In the second of his articles for FCP.co, ex FCPX Product Manager Steve Bayes gives us his thoughts on the recent IBC trade show in Amsterdam. As he led the team who launched the fastest NLE, he also gives us his suggestions on the best way to benefit from the extra editing speed from new hardware.
The first time I was in Amsterdam I decided to fake my own death. It was a simple plan, involving life insurance being split with Wife #1 and staying in Amsterdam to become Dutch. Of course, I would need a little help from my friends with the whole “Drowning in the Canal while Incredibly Drunk” story, but these seemed like minor issues in the larger scheme of fraud and learning Dutch. Common sense prevailed, however, and I went home as per the original plan.
This year was a little different, but there was an important sameness to the IBC post production presentations. The common subtext was, “We are just like all the other systems so choose us!”. This was almost an admission of defeat — after years of enticing users with shiny new features to get them to change editing systems (or not change), the major NLEs just wanted to make it slightly easier to do the same thing. No one really pays for upgrades these days because of free and subscription, so the business case for adding new functionality gets a little strange. And I’m not sure that “we’re all the same” is a winning strategy to motivate change.
No, this year was really about speed. BMD Resolve continued to promote the Cut Page as if it was a stripped down FCP X inserted into the complexity of a legacy color grading system.
Resolve has been able to use multiple GPU’s for many years and the big facilities always liked to differentiate themselves with as many GPUs as they could — Resolve’s current configuration guide shows up to 8 GPU’s. It will be interesting to see the performance improvements with the new Mac Pro, the potential for 4 GPUs, 28 CPUs and the Afterburner card.
At some point you just can’t get the frames off the disk any faster for more GPUs to be helpful in processing streaming video. You need to balance this processing with faster drive arrays too. Adding lots of power windows, tracking and other processor heavy functions on a single large frame sized shot should be nicely accelerated by adding more GPUs.
Theoretically, the BMD dedicated Editor keyboard also speeds up editing with the jog/shuttle wheel and encourages more editors to use keyboard editing techniques. Just custom mapping more edit functions to any keyboard will help with that. But when the user needs the mouse, the dominant hand takes over and the keyboard gets pushed aside. And the user always needs the mouse at some point in a complex interface.
Adobe’s demo stage announced that you could now “Get to the Finish Line Faster”. This improvement was caused by motion graphics templates, better integration with over 300 3rd party partners — and, presumably, being more stable.
Seriously, an important new feature in the last release is the System Compatibility Report as Adobe tacitly admits that much of the system’s instability is a result of your computer’s dodgy configuration.
This is just inevitable when users have such a wide range of choices to configure their own computers and the QA testing matrix must be both massive and somewhat futile.
The specific Premiere performance improvements are for HEVC and H.264 decoding, but no stats were presented. Decoding is the first step to faster rendering, exporting, transcoding and multi-stream playback. Adobe has supported multiple GPU’s for some time now, but it would be good to have baseline performance stats to judge the level of improvements.
The handoff between GPU, CPU, RAM and caching to SSD can be mysterious to the casual observer and an app has to be able to segment files across these multiple processes to get improvements. RED RAW has been optimized for Apple’s Metal (a replacement for Open GL to improve GPU performance).
Avid continues to market their new Media Composer UI update and collaboration/shared media library, Media Central. Their advertised speed improvements are due to the faster access to media on shared storage and networks. I won’t comment as to whether the Media Composer UI update hits the goal of attracting new users without pissing off the old ones. Or, if it directly leads to speed improvements because of further customization. They certainly needed an update so I’m optimistic.
Apple, as per usual, was not publicly exhibiting at the show. After seeing several other IBC demos go horribly wrong and watching users walk up to demos pods only to ask if “their bug” was fixed yet, I am still sympathetic. It is impossible to demo any randomly requested feature with the video material curated for the new features. Inevitably, those kinds of impromptu demos can backfire spectacularly and have the absolute opposite effect.
FCP X 10.4.7 was released post-IBC and the new tag line is “Work at a Faster Clip”. Also per usual, there are specific stats published on the website for performance improvements using Metal. Showing Metal’s potential is important to convince other developers to update their apps and FCP X is an important example of why you should do so.
The majority of this update seems to be getting ready for the Mac Pro and XDR monitor(s), but there are incremental improvements in color grading, simplifying HDR monitoring and, my new favorite, Sidecar support. I am definitely connecting my old 12.9” iPad Pro to my iMac Pro and look forward to my other favorite apps better utilizing this new Catalina feature.
There are several excellent FCP X videos from Mark Spencer at Ripple Training here.
A final word about speed and the perception of speed. A few years ago I watched an incredibly talented editor working on FCP X in front of a small group of agency creatives. The room was configured so the creatives could sit on the black leather couch and watch a large video monitor on the wall. The editor was behind them and only he could see the UI monitor. They were cutting a high end corporate video for a live presentation.
This editor had figured out how to edit in FCP X without stopping playback.
I’ll say that again — through an incredible knowledge of the FCP X keyboard, he could make changes in the timeline while the video playhead never stopped moving. So the creatives watched the playback of the video and shouted out their changes to the editor. As soon as the piece stopped playing, they started to get up to do something else while the editor made the changes. Instead, the editor just pressed Home and jumped back to the head of the timeline. He said, “Wait a minute. Watch this.” and started to play back with all the changes already integrated.
The creatives were a little surprised. They stopped what they were doing and commented, “Well, that was quick.” then went back to shouting out more changes. Not only did they not realize what had happened, they weren’t even that impressed. An incredible combination of skill and technology was greeted by the client with a proverbial shrug.
That is a rather long winded way of saying that “speed” is a mixed blessing and difficult to define. Of course, faster turnaround will always be a benefit for markets like news or events, but for the wider world of content it is a bit more nuanced.
Imagine if you were ready to edit a project that normally took 2 weeks, but now the facility is using FCP X with the new Mac Pro and it will only take 1 week. Incredible! Except for you. Will you double your rate to make up for the shorter timeframe? No. Will you get twice as many projects in the same time? No. Will you pad it out by making lots of excuses about how FCP X can’t do certain things? The last guy didn’t. The next guy won’t.
My advice: use speed to go deep. I know it is hard to demo depth, but if I hear one more marketer say “faster playback” I’m going to take a sip from the aquavit in my hip flask. What they mean to say is that you don’t need to pause or work at low res to instantly see your latest changes — whether this is shouted at you from the leather couch or the ones that secretly popped into your head as you watched the last run through. The dissolve at that rough transition should really be just a few frames longer, the audio needs to lead the video a touch earlier and with a softer volume ramp. You need to jump back and make the changes before clients get restless.
It is also not about how many 8K Picture in Picture streams you can play at once — it is about accelerated creativity and executing changes after review. Some clients will want to keep working with you just because everything seems to go more smoothly and the result is somehow better. Or you may have to point out exactly what you did and why it is better — know your client and work the room.
You will need to point out the benefits of going a bit slower to get the best results because watching, thinking and discussing all go at about the same speed as they always have and should not be rushed. Or just allow yourself to take the extra time for your own project for your increased satisfaction and growth as an artist.
Use the extra speed to experiment more — do more versions of the color grade with Auditions, use the benefits of Roles, add custom metadata for advanced Smart Collections, try some subtle speed ramping and slow frame repos. And, for God’s sake, use the extra time to do closed captioning. Use speed as an opportunity to take your game to a higher level, not to finish sooner. It's the best long-term strategy to stand out in a competitive field. Now where’s that hip flask?
My final takeaway from IBC is that, informally, I am hearing about large investment in local European production from all the streaming services. As the large film studios launch their own services and take back their film catalogs for themselves (RIP FilmStruck), the only way to add subscribers in each individual European country is to do more local language series.
This will be a catalyst for developing new techniques to generate high quality content on tighter budgets.
The methods created for Hollywood TV series will need to be streamlined and that means production and post production teams will need to experiment more. The force for change will be too much work, not the opposite. This favors the risk takers and, frankly, I find that refreshing and hope to see more in the coming year.
I also very much enjoyed my 2 most anticipated events - the LumaForge canal boat trip and the Atomos party at the Supper Club. It was great to be in a glass enclosed boat with a group of friends and colleagues so that you were really focussed and yet still had a beautiful, ever changing view.
We closed the Supper Club down and it was good to hang with my old Apple friends. And drinking beers sitting at Leidseplein until after midnight always reminds me how the Europeans have the whole "third space” down so much better than we do in the US.
And as the Blue Angels practice overhead in anticipation of Fleet Week, that’s it for this month. I’ll be slowly pushing out photos and short videos to IG @cuttogether from the 2 weeks in France.
Steve Bayes was the first certified instructor and principal product designer for the Avid Media Composer for almost 10 years and senior product manager for Final Cut Pro for 13 years. You can follow him on Twitter or take a look at his excellent photography on Instagram or his website www.thestevebayes.com.