Best fcpx workflow for lengthy interview clips and 500GB AVCHD media

Hello to you all.

I am a novice doco maker, video editor and FCPX user.

I'm seeking advice on preparing the editing workflow (transcoding/re-wrapping, file location) for a lengthy documentary; something I have stumbled into after meeting a fascinating 'subject' and deciding I would like to tell their story.

Some background might help:

I purchased a Canon C100 and have shot about 450GB of AVCHD clips; most of these are well lit INTERVIEWS; long takes of between 1.5 and 2.5 hours each. Some of this AVCHD footage is re-enactment scenarios and others are 'run-and-gun' location shoots where the subject is miked wirelessly and gives commentarty on each 'location'.

In addition to this I have a series of TEXT-based elements, photos, drawings etc, and some Skype interviews footage which I may wish to include occassionally.

I have chosen FCPX over Premiere Pro because I believe its metadata 'keyword/tagging' and search features will help me organise my material before structuring the story.

I have ordered a new 27-inch iMac with Retina 5K:

• 4.0GHz Quad-core Intel Core i7, Turbo Boost up to 4.4GHz
• 32GB 1600MHz DDR3 SDRAM - 4x8GB
• 1TB Flash Storage
• AMD Radeon R9 M295X 4GB GDDR5

and have my eyes on a Pegasus2 R4 diskless which I plan to populate with 4 x 4TB Hitachi 7200rpm drives.

The clips look pretty good overall but I would like to have the option to do some colour grading or at least colour balancing. I would also like to use a few 'effects' and take advantage of the lovely new 3D text feature in FCPX.

The main issues which I would like your help with are:

1) should I transcode to ProRes 422 or edit in AVCHD?

2) what is the best way to manage these large AVCHD files? eg. if I am forced to COPY (see my point 4) should I then delete the AVCHD media from its 'source' HD once import/transcode is complete? (NB: I have other copies of these 'original' AVCHD files on several ext HD's)

3) if I transcode, should I do this before import into fcpx (in either ClipWrap or EditReady) and perhaps choose ProRes LT to keep the file sizes down? OR within FCPX; the latter could result in a storage issue as I understand the ProRes 422 will more than triple the file size of AVCHD clips.

4) Due to the 'leave in place' selection being 'ghosted out' on my import options for these C100 AVCHD clips, I am already faced with the issue of having to COPY my media at least once on import. This has made me more open to considering the transcode option. I have also read a few editor's comments about some minor video artefacts persisting during editing of AVCHD cips, which are not present after transcoding within FCPX.

5) I am presuming I need to import all clips into FCPX before I can utilise the 'keyword/tagging' options. This will be an important part of my workflow as I have about 7 different interviewees discussing similar topics, so I will be labelling multiple 'ranges' within each of these lengthy clips. The only issue with this is storage space. Is there another way to 'label keywords for ranges of clips' and then only import those labelled sections into FCPX?

Thanks in advance and apologies for this lengthy post.

Feel free to chime in on any specific point if it is easier for you.

Kind regards,
Shane

Update; after further reading I have swapped my ClipWrap app for EdirReady and will do some transcode tests today on my MBP.

1. Edit in the native QuickTime file that gets created when imported from the camera card

2. Keep the entire card card archive on one drive as a backup. Import from that backup undo your main editing drive. Don't delete anything.

3. Don't transcode. You don't need EditReady,

4. You import from the archive into the library location, You can't leave camera media in place. You don;t want to.

5. You don't need to import everything at once. It's easier to import some material, keyword and tag, and then import more material. It's easier to do when you have groups of material that are being assigned the same roles or being put in the same collections.

Thank you so much, Tom.

I would like to ask you for a few minor clarifications:

1. Edit in the native QuickTime file that gets created when imported from the camera card

By 'native QuickTime file' do you mean I should leave the AVCHD file intact and NOT select the 'optimized media' button? Was my concern that editing in AVCHD can result in minor digital artefacts unfounded?

2. Keep the entire card card archive on one drive as a backup. Import from that backup...

Should I locate my Library on the RAID-Pegasus?

Shane

shanesangle wrote: ...lengthy documentary...about 450GB of AVCHD clips; most of these are well lit INTERVIEWS; long takes of between 1.5 and 2.5 hours each...

Maybe BorisFX Soundbite would help. One poster fcp.co poster tested it and had some issues but for long documentaries it might be worth a look: www.borisfx.com/products/soundbite/
www.fcp.co/forum/4-final-cut-pro-x-fcpx/...-audio-to-text#59283

You will definitely need an external RAID array. The Pegasus R4 is good -- I have an 8TB Thunderbolt 1. You will also need something to back everything up (inc'l what's on the RAID). In addition to Time Machine I also use Carbon Copy to another RAID backup drive.

FCP X is very different from other editing programs and I recommend scheduling adequate time to learn how to best leverage it. Mike Matzdorff (assistant editor of the Will Smith movie "Focus") became so frustrated at FCP X he wanted to throw his computer against the wall. He eventually appreciated it and cut the movie using FCP X. Many of us who came from other editing programs have had similar experiences.

www.fcp.co/final-cut-pro/news/1512-mike-...wood-feature-on-fcpx

OS X is not particularly friendly to AVCHD -- all the clips are within a monolithic "package" and you can't easily scrub through them at the Finder level. As a consolation Quicktime shows thumbnails and lets you play them but it's clunky. This is really OS X trying to help you since AVCHD material should be kept together. It is sometimes tempting to transcode or re-wrap everything using ClipWrap. However this is only promoting the old way of handling media which is filter it once at the file system level, then import those selects into the editor. This results in going over the material twice, maybe even taking notes and then entering those notes as tags/keywords in the editor.

It's probably better to just import everything, not transcode or re-wrap, and use the FCP X keyword/tagging tools.

Ripple Training: www.rippletraining.com/
MacPro Video: www.macprovideo.com/
Lynda.com: www.lynda.com/Final-Cut-Pro-training-tutorials/201-0.html
Larry Jordan: www.larryjordan.biz/
Tom Wolsky FCP X: Beyond the Basics: www.fcpxbook.com/

While the above are all good, there are also various classroom training providers. Those can provide an expedited, more personalized way to get up to speed.

Here is a complete narrated step-by-step walk through of an actual production edit on FCP X, with emphasis on pre-edit organization. He was using FCP X 10.0.8, which is before the new library scheme. However 99% of what he says applies to the current version: fcpxposure.com/video-editing-with-fcpx-at-the-aipp-2013/

There are many plugins for FCP X which greatly "sweeten" the final product. It is worth getting familiar with some of these. Just a few (in no particular order):

www.coremelt.com/
www.fcpeffects.com/
www.borisfx.com/
www.digitalanarchy.com/
fxfactory.com/
www.redgiant.com/products/pluraleyes/
www.izotope.com/en/products/audio-repair/rx/
www.arcticwhiteness.com/

Hello Joema,

Firstly, thank you for such a detailed lot of advice and those helpful links. You really are building good 'karma' in my book.

I'm watching the podcast with fellow Aussie Doug Suiter (FCPXposure step-by-step edit walk thru) and the Mike Matzdorff talk was great.

BorisFX's SoundBite looks amazing; I was exploring audio to text technology some weeks ago but missed this app. Will consider.

I've been ploughing (plowing) through MacProVideo tutes on FCPX and am learning much. I find Larry Jordan very clear and generous too in some of his freely available tutes, though I haven't subscribed. I will peruse the other links you sent.

Your advice to get organised before I begin is welcomed. I have been doing as much as I can to understand the task ahead, and gradually making choices to determine a good workflow.

I've purchased the Pegasus2 R4 and will now figure out a back-up for it.

Tom Wolsky (like you) suggests editing in AVCHD rather than transcoding and double-handling in EditReady. Given my files were expanding to 4.8 times their original AVCHD size after transcoding in EditReady to ProRes 422, I will follow this advice and keep all in native AVCHD.

Thanks again for such great assistance!
Shane

Yes to both. When you import AVCHD into FCP through the import window, the media is rewrapped to QuickTime, and should not need optimizing.

There was a long discussion about this on the Apple forum recently and I write a lengthy reply, which can be found here. www.fcpxbook.com/tips/optimize-proxy-and-render.html

Yes, the library and the media and cache file should be on the RAID. There no benefit in having the library on another drive.

shanesangle wrote: ....editing in AVCHD rather than transcoding and double-handling in EditReady...

My only caveat is certain codecs are so CPU-intensive that I don't get as responsive "JKL" timeline control unless they are transcoded to optimized media. That's on a top-spec 2013 iMac with 3.5Ghz i7, 32GB RAM, GTX-780m, Pegasus R4 Thunderbolt 1. Examples of this are the .mp4 files from a GoPro Hero 3 and above and the 1080p video files from DJI Phantom Vision 2+ and 4k from Vision 3 drones.

In some case it also helps editing performance (esp on 4k material) to select View->Better Performance at the upper right of the playback window.

So except for a few cases just editing in native camera format works OK. I definitely would NOT transcode or optimize speculatively on the theory that it *might* help. It usually takes a lot of time and space and doesn't help materially, at least on newer machines.

Even if transcoding were needed this need not be done on import. Documentaries have a high shooting ratio, so it's better to wait until the assembly edit phase or later to transcode -- if needed. The FCP X "skimmer" still works very fast on highly compressed material, so you can scrub through it in native format. Later -- when and if transcoding is needed -- in FCP X you just right-click on the timeline, select "reveal in browser", then right-click on the event browser clip and select "transcode". It transparently uses that material. You can later delete the optimized material with either FCP X or the 3rd party tool Final Cut Library Manager, without losing any edits.

Thank you Tom. I found that same article earlier today after Joema referred me to your site, amongst others. It is full of good info. Re-reading it now. I'm cramming in as much 'training' before the new machine arrives next week and I start importing media and meta-tagging/organising.

Thanks Joema, very informative as usual.

I've spent most of the day following the links you sent. I like the option of transcoding after the thing is mostly cut together; if at all.

I didn't realise transcoding was an option after editing had begun. But I guess that is what is meant by non-destructive editing.

It is interesting how sluggish fcpx 10.1.2 is on my current 2009 iMac, 2.6GHz i5, in spite of the SSD drive and the 32 GB Ram. The firewire port on this old dear is dead so it is using USB2 to render to my library. Man is it slow. Can't wait to compare it with the new Mac and fcpx 10.2 next week.

Have a good weekend.
S

shanesangle wrote: ...It is interesting how sluggish fcpx 10.1.2 is on my current 2009 iMac, 2.6GHz i5, in spite of the SSD drive and the 32 GB Ram....

Your new 5k iMac will be vastly faster than this. Another major advantage is the i7-4970k in your new iMac supports Intel's "Quick Sync" on-chip transcoder. When exporting to single-pass H.264, it will be about 4x or 5x faster than an *otherwise comparable* CPU. Factoring in the other improvements since the Lynnfield-generation i5, I wouldn't be surprised if exports were 15x faster.

Note to use Quick Sync you must use either H.264 single pass or "faster encode". In general I can't see any quality difference between this and "better quality".

If exporting directly from FCP X the steps are Share->Master File->Settings,
Format: Computer
Video codec: H.264 Faster Encode

If exporting to Compressor, you will need a custom preset to achieve this, since the default presets are multi-pass. Oddly exporting H.264 from FCP X is still faster than Compressor even though Compressor was supposedly GPU enhanced the last update.

Despite how fast the export is, various effects (esp. 3rd party) can be frustratingly slow, even on a top-spec iMac. However some of these effects are adding significant value, so maybe it's worthwhile. In the case of Digital Anarchy's Flicker Free, it's fixing something that would otherwise result in discarded footage, so maybe slow rendering is an OK price: www.digitalanarchy.com/Flicker/main.html

Wow. That speed bump in exporting sounds terrific!

Flickerfree looks handy. No doubt - some compromised rendering speeds on useful 3rd party 'polishing' plugins is a fair price to pay.

Thanks Joema