This was not one of the smartest things I have attempted to do in Final Cut. The correct tool was probably Motion, but I’m not that familiar with it beyond using a few behaviours on text and so on, and wouldn’t know where to start with a bigger project like this. I didn’t have that long to do this, so even though I knew that Final Cut isn’t the best for dealing with stills, nested sequences and so on, I went for the devil I knew.
We wanted some sort of trailer for our High Society exhibition which explores “mind altering drugs in history and culture”. It’s useful to have this resource to give visitors to our website a sense of what the exhibition will be like, market via YouTube and also apparently increase the likelihood of the exhibition being mentioned on blogs, since there is additional content for the blogs to tart the post up with.
The challenge we have is that before the exhibition there are no available shots of the exhibition itself because it isn’t installed yet. Usually, all we have are stills of the objects and imagery and occasionally some video. The first hurdle is therefore trying to think of something interesting to do with the stills. Classic rostrum moves (zooms and moves across the picture) just felt too dull for this subject, which seemed to be crying out for a more psychedlic treatment. I decided to take the images and cut out figures (e.g. mushrooms, people) from within them using Photoshop (thankyou hacky magic lasso tool) and then expand or move them around against the background, and generally mess with them in way that I hoped would look really trippy.
This process was a bit fiddlier than I’d hoped, but that wasn’t the main problem. What nearly killed this project was the sheer weight of it: the render time. Pulling in photoshop files with all their layers creates a sequence. Putting this into another sequence creates nested sequences which Final Cut doesn’t appear to be friendly about. The files themselves are pretty heavy, and would require a bit of rendering even before I start moving everything about within them. I was also rotating or zooming on separate image files as backgrounds, using luma keys, dissolves between multiple layers and, to cap it all off, motion templates with behaviours for the text over the top. Crunch.
This meant that I couldn’t see a change I’d made in the sequence without rendering it, which was a very slow trial and error method. Also, the rendering process started to throw up “out of memory” errors as my supposedly speedy mac stopped coping. When I came to render the final project, the only way I could complete it was by rendering just a few percent, stopping, saving, and then rendering just a few percent etc etc. It took AGES. I would dearly love to hear from anyone who knows what, if anything, I could have done to make this less painful!
Well, I got there in the end and the trailer has indeed proved very useful, racking up thousands of hits on YouTube after being embedded on various blogs across the world, even reaching the Huffington Post (if wrongly credited).
The music was from the usual place. We abandoned an original idea to also have a voice reading the dedication from Mordecai Cooke’s “The seven sisters of sleep”, since it conflicted unhappily with the music.
So, given the problems with my approach, despite a successful result, what to do next time? I’m thinking maybe something a little more lateral, taking to the streets with a relevant question and gathering interesting vox pops, maybe? Using interviews with curators/artist? Focussing on one artwork or object and using animation to do something interesting with it? We’ll see. Work on the next one begins in just a couple of months…
4 thoughts on “Finding the limits of FCP: the High Society trailer”
The classic problem with Final Cut and stills is resolution: most filters require the stills size to be below 4000px, and you rapidly run out of stills cache. You can up the RAM allocation for stills in one of the preferences dialogues, which helps, and having a kick-arse video card apparently helps too.
So, scaling images down to about 2000px on their long axis helps a lot (even smaller if you don’t need an HD project). Also, exporting the individuals layers as TIFF or PNG with transparency and dealing with them as separate objects in Final Cut seems to work better for me – I tend to use nested sequences rather sparingly, though that could just be superstition on my part.
However, eventually Final Cut just keels over with this sort of thing. As you say, the real solutions lie with Motion or After Effects, which are designed specifically to do this sort of compositing. Which is why I keep meaning to schedule a few days to futz around with them…
Nice film, though. Congrats for battling through it.
Thanks, I think generally I would do better to spend more time preparing various files for FCP. Have another project on the go which involves lots of different formats which I couldn’t be bothered to convert first, seeing as FCP is *meant* to be able to handle them on one timeline. Editing was just about OK, but when it came to final render it all fell apart and I had to render in bits again. Except this time the film was more like 8 mins instead of 1 so took 2 hours to babysit the render.
I hope I’ve learnt my lesson.
Oh, and will definitely try upping the RAM allocation for stills, thanks for the tip!
Yeah, I tend to convert everything to ProRes LT before bringing it into the project. I don’t think it saves time overall, but it does mean you’re less likely to come across a problem late in the process, which – if you’re anything like me – is when you’re least able to cope with surprises.
Also, avoiding lengthy renders is yet another good reason for making three-minute rather than 12-minute films. 🙂