[Per Holmes]

Hi,

There’s a point that I haven’t thought of that could change this FCP thing somewhat. I’ve all the time assumed that my source media would reside on the RAID along with the output, and that this duplication of media bothered me — as it should, it would double the requirements.

But if we say that we load all the clips into a sequence, and the output the sequence track by track to become multiclip sources, and that these *become the new masters*, i.e. the old footage is deleted or simply done as MXF to begin with, then making single-clip multiclip sources suddenly becomes more of a pre-processing step done at the same time as the logging.

For sure, I would have to go through this sequence of steps HUNDREDS of times, but if there’s no duplication of media, then it’s not nearly as painful.

The problems are that you still can’t used mixed sources as you can in Avid. And you’ll be paying storage for any black on a track. If one camera only runs for 2 minutes out of 45 minutes, you’ll have 43 minutes of needless rendered black.

Best,

Per

[Per Holmes]

Hi there,

For what it’s worth, FCP’s multicam is an almost direct copy of Avid’s multicam, and therefore, neither of them allow clips to start and stop. One can perhaps understand Avid because a studio-based multicamera shoot can easily have continuous sources.

But still, both programs still exclude anything remotely related to reality-type programming, where you might have 10 or 20 cameras going, and you can be virtually guaranteed that they will need to stop, for battery changes, card changes, dead air etc.

Looking at Avid, I’m still seriously considering dumping FCP and the Kona card and going back to Avid, because I’m simultaneously worried about FCP being able to handle very large projects. But I’ve heard from many FCP editors who edit gigantic multi-editor shows without breaking a sweat.

I don’t know, man, I’m very torn. I don’t think I can swallow doing this in Premiere Pro, I’ve seen that program crumble under large projects. And it’s only 4 channels.

So I guess I’m probably back to building the output on Track 10, and Adding Edits and Shift-Ctrl-Dragging stuff up to Track 10.

Hmmm.

Best,

Per

[Per Holmes]

Thanks, I appreciate it!

Best,

Per

[Per Holmes]

Hi there,

I’ve studied up a bit on Premiere Pro, and how they do it is almost exactly right. You first build a sequence with each angle on a separate track. Each clip can start and stop as much as you want.

Then you nest that clip into another sequence, and as part of the nest, you can automate which video track you’re watching.

The huge limitation is that they only allow you 4 sources, which is not enough.

The demos I’ve been watching have been mostly CS3, so I’m wondering if this is different in CS4. I’m also extremely nervous about putting a project this size through Premiere (I’m only slightly less nervous about FCP and I have total faith in Avid as I’ve done giant projects in Avid without even a hickup ever).

Man, this is difficult.

Best,

Per

[Per Holmes]

Hi,

I’m looking at the Nitris DX at B&H, and I can actually get HD out for my Avid for “just” $5,000. It was when it cost $8,000 that I balked, which made me buy an FCP system.

But looking at this, God, I miss my Avid. May I should switch back.

But reading around on the internet, multicam editing with sources that start and stop is apparently no picnic in Avid, although it IS possible to get intermittet sources from the same angle into the same multicam “track”.

Does anyone understand this article? https://viewfromthecuttingroomfloor.wordpress.com/2008/04/17/multigroups/

Is this really how hard it is to get your intermittent footage lined up into a multigroup in Avid? I’ve never personally edited multicam in Avid before, so I don’t know the workflow. Maybe this is a question ask in the Avid forum.

Thanks,

Per

[Per Holmes]

Hi Herb,

I truly appreciate that. Can I ask you a couple of questions about Premiere Pro? It’s not too late for me to switch, and I do have Creative Suite CS4, although I was never expecting I would use Premiere again (last time I used it, it was CS1).

How does Premiere handle large projects? For sure, both FCP and Avid have problems when projects/bins get too large, but I’ve seen Premiere positively go on strike when timelines got long. Is this still so?

I have to say that I’m one inch from unloading this FCP stuff and jumping back to Avid, although it means that I’ll have to swallow the pretty insane price for an HD output via Nitris DX. I can see you can get them at $4,995 at B&H if you trade in your old Mojo, which I could do. I have a Media Composer dongle, and could upgrade to the latest version, pretty easily.

Actually, maybe I ought to jump ship. Anybody wants a Kona LHi?

Best,

Per

[Per Holmes]

Hi,

I’m a little bit restricted in being able to describe the project, but it’s a training program that contains several hundred vignettes, each probably 45 minutes for each “run”, meaning 45 minutes between times where ALL cameras can stop. Some runs will be just 3 cameras, but there will be a couple of big things lasting several days of almost continuous shooting that will be 7-8 cameras.

I currently have a 12 TB RAID set up for this, which I’m expecting to almost fill up, and I might need to expand. Not all 12 TB will be multicam, because about half the TRT is a presenter on a green screen.

So still, we’re looking at probably 6-8 TB of multicam source. To export and reimport that would be *grotesque*. It would not only eat another 6-8 TB, it would take an extreme amount of time, and torpedo any tapeless benefit that ever existed prior to that. Importing as MXF instead of Log/Transfer would only be a marginal improvement over the intense data shuttling that would happen by exporting and reimporting.

So unfortunately, multiclip is not a feature that is available to me. Even if I could export as QT Reference, I don’t dare leave 6-8 TB of data outside the reach of any sort of media management, where I couldn’t even relink clips because it’s all done on the QT backend.

So I’ve resigned myself to using the Add Edit / Add Edit / Shift-Ctrl-Drag To Track 10 method of editing the multicam footage. Then I simply “select” tracks by Option-clicking the Visibility button, Add Edit for the start and end of the segment I want to use, and build the final edits on the highest track.

I don’t see any other way this can be done. I simply don’t consider exporting / reimporting to be an option because of the grotesque duplication of data. If this were a small concert, I’d have done it in a heart beat. But as a repeated workflow for hundreds of hours of footage, it’s just not realistic. For one, I’d have to add another 6-8 TB to the RAID to be able to do it at all.

On the huge upside, it also allows me to use shooters who are non-P2. I’ve had some guys with various cameras send me test footage at 720p/24, and it intercuts without a hitch with the P2 footage on a ProRes 422 timeline. That is very attractive to me, and couldn’t be done with multiclip, because they don’t allow mixed formats.

If Apple gave up the realtime playback idea, perhaps it would be easier for them, because I truly don’t need realtime thumbnails. As long as thumbnails updated once per second or something, I’d in principle be happy, as long as the angle you’re watching plays in full motion.

The suggestion to Apple is to simply allow a regular Sequence to act as a multiclip. Invariably, whether you do TC Sync or PluralEyes or timecode-slates as we do, you end up with a timeline that has each angle on its own track. The next task is inevitably to be able to switch between those tracks. There’s nothing to stop that from happening, except the ability to see the need for it, as well as giving up real-time playback of 9 thumbnails. In fact, they could give up the thumbnails altogether and just have 9 boxes labelled 1 through 9. That would remove any technical limitation to allowing a sequence to be a source for multicam editing. Because a sequence is what will in evitably be the source, for you, me, and everyone else. Turning it into a multiclip via export/import is, at best, a very heavy workaround. And it doesn’t work on large projects.

Thanks,

Per

[Per Holmes]

HI Jeremy,

I think that your MXF4Mac workflow is awesome, that’s clearly something I have to do to avoid Log & Transfer. Very impressive to watch you import 5 hours of footage in 10 seconds.

But the end result is still just a timeline with all the clips properly lined up, and I’m definitely going to investigate the time-of-day way of syncing. But you could also arrive here with timecode slates, or PluralEyes. Or using a TC feed for each camera.

So no matter which sync method we all use, we all run into the same wall that FCP will simply not let you edit it. If any Apple engineers are reading along here, it really is hoped that it’s clear how desperately important this is. Right now, all of us multicam shooters do everything right, and we do it in the only way possible, but when the footage hits the editor, we hit a brick wall.

There are only so many ways to do multicam. It is inevitable in all but the most rare and theoretical circumstances, that your clips will start and stop. At present, FCP outright denies editing of any multicam shoot that has even just a single break. But any real-life multicam shoot has hundreds of breaks, and there’s just no assisted way to edit that in FCP. The multiclip tool is simply no help at all, it is in fact so technically impossible to use that I have no experience with the functionality — I’ve never made it that far! It’s not possible to get the footage into the feature to begin with.

Jeremy, checkout the other subthreads under this topic, as I’m trying to come up with some ways of editing the multicam footage with a modicrum of multicam logic to the workflow. Since you’re undoubtedly much more experienced with FCP than I am, I’d love to hear your thoughts.

Best,

Per

[Per Holmes]

HI Jeremy,

Yes, it is an organisational challenge to be sure! I’ve never done a large project in FCP (although I’ve done plenty in Avid and it never caved), I’ve read in many places that FCP can get funky with very large databases, so this will be split up among several hundred smaller projects with categories of footage, and then separate projects for each segment of the show. I hear that this segmentation is key. Actually, Avid freaks out too if bins get too large.

So far, the footage I’ve gotten from the Sony cameras — and we’re still just testing — was somehow predigested by the videographer, because I got .mov files. So you’re right, it’s key that I investigate this with regard to MXF, because if it solves the multicam issue, that alone would be enough to require a solely P2 workflow with no other types of sources, and then the Sony guys will have to bring different cameras. This was the original intention anyway, but then I got impressed seeing the timeline mix and match different types of sources without rendering, and suddenly realized that I could get many more people to join in besides the main production, and really get an awesome amount of footage to edit from.

I’ll watch the link right now.

Cheers,

Per

[Per Holmes]

HI Kevin,

Thanks! Yes, I get heated up when I’m frustrated! It really is something that Apple should address, because it’s almost impossible to do a multicam shoot without a single break in a single camera unless you’re in a studio and a highly controlled circumstance. With all these sources, the only logical place to put them is each on their own track on a timeline, and then find a way to track-switch the timeline — it doesn’t even have to be real-time, which is probably what drove the Apple engineers to require a single clip and all matching formats. Unfortunately, it’s rather impossible to bring such ideal footage to the edit table, so it truly is a conundrum — because the fact is clear, I MUST find a way to edit it. The workflows I’ve described are barely tolerable, and no doubt, they will see improvement as well.

Thanks for your compliments.

Per