Final Cut Studio 3 and Mac OS X Lion - Possible but Not Practical

Apple aren't known for their sympathy towards legacy products. When they introduced the first iMac back in 1998, many customers were dismayed to see that they'd neglected to include the option for a floppy disc drive. Years later, and the removal of packaged Apple software from its stores reinforces Apple's stance on what it sees as dated tech; it's over - move on.

Apple have been just as ruthless with their programming as they have with hardware. With the introduction of Mac OS X Lion, Rosetta (a dynamic binary translator that allows you to run applications programmed for PowerPC system architectures) is no more.

When I first heard about this, I was pretty concerned. I've been using Macs now for eight years now, and I've accumulated quite a number of PowerPC applications over that time. What's more concerning from a professional stance, is that Final Cut Studio, a software suite that I've spent many years learning and making a living from, will no longer function under the new Mac OS X environment.

But all is not lost! If, like many people, you have a Snow Leopard install disc, you use it to install Rosetta on Lion, and breathe life back into your old applications.

So you'd assume that that was a problem solved. Chalk one up for old-skool Mac users worldwide. Well, not exactly.

In my current job, I've been trying out how Mac OS X Lion plays with Final Cut Studio 3, and its not always pretty. It all comes down to Lion's apparent improvements in dividing processes between multi-core processors, which solves some problems, but introduces others with general use.

Final Cut Studio was not built with multi-core system architecture in mind. For that reason, you'll find that it's rather infuriatingly inefficient at making the most of your system resources. Slam something into Compressor, and watch in Activity Monitor as your 16-core Intel Mac Pro barely bats an eye-lid at your request. It is possible in this scenario to use the virtual clusters trick to force your Mac to work harder, but this is a bit of a dark art, and can render your entire system unusable if set-up incorrectly.

Lion, however, eats these kind of complicated calculations for breakfast, taking the guess-work out of all of this, carefully allocating processor time to all of the tasks presented to it. Sounds great on paper, but then when you want to edit in Final Cut and transcode in Compressor at the same time, Lion doesn't know which tasks deserves more attention. What you're left with therefore, is a slower, less-responsive and asthmatic editing experience, where every render takes considerably longer.

The problems don't stop there, either. In my experience, Final Cut Pro acts erratically even when you haven't got anything else going on in the background. Recently I was editing together a five-minute piece consisting of nothing but ProRes Proxy QuickTimes and a couple of AIFFs on a ProRes timeline, and consistently found that Final Cut would drop frames on two occasions during playback (anything from about 5-50 frames) before I paused and resumed playback again. All this on a brand-new 17" MacBook Pro with a 2.4GHz quad-core Intel Core i7, 1GB of AMD graphics memory and 4GB of RAM - specs that dwarf the listed minimum system requirements of Final Cut Studio.

My advice? Don't upgrade at this time if you're still hanging onto Final Cut Studio 3 and below. Save the £21.99 that you would be spending on Lion and put it towards something useful, like some FCP X, Avid or Adobe Premiere tutorials. Just remember that if you choose to stick it out with Snow Leopard for the long haul, then it won't be long before Apple leave you and your aging system behind to fend for yourself.

Bristol Old Vic - Good Clown Bad Clown

I've never really fancied myself as much of a director. That said, a couple of weeks ago I popped down to the Bristol Old Vic on a Saturday to direct an online trailer for Good Clown Bad Clown, a childrens' performance for the winter season. Shot on a Panasonic HDX-900, and then edited by myself in Final Cut Studio with some basic work in Motion, it's 63 seconds of white-faced, red-nosed, horn-sounding fun.

Digital Workflows - The Dream and The Nightmare

A long while ago, I wrote an article on here about the advantages of digital workflows, spouting on about the wondrous new range of Sony XDCAM cameras. Of course, at that time, I hadn't really tried out any other digital products on the market, and it was a bit naive to get swept up into the Sony hype that they would revolutionise the digital video industry.

Digital formats are the future, but they're not as simple to use as I initially assumed, and it all comes down to the multitude of compression formats that different brands and types of cameras use. When you're out and about, you want a camera to record and the highest possible quality, whilst taking up the smallest amount of space on your memory cards. This isn't so important when you're in the comfort of a studio, but it is when you're half way around the world, in the middle of a humid jungle in Indonesia.

In just such a situation, you're likely going to end up recycling your cards on a daily basis, and whilst all your data can be backed up with the use of just a laptop, you have to consider the transfer speeds of your media, because these are likely to be a real bottleneck in your workflow. I recently had a plea for help from one of my friends and colleagues, who was on a remote shoot in a foreign country, having to back up sometimes hours of Canon 5D footage from Compact Flash cards at the end of each day. Capturing one card through a USB 2.0 memory card reader would take hours, then he'd have to get up in the middle of the night in order to switch to the next card... and then the next. In order to get around this, I had a Lexar ExpressCard/34 memory card reader sent out to him, which saved a lot of time (and sleep) at his end.

With tape-based media, you can only capture in real-time. Therefore, if you've got 24 hours of footage, you'll need to spend 24 hours playing it back on a deck for a computer to capture all of your rushes (not allowing for all the fussing with swapping cassettes over and inevitable technical hurdles that you're almost bound to encounter). With file-based formats, manufacturers would have you believe that the footage is ready to be sent directly to the edit, but that just isn't the case. If you haven't already (because you live your life on the edge - rock on), you need to backup all that material that you shot. Lets assume that you shot 35GB of footage every day for fourteen days. On a Canon XF305, that's 87mins of footage at full-quality 50mbps and a grand total of 500GB of footage. Based upon LaCie's claimed transfer speeds of 57MB/sec over FireWire 800, it would take you about two hours and twenty minutes to transfer that data from one of their Rugged hard-disks. But then you've STILL got to transcode all your footage into something that's suitable to edit with.

Sony's XDCAM Transfer footage will quite happily mow through all your data and wrap your raw files into playable XDCAM QuickTimes, which are supported by Final Cut Studio, but not ideal. XDCAM is, as I've learnt, a bit of a vile with support that's been shoe-horned into Final Cut. It's fine on it's own, but if you start mixing formats on a timeline, FCP is likely to die on you repeatedly. You'll therefore need to compress the footage into something that's a lot more palatable to Final Cut, such as Apple ProRes 422. Whilst I don't have any official figures to hand, let me assure you that this takes a loooooong time and a lot of drive space. One way to get past this is to encode everything into ProRes Proxies, but this means that you're going to need to conform all of your footage to ProRes 422 or ProRes 422 HQ before you go into an online edit... so that's an extra step in your workflow that I'd generally allow about half a day of work for. If you want to avoid all the hassle of a conform, consider spending considerably more on more hard-disks.

There are ways to use distributed processing to force the very best out of Apple's Compressor in order to transcode footage faster, but my experience of the XDCAM format is that no-matter what you try, Final Cut really doesn't like the stuff. I've had better success with Canon's XF305, which comes with lightweight software and a Final Cut plugin that allows you to use the Log and Transfer window in Final Cut Pro. (Sony state that you can use the Log and Transfer window in Final Cut, but I've yet to see it actually accept any XDCAM footage. Sony's software and plugins are, quite simply, a bit of a mess in my experience).

If you're shooting with a camera that records into H.264 QuickTimes, for instance with the Canon 5D MkII, you're into a whole world of pain because of the heavy compression of the files. Final Cut Pro will spit at you if you try to drop the files directly onto the timeline, and Compressor will become particularly unresponsive if it even so much as smells a batch of more than about ten clips. It's therefore better to use another program like the horrendously useful MPEG Streamclip to batch convert your files.

Just remember to consider how long this is all taking to convert your files. If you do have a catastrophic hard-disk failure on your working drive, you'd best make sure that you have a backup of all your converted files and not just your raw files, otherwise you're going to have to do aaaaaaaaaaaaaall this again and find a remedy to the nervous-breakdown you'll be having.

Once all your conversions are complete, however, you can finally breath a sigh of relief - you're finally ready to start editing. Wasn't that quicker and easier than sticking a few tapes in? Well... maybe not actually.

Development Films

For the past few weeks, most of my time has been dedicated to twelve training programmes for Development Films. It's been a long while since I filmed or edited something based on a narrative script (the vast majority of projects that I undertake at Blue Sky are shot and assembled using the same documentary methods that a news and current affairs programme might use), and it's made a welcome change.

What makes this project particularly special to me is that I not only performed the role of second camera and editor (shooting on a Sony PMW-EX3 and editing on a Mac Pro with Final Cut Studio 2, but I also took on quite a bit of producing work, finding suitable locations and cast according to the guidelines set by the client.

The video presented below is currently being shown at exhibitions across the country (Adrian, Kay and Richard will be at the NEC in London tomorrow).



The first series of Development Films is due to go on sale this Autumn, available as a DVD or digital download from The Development Company.

Somewhere Beyond the Sea...

Uncannily, it's a double bill of yet MORE watery-themed projects to talk about today.

I'll kick things off by first introducing this clip from Team Insight's video, charting their run up to their epic rowing voyage across the Atlantic Ocean. That's right - rowing 2935 miles, from La Gomera in the Canary Islands, to Antigua! And all in aid of charity!



I have to say, this has been one of my favourite projects to edit together, partly because it provided me with an opportunity to experiment with a few fancy effects, such as compositing and a bit of Apple Motion, but also because the guys had a really good sense of humour - something that I capitalised upon.

Their eventful trip was also blessed with some pretty decent footage. Three cameras were on-board the boat throughout the journey, two of which were fully waterproof and able to capture the action below the waves, as well as above.

You can find out more about Team Insight and their incredible feat at their website.

From one favourite, to the next. Students at Churchdown School recently participated in a project that melded together dance and illustrative art into a theatrical performance. Pupils were encouraged to explore new forms of expression in response to natural and man-made architectural forms in and around Gloucester. Gloucester being a city that grew up around docks, this mostly involves a lot of water.

When I was actually filming the performance, I had this underlying fear that I would never be able to recreate the dance in a flattering light. Stage performances do not always transfer themselves seamlessly onto film - the camera almost bottlenecks the performance and (for lack of better words) seems to suck the soul out of it, so this was my attempt to compensate for that loss. Rather than merely documenting the performance, it's an extension of it. Cue a load more compositing, filters, masks and colour-correction.

Enough talk and philosophising! Watch and enjoy!

Canon Eos 5D MkII Video Test

Recently, my boss at Blue Sky Event & Media bought a brand-spanking new Canon 5D. You might know the 5D as one of the most exciting DSLRs to hit the market for quite some time, sending videographers and directors weak at the knees over its mighty fine HD video recording and jaw-droppingly shallow depth of field shots (which appears to be a very 'in' style of shooting at the moment judging by the new series of Doctor Who and Top Gear). Producing 'cinema quality' images for a fraction of the price of your typical, broadcast quality video-camcorder, it has contributed heavily to a flurry of HD art-house films hitting the likes of Vimeo. The Canon 5D produces remarkably beautiful images. All that has been added to the raw footage has been a basic vignette and colour grading. The whole film probably took only about half a day to shoot, edit and export.

Even so, a few problems reared their ugly head whilst I was making this film. The first being that the 5D is not ergonomically suitable in any way to shoot video. Adaptive autofocus is absent when the camera is in Live-View mode and focusing a HD image, with a low depth-of-field, on a screen of about 1.5" is a hit and miss process. If you want to expand that cinematic depth-of-field, the camera will adapt by raising the ISO, so you'll likely end up with a grainier image. Oh, and forget about any decent sound from the tiny internal microphone - it's not intended for any serious use and an external microphone or separate sound kit is a must have if you are considering making films on it.

The second dose of problems becomes apparent when you start to edit. The H.264 encoded files that the camera records to are not a suitable format to edit with on most systems (believe me, I tried), so everything must be converted to Apple ProRes 422 before Final Cut can touch it. This sensibly requires Compressor, which knocks my dual-core MacBook Pro with Final Cut Express out of the picture fairly early. Converting to ProRes takes up a lot of time and storage, certainly in comparison to capturing DV from tape, so if you've recorded a lot you'll either want to be selective about what you convert or leave your machine chugging through it overnight.

Editing with such large files requires some expansive system resources, so the quad-core Intel Mac Pro that I was editing with suddenly presented itself as a bit old-hat, Final Cut proclaiming that certain filters and transitions weren't possible to render on such a weedy graphics card. Those renders that can be performed have processing times that are roughly doubled (or even tripled) when compared so working on a standard-definition timeline.

If you're experiencing similar problems of Final Cut complaining about your graphics card, you might like to know that graphics card upgrades for the Mac Pro are not cheap or easy to come by if Apple no longer stocks the part. Few graphics cards have been produced that support the chipsets in Apple machines. Of course, if you're getting these problems on any other Apple machine, such as an iMac, MacBook or MacBook Pro then you might be even more irritated to be learn that only Mac Pros have upgradable graphics circuitry (assuming that you didn't know that already).

Once you've completed wrestling together your masterpiece, you then have to export it as a whopping Quicktime Movie. These 43 seconds alone spanned across around 850MB of hard-disk space. The irony of it all is that you're probably going to do what I did and export the thing to a video sharing website such as Vimeo, which means that you'll be wanting to convert it back to its original H.264 format.

Work-flows will undoubtedly improve as system specs catch up with HD requirements, but for the moment this is a very time consuming process to produce what might be very little.