Thinking About the Future of Video Editing

Posted on by Larry

There are people who make a living predicting the future. I am not one of them. I’m pretty grounded in the here-and-now. My focus is on what we need to know today to complete our projects and pay the bills.

Still, I want to share some thoughts about the future of video editing – as much to read your comments as to reflect my thinking.

COMPETITION & AI

In the past, those of us who earned a living in this industry needed to worry about lower-cost competition stealing our clients and lowering budgets. That pressure hasn’t changed. One of the most common ways to launch a career in media is to do the same work cheaper.

What has changed is the rise of wholly new competition: Artificial Intelligence (AI). I was reading an article in Ars Technica this morning that “[w]ith neural networks now tackling text, speech, pictures, video, and now handwriting, it seems like no corner of human creative output is beyond the reach of generative AI.”

Personally, I find AI both interesting and troubling. Mostly, troubling. As Eric Schmidt, former CEO of Google, said recently on the Tim Ferriss podcast, “One of the real concerns about AI is that it will be very difficult to tell the difference between information and misinformation.”

As exciting as the new capabilities of AI appear to be, I expect that, instead, we will quickly get inundated with deeply deceptive misrepresentations of the truth. (Worse, because AI can easily create seemingly-real video, clients will think they can do without us and simply let “AI” create their projects. They may even be right.)

While I do think there is a bright side to AI, we have a very challenging time ahead of us before that arrives. In the absence of any clear regulation or sense of propriety on the part of developers or creators, there are some struggles ahead of us. I am not looking forward to it.

FRAME SIZES WILL INCREASE

I read somewhere that the human eye sees essentially a 12K image. My guess is that we are ultimately headed for a 12K frame size; though there will be an intermediate stop at 8K. Already, I’m getting notes from producers and editors who are regularly shooting or editing 8K.

Now, the point could be made that we don’t NEED 8K. (As a test, name any actor that wants their face in a closeup with 8K of resolution.) Nor, for that matter, can we perceive 8K in any mobile, computer or living room display; even large-screen theatrical may not have a big enough screen to show an 8K image at full resolution. However, I do think that for many media productions, we are heading for masters at 8K, and production in 12K.

Our computers are fast enough to edit these formats, but our storage needs to get bigger and much faster.

NOTE: 8K ProRes 422 (8192 x 4320 pixels at 30 fps) requires 2.5 GB/second for playback. Thunderbolt 3/4 tops out at 2.8 GB/second. 8K also requires 1.1 TB/hour for storage. Hard disks and servers can easily hold this much data, but they are nowhere near fast enough.

COMPUTERS & STORAGE

Computers are finally fast enough for any video editing we need to do. While visual effects still benefit from more computer power, media editing no longer requires state-of-the-art hardware. We may want it, but we don’t need it.

The challenge is not maximizing our computer core counts, but maximizing our storage. As frame sizes and the amount of media shot for each project increase, the amount of storage, along with need for high-speed storage, become ever more critical. Hard disks can be used for long-term storage, but editing will need to shift to all-SSD storage

NOTE: It would be really helpful if LTO drives were cheaper and connected via Thunderbolt, rather than SAS; but I don’t see any push for that happening.

Thunderbolt 3/4 is fast enough for single camera editing, but we will need a faster protocol and storage for multicam editing as frame sizes reach 8K, then exceed it.

PRIVACY

Privacy will either become regulated or it will disappear entirely. Too much of the software industry is built on a business model of selling user data to the highest bidder.  To shift away from that model will require us paying more for software. Whatever happens to limit privacy, though, will be driven from outside the US, because the largest tech companies (which are based in the US) provide too many obstacles to create an effective regulatory environment here.

MAKING A LIVING IN MEDIA

This, to me, is the biggest challenge. Competition is nothing new, but the rise of AI makes it much more complex. When AI can create images and videos that are “good enough,” where does that leave us?

In the past, success depended upon the relationships we built with our clients and leveraging them to find new clients. We still need to do this, but it is no longer enough. I think the key is that we need to know more than just media production and editing. I tell my students that media should be their minor. They need to major in a subject that they then use media to express.

That’s good advice for us grown-ups, too. If all we know is how to edit, we are a tool easily replaced by something cheaper. If we know a subject – say, farming, or mathematics, or history, or any field of study that needs to explain itself to the world – that knowledge becomes a fundamental asset we and our clients can tap into to provide more informed video projects.

It is critical that we know more than editing to differentiate ourselves from the competition – both human and computer – that we face today. AI can parrot, but only we can create. We need to be more than an editing “tool.” We need to inform that tool with our creativity and our knowledge of the greater world.

I’m interested in your thoughts.


Bookmark the permalink.

23 Responses to Thinking About the Future of Video Editing

Newer Comments →
  1. Hanno Hart says:

    Dear Larry,
    thanks for your thoughs! Multitasking ist the answer. Basicly as DOP for TV and independend arthouse, I picked up editing with Betacam TV and by educational background I evolved in producing teacher-training videos. This combined with corresponding institutions might be a 15 year-perspective to rise children. Might also not.

  2. Simon Morice says:

    During the early days of high definition (HD) display technology demonstrations involving LG, I had the opportunity to attend a lecture by a BBC engineer on the subject of 4K systems at a Royal Television Society event. The speaker discussed the resolution capability of the human eye, which is approximately 1 arcminute or 1/60 of a degree, roughly 30cm at a distance of 1 km.

    He explained that traditional standard definition (SD) television sets could be positioned at a viewing distance of six screen heights from the audience to achieve the maximum resolution. This made them suitable for family viewing, as there was ample space for all viewers to be at the optimal viewing distance. However, as the viewer approached the screen, the resolution of the display would become less than that of their eye.

    In comparison, Full HD displays have roughly twice the linear resolution, allowing maximum detail to be viewed at half the distance, or three screen heights. At 4K, this distance roughly halves again to 1.5 screen heights, and at 8K, it’s 0.75 of a screen height.

    The speaker emphasized that the dynamic range of the pixels was of greater significance to the overall viewing experience compared to their size. He argued that while higher resolutions are important, they are not the only factor to consider in optimizing the viewing experience.

    …I guess there will be implications for volumetric production as well as for AR and VR systems. Since increasing volumes of video are being consumed on phone screens, presentation resolution is almost moot.

    Note: In light of your mention of AI, I thought it might amuse you to know that I ran my initial comment through ChatGPT asking the AI to rewrite for a technical audience, and this is the product.

    • Larry says:

      Simon:

      Thank you and ChatGPT for your comments.

      The fact that the second comment to this post was created by AI is a statement in itself.

      Larry

  3. Jim McQuaid says:

    Thanks for sharing these fundamental insights. Privacy and surveillance capitalism are huge problems that will have to addressed; we who create video are a tiny part of that, alas.

    As for surviving in commerce, I run into young camera guys you are thinking they will go into business doing various kinds of commercial video. I tell them, fine, but understand that only 20% of that work will be about how good you are with a camera or editing. 80% is about understanding the customer, their goals, reality-checking them in a useful way and dealing with changes they discover they want to make. That’s the heart of being in business by yourself.

    And — if you happen to enjoy a good mystery novel — my friend Sarah Johnson has written 3 mysteries set in New Zealand (Molten Mud Murder, The Bones Remember, and The Bone Track). Enjoy your trip.

  4. Technology has always been the great disruptor. We experienced that with Claude E. Shannon when he changed the digital playing field with his information theory that propagated down to bits and bytes or how the Amiga Toaster leveled the playing field of NLE systems at the consumer level. Bottom line is that will need to adjust/upgrade our skill sets to this new A.I. business model if you want to stay employed! It’s a scary time for editors but the future has always pushed creative people into unknown areas – often forcing them to wear many other tech hats. Good or bad, we are going to find out soon if we can adjust and sustain in the marketplace. This might be a good time to have a backup plan!

  5. H. Nelson says:

    Just a couple random thoughts: I have a feeling that as AI becomes full-blown, we’ll have to incorporate some version of it into our filmmaking toolkits. I see this already beginning in the art world. AI such as MidJourney is wonderful, but it’s only those artists that have mastered HOW to use its prompts that are creating results that go beyond “Hey, that’s cool” to generating something that is commercially usable/sellable.

    For now, AI still needs a user to give it input: a mission or purpose. And someone needs to sift through the end results to see what is usable or desirable. AI can create 24/7, yes, but if it creates something no one is interested in–what’s the point?

  6. Christian says:

    Thank you Larry for you thoughts. I am quite sure that in the next 2-3 years several companies will look for average quantity/quality productions with AI. I fear that the quantity of experienced media workers is becoming bigger than the market request. The big question is for the guys who are not old or young enough. When you are around 40 y.o. with a great experience, but still with 25-30 years of work in front of you. Should we all become teachers at some point or should we look for a plan B?

    • Larry says:

      Christian:

      I suspect you are right. Opportunities for full time work are diminishing. But part time work still seems to be growing.

      And having a Plan B in our industry is always wise. I’m just not sure what that is quite yet.

      Larry

  7. We are “programmed” to think of AI as a replacement for us and therefore threatening. I’m not saying there’s no truth to that, but AI can also be really helpful in some aspects of our work. Any editor who isn’t familiar with products such as Topaz Labs’ image enhancement technologies (https://www.topazlabs.com) is out of date.

    • Larry says:

      Jacquie:

      Thanks for your thoughts. I’ve played with Topaz Labs tool but haven’t seen any significant results. Probably operator eeeor.

      Larry

  8. AI is no doubt a tool that will be useful to some and others not so much. Before I retired to work on my own projects, a segment of my customer base was known as difficult to work with and that became a niche. I became very good at helping those customers succeed. I wish AI all the best working with that group, you’re going to need all the algorithms you got.

  9. Emerson says:

    Regarding frame sizes, I think it is a matter of marketing. We all lead to know that (it doesn’t matter which equipment) the more the merrier, which means that the bigger the frame size the better the quality and this is because the technology has no new place to go since we reached the “best” quality image since the beginning of video editing. So how can I sell the “newest” camera model? let’s increase the frame size and add gazillions of new pixels (even though we don’t need them). How can I sell the new best computer and/or ask you to update your (more than enough) “old” machine? let’s sell a computer that handles that frame size and those gazillions of pixels…so from the client’s and consumer’s perspective technicalities don’t matter: they want the “best” and the best is always more (in their mind). So unfortunately we just need to ride along until we start using a movie theatre as our monitor to edit. About the AI I don’t think we need to suffer in advance, as you said AI is just parrots so they need somebody to copy so why can’t we be that person? we could incorporate AI tools into our editing, as we use templates and plugins to save us time, we could use AI tools as well…Just some positive thinking…

    • Larry says:

      Emerson:

      Positive thinking is always good.

      And I’ve already starting chatting with my wife about converting the spare barn out back into a movie theater so I can properly monitor 8K – dare I even suggest … 12K – media properly.

      Larry

  10. Dirk Baumann says:

    I remember when I took the first semester of computer science in 1975. “Computers will never replace jobs!”
    I do not remember hearing anything about Al.

    Where is the future, is Al it?

    Dirk

    • Larry Jordan says:

      Dirk:

      Technology and computers have always had an impact on our lives and jobs. That pressure continues. What makes AI different is its illusion of creativity. As creative artists ourselves, we need to consider how to help our clients distinguish the benefits of human vs. machine creativity. For many tasks, machines will be sufficient – which would have an impact on jobs.

      Larry

Newer Comments →

Leave a Reply

Your email address will not be published. Required fields are marked *

Larry Recommends:

FCPX Complete

NEW & Updated!

Edit smarter with Larry’s latest training, all available in our store.

Access over 1,900 on-demand video editing courses. Become a member of our Video Training Library today!

JOIN NOW

Subscribe to Larry's FREE weekly newsletter and save 10%
on your first purchase.