[Joe Marler]

[Wu Zhengjiang] “I usually convert all my footages into mov by Final Cut Pro X.
The reasons are: 1, we only serve local clients, so many footages are reusable. Converting MXF into mov gives me the convenience of quick look in Finder. Since MXF files won’t display thumbnails in Finder, which is very annoying when I try to find certain footages in Finder…FCPX does a good job of converting footages, the Slog files will remain same quality after converting. But there is one trouble. Those mov files could only be opened in FCPX, DaVinci Resolve and quick look in Finder, but not Premiere, After Effects, and Fusion, also almost all video players couldn’t play them, except Quicktime..”

MXF is a wrapper format so there could be various codecs inside. The 4k 4:2:2 300 mbps MXF material I have from a Canon XC10 plays OK via Quick Look and Quicktime Player, however the thumbnails don’t show in Finder. For me this isn’t a major issue since I always import all content to FCPX with “leave files in place” and organize it there. The FCPX skimmer is much faster than any external product for viewing and categorizing material.

You may need Apple Pro Video Formats installed to enable MXF playback or maybe you have a version where the codec isn’t supported at the Finder level: https://www.newsshooter.com/2016/11/22/apple-pro-video-formats-2-0-5-update-adds-mxf-integration-to-macos/

Rewrapping (not transcoding) MXF can be very fast if using a tool like EditReady: https://www.divergentmedia.com/editready/understandingmxf

[Joe Marler]

[Stuart Christensen] “….after I reformat the P2 cards, the random nature of sequential ordering performed by the camera will change, and if I import those new clips into the main import window of my project, they get scattered in between all of the footage that is already imported and marked up with I/O points…”

The files listed in the FCPX import window will by ordered by whatever criteria you want — name, date, size, etc. If you order by ascending date, the new files should be at the bottom — not scattered in between previously-imported files.

By “main import window” if you are talking about the Event Browser, the files will likewise be shown in whatever order you want. You can order or group them various ways. Use menu option View>Browser>Sort By and/or Group Clips By. Or you can change the Event Browser to List View and simply click on the heading.

I normally copy each day’s material to a separate date-named folder to avoid having to pick through one big folder for new material. You can have FCPX automatically tag the incoming material with the folder name. See the import dialog or FCPX>Preferences>Import>Keywords: From folders.

You can also tag the files at the Finder level and have FCPX apply those tags as keywords during import: https://www.fcp.co/final-cut-pro/tutorials/1456-using-finder-tags-to-help-with-importing-media-into-final-cut-pro-x

However it’s not necessary to even use the FCPX import dialog. You can drag/drop files directly from Finder on top of the event in FCPX. The various import behaviors are then determined by the FCPX>Preferences>Import settings.

[Stuart Christensen] “….So, I was wondering if anyone has a workflow or solution to create a folder or (Bin) in the main window of the project so I won’t have to keep creating “events” and copy and paste the clips back into the main timeline. This process is seriously tedious and maddening….any help would be greatly appreciated. I just started using FCPX 10.3.4 about 4 months ago”

For my doc team, I normally copy all media for each day’s shooting to a date-named folder, and within that by folders by camera name. I never select import ranges — that feature is only available for tree-based media like AVCHD or XAVC-S. I import all media with “leave files in place” and auto-assign initial keywords based on folder names on disk (mainly shooting day and camera name). Since multicam requires each camera angle be named in FCPX and since we have several cameras which make similar filenames, this allows quickly isolating the clips from an individual camera to batch-assign camera name or angle in the FCPX metadata: https://support.apple.com/kb/PH12658?locale=en_US&viewlocale=en_US

I rarely even look at the material outside of FCPX, except for confidence checks. I don’t spend time trying to meticulously organize the media by files and folders. This would require looking at the material with some viewer which is vastly slower than the FCPX skimmer. After importing I use FCPX to favorite, reject and keyword the material.

For some projects I use a single event and use FCPX database methods to organize if after import. This is extensively organized in the Event Browser before any timeline editing is done. It is not uncommon to spend over 60% of the total editing time on organization, before even using the timeline. Then the timeline editing phase goes much faster.

For other projects (say a year-long production filmed at various sites separated by months of time), I use an event per physical shooting site. Strictly speaking this isn’t necessary — the material at each shooting location could simply be keyworded. However having separate events helps facilitate multi-editor workflows where various assistant editors are keywording and rating the material in assigned events, which are then consolidated by sending an event XML to a master copy of the library.

[Joe Marler]

[Michael Hadley] “With an AMD 580 card, Bruce x goes from 24 secs to 16 secs. That seems like a pretty robust improvement.”

My 2017 top-spec iMac 27 does BruceX in 15.8 sec, and transcodes from 4k H264 to proxy about 2x faster than a 12-core D700 Mac Pro. A 512GB SSD version with 8GB RAM is about $2700, or roughly $2300 if available on the Apple refurbished site. A 3rd party 32GB RAM upgrade would be about $350 more.

Of course the Sonnet eGPU box plus RX 580 is only about $700, but that won’t speed up H264 or H265 at all.

A 10-core Vega 64 iMac Pro might be about 1.8x or 2x faster on both CPU and GPU than the top-spec iMac 27 but nobody yet knows, and how that translates to real-world performance is anyone’s guess. It would also likely be quite expensive, maybe around $7000.

[Joe Marler]

[Ray Sherman] “Would you suggest transcoding to .MOV? Thanks again, Ray….”

As I previously said, you simply need to import your camera’s XAVC-S (a form of H.264). That is your camera’s native codec, and is the best available quality for your camera at 4k. You do not need to transcode to optimized media for quality reasons. If you need to transcode to proxy for performance reasons, that is OK. You do not need to externally transcode to .MOV (maybe you mean ProRes). .MOV is a Quicktime container format that may contain either H.264 or ProRes.

You can optionally re-wrap (not transcode) the XAVC-S before import using EditReady, but IMO this isn’t needed. You can import XAVC-S content with “leave files in place” if you copy the video files out of the media folder before import. Using EditReady does preserve all original metadata prior to in-place import, but so far I haven’t seen problems with not using it (on XAVC-S). AVCHD is very different however — the bare files should never be imported to FCPX using “leave files in place”. In that case using EditReady should be mandatory: https://www.divergentmedia.com/editready

Transcoding to optimized ProRes media will increase your media size by about 6x, greatly increase the I/O burden, and will not help quality any. You cannot create resolution or color data from nothing. It will help editing performance but so will proxy and without the size penalty and I/O load.

Since FCPX always edits H264 using a ProRes buffer, editing H264 media does not handicap color correction, compositing, gradients or most anything else.
Transcoding to ProRes does not reduce the need for a powerful GPU. E.g, compute-intensive effects such as Neat Video are just as slow whether the material is ProRes or H264. It is true that some editing software which does not use Quick Sync can struggle with 4k H264 or even FCPX can on a Xeon-based Mac Pro since Xeon CPUs (with few exceptions) don’t have Quick Sync. However transcoding to proxy will greatly increase performance.

Proxy does not support alpha channel (ie transparency) data so in that specific case using optimized ProRes might be necessary. Optimized media might be needed if editing mixed H264 1080p and 4k material. FCPX has a global proxy mode so all content is viewed either as proxy or optimized/original. Proxy of 4k is 1/2 the linear resolution (1/4 the total pixel resolution), or 1080p. In that case there’s plenty of resolution for editing decisions. However proxy of 1080p is only 960 x 540, which can be a little low for some editing decisions. This might necessitate transcoding to optimized media to obtain good performance on 4k yet retain original resolution on 1080p for editing decisions.

You can edit the 8-bit 4:2:0 4k H264 material in a 1080p timeline and export it as 10-bit 4:4:4 1080p. That is not creating bit depth and improved chroma sampling from nothing, but *trading* resolution for it. If you don’t crop it and export to ProRes it will supposedly become true 1080p 10-bit 4:4:4 on output, although I haven’t put this under a microscope to test it. See white paper by Barry Green “The Benefits of Shooting in 4k”: ftp://ftp.panasonic.com/provideo/agdvx200pj/4kbenefits_techbrief.pdf

[Joe Marler]

As Jeff said, it could be hardware, system layer or software. A common problem is running out of disk space which then destabilizes the OS or the app. Check you have plenty of free disk space on all drives.

Before you start FCPX, verify you are getting expected performance on each drive by running Black Magic speed test: https://itunes.apple.com/us/app/blackmagic-disk-speed-test/id425264550?mt=12

General FCPX troubleshooting recommendations from Apple: https://support.apple.com/en-us/HT203477

My own suggestions (inc’l some of the above):

(1) Locate and make a manual backup of the current library. Reason: even though FCPX automatically makes backup copies of the library every few minutes, this is an additional safety step before proceeding further. Note: this assumes you’re using only external media. Duplicating the library may not be practical if you are using a managed library with all media inside — it is likely too large to duplicate.

(2) When located, right-click that library in Finder and select “duplicate”; wait a few seconds. That makes a duplicate copy with a numeric suffix. This doesn’t fix anything it’s just a contingency backup.

(3) If FCPX won’t start normally, reset FCPX preferences. This discards a few user-configurable preferences that can easily be set again. FCPX is known to become erratic if this file is damaged. Procedure: while FCPX is not running, press OPT+CMD and click on FCPX in the dock at bottom. It will ask do you want to reset preferences. You say yes.

(4) After a minute it should start up looking like a new installation with “Welcome to FCPX”. If it won’t start even, then it’s a deeper problem. If it does start the problem was data-specific, probably inside the library or due to a damaged preference file.

(5) Set your preferences back to preferred values in the menu Final Cut Pro>Preferences, esp turn off background rendering while you are troubleshooting.

(6) Resetting preferences deletes the last known location of the library, so you can load that from Finder by double-clicking on it. It should load into FCPX. If it does load OK the problem was corruption in the preference file. If it does not load normally the problem was in the library.

(7) If it does not load properly or FCPX hangs, you’ll need to force down FCPX by doing OPT+CMD+ESC. IF that doesn’t work you can gracefully force down macOS by holding down CTRL+OPT+CMD+Eject key (at upper right).

(8) If FCPX won’t launch with your main library, you must launch it without that. While holding the OPT key, click on FCPX in the dock. This makes it starts without loading any specific library, and gives you the option of what to load. Click “new” to create a new test library and make sure the program will at least start with a blank library.

(9) If it starts with a blank library the next step is try the previous auto-backup version of the main library. To do this:

(a) With FCPX up, pick menu File>Open Library>From Backup, then pick “Other”
(b) A list of library backups will be shown. Pick the most recent one.
(c) If that loads OK, examine your most recent work and see what is there.
(d) If it does not load OK or FCPX hangs, repeat the above procedure and pick the next most recent library.

(10) If none of those work, some users have reported deleting the “flexolibrary” file inside the library bundle has fixed some problems. Always have a backup copy of the library, although if it’s very large due to managed media this could be difficult.

[Joe Marler]

[Nikos Papadopoulos] “Usually this happens if the system has NVIDIA instead of AMD”

I have four Macs with AMD GPUs and I periodically see this problem on all of them. As soon as 10.4 is released I’ll test that.

[Joe Marler]

[Andrew Mass] “dilemma I’m having with FCP: I’m editing a documentary that I set up with many libraries (one per interview, etc)….This is now causing some headaches as editing a project from many libraries is seemingly forcing a lot of duplication – and just feels sloppy….Making one giant library also seems like trouble just because of sheer size.”

I’m editing a documentary composed of 6,000 clips or 200 hr of 4K H264 material, equating to 8TB of original media and 5TB proxies. This will eventually be about 30% larger or about 18 terabytes peak and 260 hr of material. This is organized mostly in a single library composed of about 10 events. It is running on a top-spec 2017 iMac 27 and a 32TB RAID-0 Thunderbolt array (with identical backup), and performance and reliability is generally very good.

Having a single large library works well, at least on this hardware and up to this size. As you discovered, each library is a stand-alone entity and you cannot search or use database management across libraries.

[Joe Marler]

[Larry Watts] ” footage shot a 1080 x 1920 but have some interviews I need to intersperse which are 4K. (Both have the same frame rate)….The 4K will be great to do pans and zooms. Will they behave on the timeline okay?”

In a 1080 timeline the 4k clip resolution will be used for pan/zoom, and if you export to 1080 the underlying 4k resolution will be visible. IOW if you scale a 4k clip 200% in a 1080 timeline, even though you may not see the 4k resolution in the viewer, when you export it the resolution will be there. It will look just like it was shot at 1080 and not scaled. To avoid zooming in too far and losing resolution you keep track of the scale factor as shown in the Inspector. 200% is the limit for UHD 4k to 1080. If you are exporting 4k to 720 the limit is 300%.

The problem with intermixing H264 1080p and 4k is sometimes proxies are required to obtain good performance on 4k. Proxy of 4k is 1080, so plenty of resolution for editing decisions. But proxy of 1080p is only 960×540, so it’s somewhat rough.

Since the FCPX proxy mode is a global switch, you can’t view the 1080 clips at full res and the 4k as proxy in a mixed configuration. One solution is to transcode the 4k to optimized ProRes but that can take a lot of space. But with mixed 1080/4k media that would provide better editing performance for the 4k clips while staying out of proxy mode and show the 1080 clips at full res for better editing decisions.

[Joe Marler]

[Bret Williams] “If you can create it with just Motion, it will perform better. Every plugin has to go through the Motion engine, which is where FXPlug is supported.”

It is my understanding there are two categories of plugins: (1) Those created solely with Motion and (2) Those built using a programming language. These may be exposed to FCPX through a Motion interface but the core functionality is custom programmed.

Example of #1 would be Ripple Callouts: https://www.rippletraining.com/products/plugins/callouts-3-0/

Examples of #2 would be things like Digital Anarchy’s Flicker Free, Neat Video noise reduction and Imagenomic Portraiture. I don’t think these can be created solely with Motion because they contain sophisticated image processing logic. However I don’t know what programming languages were used to build them, or what the software development environment is, or the plugin architecture (e.g, do they run in-process or out-of-process). It would be interesting to know that.

https://digitalanarchy.com/Flicker/main.html
https://www.neatvideo.com/
https://www.imagenomic.com/VideoPluginSuite.aspx

[Joe Marler]

[dallas kruse] “Here’s a screenshot of the Quicktime info for the format…I dragged my footage from my Video Drive into FCPX using Proxy. Seems to be no issues….the file names for future projects might have the same name? I’m hearing that Sony doesn’t automatically parse the file names into sequential order?…For future reference, which format SHOULD I be filming in when renting a Sony a7s ii?”

That is XAVC-S HD. You do not need proxy for that, as almost any computer can edit H264 1080p without transcoding. It also does not require rewrapping with EditReady. However FCPX will not import with “leave files in place” from any tree-oriented file format. To achieve that you must copy the .mp4 files out of the folder structure and import from there. That is considered a poor practice but in my experience it causes no problem with files from the Sony Alpha E-mount series. That should never be done with AVCHD media.

The A7SII will reuse filenames starting over with C0001.mp4 whenever you insert an empty SD card. That cannot be overridden, however it’s only a problem if you will be doing metadata transfers via event or library XMLs. In that case it can create spurious duplicate clips in the Event Browser. However as a good operating practice, it’s often a good idea to rename the files before import to (1) Make them unique and (2) Add additional info such as camera type or owner. This is useful in multi-camera shoots when you have several cameras producing similar filenames.

Re what format to use, I would generally suggest using XAVC-S 4k, whether you will be distributing in 4k or not. Admittedly that will require creating proxy files to achieve good editing performance, but FCPX handles that very well. Just remember to set the viewer from proxy back to optimized/original before exporting, else the output file will be at proxy resolution.

Using 4k gives you the ability to crop and recompose in post without losing resolution for 1080p distribution, each frame is an 8 megapixel still so it’s useful for frame grabs, and also some 4k cameras don’t produce good-looking 1080p.