DaVinci Resolve's Linux system requirements is that these are so outdated that basically every developer links their plugins with static libraries they build themselves; and Resolve itself bundles lots of .so files with the app itself. Also Rocky Linux 8 is a good baseline Linux distribution: simple system libraries interdependence, so trivial to export and bundle those.
In other words, binaries created for RockyLinux8 with portability in mind are easily portable to sane modern Linux distributions.
(That doesn't make developing on Rocky Linux 8 fun at all)
Installing on Mint/Ubuntu is easy with MakeResolveDeb. https://www.danieltufvesson.com/makeresolvedeb
The biggest issues are the supports of video cards (nvidia is the king) and different popular codecs. The problem is partly solved by using paid version, but the gpu/codec support is still not even with Windows, so you have to use ffmpeg to re-encode from/into supported formats.
Runs flawlessly on my strix halo Ubuntu 24 machine after fixing the install process [1]. I have an ffmpeg script that batch converts all my files into the supported codec.
[1] https://github.com/nhma20/strix-halo-resolve-24-04/
Tried to work with ut for 5 months on Ubuntu. Installation was painful. Have to use some tutorial. No AAC audio codec support as well, which is the code of my smartphone camera. Randomly decided to not work after Ubuntu updates. I'm on Windows now...
I don't know about recent versions, but I managed to install 18 on Linux Mint.
Making it run wasn't easy, and it can't decode MP3 and some video codecs because of licensing issues. It also can't decode WebM. This is kind of a problem because of a lot of free music you can download to add to videos and a lot of video clips you can download from Pexels for example are encoded with codecs it can't decode, so you have to re-encode it with FFMPEG first to a supported codec.
Seems like an unusually boring Resolve release. There's nothing there I particularly want to try out. Nothing wrong with that, I've just got used to every release having a ton of shiny new stuff in it.
I guess the agentic stuff is probably exciting to somebody.
It's a .1 release
Can they support something like modern linux? Every single youtuber talks sh/t about davinci and they still do not care
Talking shit about Linux support or the software itself? I know at least 4 YouTubers that do all of their edits in resolve
I am never excited as I am when I see a Davinci Resolve Update
Anyone knows if there os support for Intel integrated graphics in Resolve now?
I tried it a few times but I think it didn't work at all with those cards.
Resolve is extremely RAM hungry. For 4K video I wouldn't want to be editing with less than 12GB of VRAM. I'm guessing that excludes most integrated graphics.
Source - I work as a professional editor (amongst other things). Please commenters, do not come at me with your five minute low bit rate projects that required only a LUT and some contrast and 'worked fine' in DaVinci on a potato.
Integrated graphics use the main system ram so they basically all have access to more memory than the average GPU.
After Effects still has some features that keep me stuck with it, but I'm excited for more competition. Adobe seems to be paying attention somewhat.
They must be planning to be bought by one of these companies. Otherwise I dont see why they would not just continue with their strategy of integrating AI tools within their product. They just reduced DR to just another tool. This is like opening a five star restaurant in a food court at the mall.
Is it still a PITA to install? If I have to install a custom Linux distro to be able to run your app, it's not a good time.
Use MakeResolveDeb to easily install on Ubuntu/Mint.
That's what distrobox is for ;)
It installs on Windows 11 with one or two clicks.
I've tried it a bunch of times, it just doesn't work and always crashes on Artix.
LightWorks works flawlessly on the other hand.
For people like me who have no idea what this is, I looked it up:
DaVinci Resolve is a proprietary application for non-linear video editing, color correction, color grading, visual effects, and audio post-production. It is developed by the Australian company Blackmagic Design for macOS, Windows, iPadOS and Linux.
...
DaVinci Resolve has been used for the color grading and/or editing of feature films such as Alien: Covenant, Avatar, Best of Enemies, Deadpool 2, Jason Bourne, Kingsman: The Golden Circle, La La Land, Love & Mercy, the Pirates of the Caribbean series, Prometheus, Anora, Robin Hood, Spectre, Star Wars: The Last Jedi, and X-Men: Apocalypse.
(Wikipedia)
It’s the de-facto “standard” for color grading and it’s getting more and more capable for things like editing, audio mixing and VFX.
My latest 'free' computer motherboard has an i7-6700K, with integrated graphics. I tried to run Resolve 20 but it complained about my not having a real GPU. I ended up getting an Intel Arc B580 for around $300, which I thought was a better value than getting a used GPU off eBay.
Going to look at this version, as I haven't gone through the tutorials yet, so the assistants would be quite helpful.
This team ships.
AI? Yawn...
apple
I don't see how companies like this can survive going forwards. DaVinci Resolve is effectively ~200 Astra/Fable prompts, maybe even fewer if you're willing to forgo polish.
I'd give you $100K no questions asked if you manage to get something even close to Resolve with 2K Astra/Fable prompts, polish included. You vastly underestimate how much work goes into "polish" when software with huge surface area is decades old.
I think there's a really interesting tension between what you're saying and what the parent post is saying, and it's something that's been discussed for a long time but LLMs are throwing a new wrench into it: a given user probably only uses 5-10% of the features of a piece of software, but across the whole body of users you get close to full coverage.
I use DaVinci Resolve for editing GoPro, Canon DSLR, iPhone, and Drone FPV footage. I import clips, use the razor to cut them and rearrange them, occasionally crop and zoom, and then encode them down to h264 or whatever. There are a bazillion features that I sort of know exist but haven't ever had any need to use. Could I make a tool using an LLM that covers my DaVinci workflow? Almost certainly, and I don't think it would take anywhere near 200 prompts. I've made significantly more complex bespoke data analysis pipelines for drone flight telemetry in like... a dozen prompts. But I have no illusions that what I would make as my "DaVinci Clone" would have anywhere near the functionality of DaVinci, just the tiny piece of that functionality that I use.
I'm pretty sure your editing workflow can be achieved in the Preview app on macOS. If your editing workflow is just basic chopping clips out of a longer video then yes, you could vibe code it.
> a given user probably only uses 5-10% of the features of a piece of software
Then vibe your own 5% editor, but saying that a software so mature and complete as Resolve can be done with some AI and time is something I expect coming from the mouth of a teenager, not an adult.
regardless of whether it’s possible, there is enormous value in the community, ecosystem, and learning materials centred around Reaolve as it is right now. the workflow has been refined over years. no one shot slop effort will have that network effect behind it, and if the workflow isn’t a drop-in replacement it’s dead in the water anyway.
You can say anything, why stop at claiming 200 prompts, why not say you think Astra/Fable can one shot Davinci Resolve! And you can EASILY make the BM cinema cameras and other hardware by prompting Fable to run a 3D printer. (≧▽≦)
The free tier is very generous and it’s only $300, no subscription. Those Fable credits would cost a lot more than that. You’d spend thousands of dollars to save $300.
Moreover, not everyone wants to have the hassle of building and maintaining their own version of a thing that isn’t compatible with anyone else’s version of the same thing. And if you hire anyone to help you, now you have to train them on how to use your app. If they leave, those skills are useless for their next job.
(I’m not even covering the idea that you could do it in 200 prompts. I got Claude to create a video player and that alone took over 300 prompts.)
Black Magic Design's primary business is selling cinematography hardware. Anyways, those ~200 prompts are going to cost you more money than a license once you factor in the cost of your own time. Value proposition of AI still ain't there, and it probably won't be until the hardware bubble pops, and it very well might never pop without market intervention. This is assuming your 200 prompts is a realistic figure, which I'm frankly skeptical of.
you do know this company also sells myriad of hardware , can't Fable those just yet. and I am sure majority of their income comes from that part of their business, selling equipment to studios etc.
LOL. I wish people could make good software that easy!