Simon Ubsdell
6
COW Coins
Forum Replies Created
-
[Joe Marler] “Except for these cases, FCPX seems to import extremely quickly. It generally doesn’t strike me as slow to import when using “leave files in place” on a locally-attached RAID array.”
But compared to doing the exact same thing in FCP 7, it was slow in this particular case. 7 was virtually instant – X has to think about waveforms and other nonsense, that takes up time I’d rather be using editing. Premiere Pro is not necessarily any better.
Notably in this case, we ended up cutting the H.264 material directly (no transcoding) in FCP 7 – which theoretically should have been a disaster. But it wasn’t. But you have to know what you’re doing.
Simon Ubsdell
tokyo-uk.com -
[Joe Marler] “I tested a 20 hr H264 1080p timeline on FCP X and it ran but was very sluggish. It is definitely not optimized for that.”
Yes, FCP X is dog slow when it comes to handling lots of material like this.
Which is why I moved over to FCP 7 where all was as smooth as butter.
[Joe Marler] “By contrast when I imported 200GB (20 hr) of that material into FCP X with “leave files in place” this took 5 min”
Yes, those slow import times are why i moved this project over to FCP 7 where the import was virtually instant.
Thanks for your considered answer to the question, but I’m not sure you have done anything to convince me that there is a quicker or more effective way of doing this than the string-out method. I think I have addressed the differences both qualitatively and quantitatively in my post to Bill here:
https://forums.creativecow.net/readpost/335/85854
Simon Ubsdell
tokyo-uk.com -
[Oliver Peters] “it still doesn’t provide the human-ordered visual juxtaposition that re-arranging frames in a bin – or images in a lightbox, for that matter – gives you”
Quite so. I don’t use this method myself but I know of very many editors who do, and clearly it has value. Ideally, every NLE should be able to offer it.
Simon Ubsdell
tokyo-uk.com -
[Oliver Peters] “What do you mean “tag from Space A”?”
He means that non-alpha-numeric tags will float higher than alpha-numeric ones. Just as numeric tags float higher than alphabetic ones.
In FCP X, just like anywhere else, of course.
EDIT: But he clearly meant to say something like UnderscoreA, as FCP X simply ignores an actual space when initialising keywords, although they can be subsequently edited with a space. Why you’d use a space as against any of the many other non-alphanumeric options available I really don’t know. Especially since you can initialise these from the off. Not to mention how hard it is to distinguish a space from the lack of a space visually. Of course, editors have been using this method to sort items for years (computing skills 101) and there is nothing magic about its availability in FCP X.
Simon Ubsdell
tokyo-uk.com -
Hi Jeremy
I totally agree with all of this.
The only point I was trying to make was that having a variety of strategies – which is exactly what you are talking about – is a much healthier editing philosophy, than rigid adherence to one principle or other.
Sometimes, as in the case I cited, string-outs are better and sometimes they’re less useful, or indeed not useful at all. There are certainly instances where I don’t use them, though given the kind of editing I mostly do, they constitute an invaluable and often essential working method.
FCP X does genuinely open up different ways of working – sometimes they are an obvious improvement, but sometimes the old school methods are more appropriate. It’s a shame that FCP X is less well suited to the string-out method (and pancaking, and other associated string-out workflows) than other NLEs. I think it’s only fair to recognise,as you do, that this is a limitation, albeit not a fatal one.
The really important thing to remember is that not all editing is the same. What works for a particular type of editing project, might not work, or might work less well for another type of project. I frequently make a judgement call about which NLE I’m going to use based on the nature of the project, just as much as any personal preference.
On the other hand, it’s hard to argue with personal preference and the way an NLE “feels” to you. Does it free you up or bog you down? It’s a very valid criterion, as Walter alludes to in another post here.
Simon Ubsdell
tokyo-uk.com -
[Walter Soyka] “The emotional penalty for friction in a creative application is certainly high.”
That’s a really great line!!!
[Walter Soyka] “But I wonder if disruption to flow [link] may be a very real speed-killer.”
A really good point, yes. Flow is certainly hard to put quantifiable values on, but has to be a huge component of the editing experience.
Almost all applications that I know of, however good, have noticeable speed bumps at certain points which seem to break the flow.
Which things will be speed bumps will depend a lot on really subjective factors though. You might glide smoothly over my speed bumps and vice versa.
Simon Ubsdell
tokyo-uk.com -
[Bill Davis] “You sit down with a cup of coffee. Load your first clip. IMMEDIATELY put an IN point on the FIRST FRAME.
Then – depending on the “density” of your footage, you can either JKL or use the skimmer to browse ahead. At the moment you see something potentially useful, you just stop. Hit OUT and keystroke the REJECT tag onto everything useless you just sped/scrubbed over. Tap I to set a new IN. Watch the content. When you reach the END of whatever you decide is potentially MORE interesting – hit OUT and make a binary decision. If you LOVE what you’ve just saw – Hit F to Favorite it. If you’re Unsure – Just hit I again to re-set your range select in point. And plunge ahead to REJECT more footage. “
Because I’ve got a bit of time on my hands, I thought it would be useful to count keystrokes – and fallacies. Let’s start with:
IMMEDIATELY put an IN point on the FIRST FRAME.
You don’t seem to know how FCP X works here. You absolutely don’t need to set an IN on the first frame – you can simply set an OUT when you arrive at the end of the section you want to reject. Try to get out of this habit of yours and you will certainly save a lot of time.
But let’s keep that in there for now so we can count the stages in your method.
(1) Load Clip; (2) Set IN; (3) Skim or JKL; (4) Set OUT; (5) Hit Reject; (6) Skim or JKL (you missed this); (7) Set an IN; (8) Watch the content; (9) Set OUT; (10) Hit F or don’t hit F; Hit IN … rinse and repeat.
Not forgetting, which you did, the need to (11) load up each new clip before carrying out this process.
And also not forgetting (12) that your selected clips will at some point need to be added to a timeline. Not least if you want to review them in a continuous fashion, as you absolutely should.
Compare my method – using FCPX, which I also like to do. (Did I mention that I use FCP X all the time and find it very useful for some projects, less so for others?)
(1) Load all clips into the timeline; (2) Skim or JKL; (3) Set an OUT where you want your first selection to start; (4) Delete; (5) Skim or JKL; (6) set an IN where you want your selection to end …
Start again at (2)
I’m not convinced that your method is quicker … in fact, I my limited maths skills tell me that your method is TWICE AS SLOW as mine. And in FCP X itself at that.
That’s TWICE AS SLOW.
But really, as I’ve tried to make clear, it’s not about keystrokes ultimately, it’s about the value of the result you are left with at the end of each process. I really don’t think there is as much value in your result – for the purposes of the project we have been talking about – as there is in mine.
This isn’t about whether or not I like FCP X – it’s about your oft-promoted concept of editing in the Browser, which I believe has undesirable limitations in certain circumstances, including this one.
Simon Ubsdell
tokyo-uk.com -
[Bill Davis] “It would like somebody saying Schoenburg was WRONG because he didn’t use a “proper” scale. “
It’s interesting that you invoke Schoenberg (I guess from your spelling of his name that perhaps you’re not as much of an avid devotee of his music as I am).
At a time when totalitarianism was starting to sweep across the world and stamp on the face of difference and freedom, good old Arnold was devising his twelve tone system, also known as serialism, to rigidly codify atonalism and banish conventional tonality from the world forever. (Exaggeration for rhetorical effect BTW).
A hundred years later, despite the undoubted serial masterpieces produced by the Second Viennese School and a very small handful of notable epigones, serialism can be safely said to have died and has had no lasting influence on the history of Western music other than to stifle a lot of the creativity of the 20th century with its doctrinaire certainties.
The moral I think is to always beware totalitarian systems.
I wonder what analogy we could draw from this?
Of course, Schoenberg was a titanic genius – the real harm, such as it was, came from his terrifyingly ardent followers …)
Simon Ubsdell
tokyo-uk.com -
[Bill Davis] ”
Don’t worry Simon. It’s impossible to offend me on this. And I’m sorry it’s causing you such Angst. “Thanks for your reply, Bill.
As you know, we both have a history of spectacularly misunderstanding one another, but this time my argument seems to have whooshed majestically over your head like some space vehicle out of Star Wars.
I thought I’d made it clear what I was saying right at the start of my post, but perhaps I should have used capitals, so here goes again:
THERE ARE MANY DIFFERENT TYPES OF EDITING PROJECTS.
I don’t mean there are different workflows, techniques, methods, though of course there a huge variety of these too – what I mean is that different types of editing project have different needs and are suited to different strategies. Not all strategies will work equally well for all types of project – this is the point that you seem to overlook. And it’s a very important one.
Because so much of this can ascend/descend into airy theorising, let’s take a concrete example from a project we worked on for this year’s American Film Market.
We were given sixteen hours of dailies from which to cut a promo and we had just a couple of days to do it. Because we work collaboratively, I volunteered for the enviable job of preselecting from those sixteen hours – and I had only three or four hours in which to do it. Crazy, huh? None of us had seen any of the material before, so had no idea what to expect. However, using the string-out method, I was able to put together a longish selects reel within that ridiculously short space of time. One of my other colleagues spent another couple of hours refining the string-out(s), before handing over to a third colleague who actually cut the promo. Within about 24 hours we had something that the client and film-makers absolutely loved, were bowled over by, couldn’t believe we’d managed to achieve in the time, etc., etc.
This would simply have not been possible without the string-out method, and conversely, the keyword method would have frittered away those precious hours for absolutely no purpose whatsoever.
By working with string-outs we were able to easily pass the project from one editor to another in a way that no amount of the most perfect keywording would have allowed. The previous editor’s intentions and choices were very easy for the next editor to read.
But what would we have done with keywords? What possible use would it have been if I’d tagged the shot of two grumpy men in hats on a beach? 2-SHOT, MEN, BEACH, HATS, GRUMPY – all or any of these would have been pointless, even if I had been the one cutting the final piece. Let’s ignore the time it would take to set those keywords, none of it would be information that would help me with the edit in any way whatsoever. But stringing the shot into a sequence which was already the beginning of an assembly of the scene from which it derived was an entirely self-explanatory system – it required no metadata for my colleagues to understand the logic of the shot and its possible usefulness.
One of the most important considerations for me is that you can sit down and watch a selects reel (string-out) and start to get a very good idea of how the final edit might come together. And that is because you are experiencing your material in real time, sequentially, just as the final result will be experienced. It’s hard to get away from the fact that editing is a sequential process, an unfolding of the material in real time. The string-out process puts that front and centre and works well for that reason.
Of course, there is no doubt that keywording/subclipping/browser organisation is a great boon for many kinds of editing projects – corporates and documentaries are obviously top of the list of types of editing project where this is true, and I speak from experience of both. Sometimes, sifting and sorting clips is a very large part of the editing process, as it clearly is for the work that you do. But to elevate it to a guiding principle that trumps all other ways of working, seems to me to be fundamentally mistaken and not a little short-sighted.
I’ve admitted documentaries to the list of editing project that are well served by the keywording method, but let’s not forget that there is a very distinguished school of thought in documentary film-making where a form of the string-out method has always been central to the creative process. Again, not all documentary editing projects are the same, and lumping them all together is just too crude. And this brings me back to my original point – there are many different types of editing project and it doesn’t make sense to dictate a particular editing strategy (or software solution) as being preferable across the board.
Incidentally, to add further anecdotal colour to my example above, it’s worth mentioning what we used to actually cut the project. As it happens, I began in FCP X .. but gave it up because the import was taking too long. I then moved over to Premiere … but gave it up because the import was taking too long. So I moved over the FCP 7 … where I was up and running in just a few seconds.
And one other point – because we were sharing huge amounts of media and the master material was too big to share conveniently, we “offlined” using the H.264 proxies supplied to us, and then relinked to the master media for the final grade and delivery. FCP7’s outstandingly flexible relinking made this extremely quick and easy, but it would have been simply impossible in FCP X due to its inflexible file handling – far too much differed between the proxies and the master clips for X to have been able to deal with it. Incidentally, the great Ronny Courtens did suggest to me a way around FCP X’s relinking weakness in situations like this, and that was to multiclip the proxy with the master clip and edit from the result and then simply swap the multiclip around at the end of the process. It’s a great tip and it’s worked for me when I’ve had the time to do the preparation. Quite obviously the volume of material and the time constraint in this case meant that this was a total non-starter for us. (I’m so glad that my original intention of cutting in FCP X backfired early on, as I would have roundly cursed the decision later!)
I should finish by saying that FCP X is great and there are several kinds of editing project where I find it a very natural fit and where its particular editing model means I can work much, much faster.
But there are other types of editing project where it just doesn’t fit and its editing model makes for a much slower process – or indeed where its continued lack of key features means that I can’t use it at all, whether it otherwise fits or not.
Simon Ubsdell
tokyo-uk.com -