Joe Marler
22
Apple Final Cut Pro X Coins
39
COW Coins
Forum Replies Created
-
[Brian Dunning] “is there ANY way to automate/improve this? I’d be happy with “somewhat improved” and don’t expect perfect. Thanks.”
In some cases you can use Digital Anarchy’s Flicker Free to improve this. I have used it on drone videos which have step-like exposure fluctuations and it helped considerably: https://www.digitalanarchy.com/Flicker/main.html
OTOH I had an interview case similar to yours and it was less useful there — I ended up keyframing the whole thing by hand. I think Digital Anarchy has a demo version so you could try it. Be advised it is very compute-intensive and takes a while to run, plus there are various tuning parameters so you often end up running it multiple times to adjust it optimally.
This video describes how to use the After Effects color stabilizer to fix or improve this type of problem:
https://www.youtube.com/watch?v=bYqaw39ADto -
This problem was resolved by renaming the new RAID volume name to the original name. When using external non-managed media combined with external proxies and moving content to a new volume name, FCPX relink does not work reliably. The workaround is renaming the new volume name to the old name. This entire area needs to be improved, including specific UI management of external proxies to fully support a proxy-only workflow. A somewhat cumbersome workaround is described in Ripple Training’s 10.3 Media Management tutorial.
-
[Tony West] “….X has other tools that are cool like the skimmer and background rendering and many more, but that’s not why most people decide to use X or not use X…It’s that fundamental concept of how the timeline works through connected clips that people either like or don’t like. it’s what makes X different….”
This is often said but I don’t fully agree, and I hope the documentary does not focus excessively on the timeline aspect. I’ve heard of several Premiere editors who use FCPX for content organization then edit in Premiere. Even FCPX advocate Sam Mestman said he’s not convinced the magnetic timeline is as big an advantage as the database, skimming and organizational features.
FCPX was apparently a convergence of several concepts: (1) The software should help organize the data, not just run the timeline. This includes deeply integrated database functionality (2) A fluid, extremely high performance skimmer can facilitate high-speed visual browsing of the media (3) The skimmer performance and Event Browser functionality largely make the traditional two-monitor view of source and program monitors unnecessary (4) The software should assist the user with physical media management and minimize missing media cases (5) It should have improved native codec support and high-performance playback like Premiere and not mandate transcoding to ProRes except in special cases (6) Overall UI should be simplified and use “reveal as needed” elements (7) Magnetic timeline, including the “storyline” concept of embedded or collapsed content rather than stacked up tracks
During the FCPX conceptual phase, did separate individuals advocate each of these and a blend was achieved? Or did one or two people equally promote all these individual elements? In hindsight which of these were most/least successful?
In Walter Murch’s book Behind The Seen, he discusses the burden of organizing, logging and trying to stay in control of the huge volume of material. He lamented the time-consuming need to make a huge assembly edit and start trimming that down. Using legacy FCP on Cold Mountain, he tried to manage the material by printing out filmstrip thumbnails and pasting these next to his workstation. He wrote metadata on post-it notes. He entered metadata in Filemaker Pro which he queried while editing to find clips. This is essentially a crude, manual version of FCPX: https://designingsound.org/wp-content/uploads/2013/02/walter_murch.jpg
When reading this book it almost seemed like Apple used that and similar feedback to partially form the feature list and conceptual approach for FCPX. It would be interesting to know if that played any part.
-
[Jacob Holcomb] “latest top spec 27” Imac…documentary project….is spread out over 12+ one tb disks, plus some other stuff in random external drives. Most of the footie was converted to Prores so it would run on my 2008 hackintosh before it exploded…I intend to re-import only the masters and limit transcoding to what is necessary using the 10.3 organizational workflow….Pegusus3 R6 is $2300 on b&h right now and comes with an extra 4tb backup thingie = 20gb raid5…
I’m a documentary 1st AE so I deal with lots of technical, media and workflow issues. Your new iMac is vastly faster than your old Hackintosh, especially on H264 since it has Quick Sync. Promise arrays are great (I have an R4) but I have switched to OWC Thunderbay 4 arrays since SoftRAID is just as fast and it avoids locking you into a proprietary hardware RAID format. E.g, if any of my Thunderbay chassis fail I can put those drives in any other JBOD chassis and access the data. If a Promise chassis fails those drives can only be accessed in another Pegasus. However that is an uncommon scenario. OTOH if you need 20TB you’d have to get a 32TB Thunderbay 4 to provide 24TB of RAID-5 capacity.
Yet another viewpoint is just ditch RAID-5 and use RAID-0 which gives more capacity and speed. Normally you’d never do this on a four drive array but in a sense your media itself doesn’t need redundancy since you obviously already have it somewhere — that itself is your backup. You mainly need to backup your library and edits. I’m not saying do this but it’s another possibility. But if your current media is on aging and questionable drives, those need backing up — even IF you copy it to RAID-5. So RAID-5 does not eliminate the need for backup.
But there is nothing wrong with the R6 — it is a great array — very quiet and good performance. It is a known and trusted item and lots of productions use those.
[Jacob Holcomb] …started out with miniDV, then 720, then 1080 in several different flavors from gopro to avchd to xdcam. There’s definitely new material coming in, including possible 4k b-roll, which explains the urgency of “oh crap, I need to figure out where to put this stuff!” as well as do a rescue of the aging drives and cross my fingers that I made enough backups….how to you back up the new Library files? They could get pretty huge to the point where a single one could fill up several disks….
In general you want to import your media with “leave files in place”, and you don’t normally need to generate optimized or proxy media. This makes the import much faster and avoids taking extra disk space and time for transcodes. FCPX is fast enough to smoothly edit most camera native content. But there are a few exceptions.
If you have AVCHD content be absolutely certain you do not import those with “leave files in place”. FCPX does not handle that well and it can cause I/O performance problems. If you have the original folder trees FCPX will not allow import in place but will copy those to the library. If you only have the .MTS files you can import in place but never do this. The best solution for AVCHD (whether trees or files) is rewrap those using the EdityReady utility, then import them with “leave files in place”. This is the fastest approach and takes the least space: https://www.divergentmedia.com/editready
Even your top-spec 2017 iMac 27 may not be fast enough to smoothly edit or scrub 4k H264, depending on the exact codec variant. This may require transcoding to proxy. You do this after import, so you can evaluate the performance beforehand. By default proxies go into your library so greatly swells the size. There is a procedure to keep the proxies outside that’s documented in Ripple Training’s 10.3 Media Management class, but IMO it’s somewhat cumbersome: https://www.rippletraining.com/products/final-cut-pro/media-management-in-final-cut-pro-10-3/
[Jacob Holcomb] …8 years ago I was forced to transcode everything just to get it all to play together, but sounds like it’s a new era where a lot of that stuff will exist on the same timeline (or project or whatever they call it now) at least until it’s time to color….
FCPX is a great tool for documentary since it allows rapidly skimming through vast amounts of content and rating or keywording this. In general FCPX can handle a wide variety of media. However with such diverse media you will probably encounter a few bumps and will need to figure out a few things.
For interlaced DV content it’s often best to import this into a progressive timeline and manually deinterlace that. Select the clip in the timeline, in the Inspector at the bottom right change from Basic to Settings, then above that select the deinterlace checkbox. For an interlaced project this doesn’t do anything since by definition it must remain interlaced. For a progressive project FCPX does not automatically deinterlace, so this can help.
When dealing with mixed progressive and interlaced content, always do test exports and evaluate this with a tool like VLC which allows selective control over deinterlaced playback. Mistakes on interlaced content are common. Just yesterday Fox News posted this piece which has “baked in” interlacing artifacts. This is especially ironic since Fox, ABC and ESPN record and broadcast in 720p/60: https://video.foxnews.com/v/5487990681001/?#sp=show-clips
-
[Craig Seeman] “Export either Quicktime Movie H.264 or DNxHD of HD raster (1080×1920) 59.94fps.”
The review on ProVideoCoalition says it can only export up to 720p not 1080 and posts screen caps of this: https://www.provideocoalition.com/media-composer-first-free-and-finally-shipping-after-all-these-years/
However the Avid comparison page says it can export up to 1080 resolution, just not 4k. They may have updated it after the above review: https://www.avid.com/media-composer-first/comparison#General
I can’t tell whether Media Composer One supports proxy mode but if it can’t do 4k there would be less need for this. It’s great for editors and educators to have another fairly full featured free option.
-
[Andrew Kimery] “….I’d be most interested in hearing about is what took place behind closed doors deep inside the bowels of Apple….pretty much everyone interviewed for the doc I’ve already read their books, seen their demos, attended their panels, listened to their pod casts, followed them on Twitter, etc….I’m not really that interested in watching a 90min recap of 6yrs worth of events that I already followed on a daily basis…”
These are good points. Brad said he wanted to answer “Why did Apple decide to make this change?”
Was it a focus group? Was it Randy Ubillos’ alleged difficulty organizing his own vacation videos? Were there influential academic white papers? Was there a “Xerox PARC visit” event? For the origins of the Mac, these have been well covered historically. The jump to FCPX was nearly as great a disruption.
Some of the commentary in the trailer is interesting, but it’s not very deep. Thus far the emphasis has been more about why are people resistant than what were Apple’s conceptual and philosophical drivers for the redesign.
I am concerned about statements like “case study of how creative individuals react to disruptive technological changes”, and “found myself in the minority as many other creative professionals felt abandoned, betrayed, and even belittled by Apple’s update”, and “led to a lot of misinformation that persists today”. This implies a motivation for advocacy which is understandable. However it may risk shaping the documentary toward a rehash of what we already know, vs an informational piece of more lasting historical value.
Tracy Kidder won a Pulitzer Prize for his 1981 novel Soul of a New Machine, about the inside story of designing a radical new computer. It is required reading in some educational curriculums today. If Off the Tracks is shaped more like this book and less like an observational study of “the debate”, it will probably have more lasting value and ultimately will be more informative.
-
[Jacqueline Nauman] “…Sometimes with the hub a drive will drop off for no reason. I think it’s probably too many working drives…”
If these are bus-powered USB drives, most hubs have significant current limit which is lower than the current limit most Macs have. Macs can supply higher current to charge iPads and Apple uses several USB current profiles, denoted 1A, 2A and 2.4A. You can have (for example) four bus-powered USB drives plugged into an iMac and they’ll work OK, but maybe not if plugged into some USB hubs, even if the port is externally powered. You never know since it depends on the hub, the drive, peak simultaneous current draw, etc.
This is a big issue on the 2016 and later MacBook Pro which only has USB-C, thus hubs are more common. Even the plug-in “hyper docks” which use two USB-C ports will not provide the USB-A current that previous Macs did, hence you can’t always run the multiple bus-powered USB 3 drives you did previously.
[Jacqueline Nauman] “…Is the Trash Can really worth it?….I need something to last and work hard with FCP X….Mostly my editing is corporate or documentary video. I have yet to do any 4k editing….not sure that I ever will but it would be nice to have the option….
I have a 12-core D700 nMP on my desk right now I’m testing against a 2015 top-spec iMac 27 in FCPX. If you use or plan on using any H264 codecs — including AVCHD, Sony XAVC-S, most drones, GoPros, most DSLRs, many camcorders, etc — I would definitely not get the Mac Pro. It is an aesthetically beautiful machine, very quiet and with good thermal control. However due to the Xeon CPU it does not have Quick Sync which is a major handicap dealing with H264. In my tests it is 1/2 as fast as my 2015 iMac 27 at transcoding 4k H264 to ProRes proxy. It is only a little faster on Neat Video noise reduction, no faster on Digital Anarchy flicker removal, and no faster on stabilization or sharpening. It is a little faster on some effects like Aged Film, and if using a ProRes timeline some of the above effects will render faster than an iMac.
On the 12-core nMP, multicam 4k H264 requires transcoding to proxy for smooth editing — just like an iMac. The only problem is the nMP transcodes to proxy at 1/2 the rate. In short the nMP is a great machine if you use all ProRes acquisition. In my doc team we have 12 GoPros, four drones, cable cams, crane cams, gimbal cams, 3-axis motion control cams, etc — all 4k H264. We have Atomos recorders but prefer not using those in the field unless absolutely required because of complexity and HDMI fragility. The Inspire 2 drone can shoot ProRes internally but this doesn’t help if everything else is H264. So the idea of “just use ProRes acquisition” may work on scripted narratives, short commercials and lower shooting ratios but it doesn’t work for us, thus the need for faster transcoding to proxy or a machine that can edit 4k H264 smoothly without transcoding.
For H264 1080p, almost any machine or editing software is fast enough. A recent iMac 27 would be significantly faster than your old Mac Pro for H264. In your position I’d probably get a 2017 iMac 27 of some kind.
-
[Robin S. Kurz] “I highly doubt that any mentionable performance differences will merely be due to the SSD as opposed to a Fusion drive by any relevant level. There are many many other specs that will play into that and make the real difference, whatever it may be.”
This is correct. I have a 2013 and 2015 iMac 27, one with 3TB Fusion Drive and one with 1TB SSD. Both have the top CPU and GPU and both have 32GB RAM. I have tested them extensively side-by-side and cannot tell any difference in I/O-related FCPX performance when the media is on an external drive. Re FCPX boot time, it’s usually about the same between the Fusion Drive and SSD iMac.
If media is on the boot drive there are some cases where SSD is faster but (a) You normally don’t want media on the boot drive, and (b) Even 1TB of SSD is too small to put much media on it.
Thus the dilemma is a 3TB Fusion Drive is big enough to put some media on it, but you shouldn’t do that, whereas the 1TB SSD is fast enough but not big enough. If all media is on external storage, either Fusion Drive or SSD will generally deliver about the same real-world FCPX performance. In that case I would mildly prefer SSD because it’s less complex and might help non-FCPX performance in some edge cases.
OTOH spending all your money on an SSD iMac then putting your media on a cheap, slow, bus-powered USB drive can often produce poor performance.
-
[Don Walker] ” Robin says that the choice of an i7 is not that critical over the i5. Peter says the i7 is crucial. Which is it? Is an FCPX editor really going to see much difference between the two, all other options being equal?”
The two options discussed were a quad-core 3.8 Ghz i5 vs 4.2 Ghz i7 for 4k editing in FCPX. The clock speed difference is about 11%. Benefit from i7 hyperthreading will vary from zero to significant. I have tested FCPX export performance with hyperthreading on vs off on the same i7, by using the 3rd party CPUSetter utility. Hyperthreading improved export performance by 30%, so the total i5 vs i7 performance difference would be about 40% — for the export case.
I didn’t test other FCPX CPU-bound operations such as transcoding, or CPU-bound effects such as stabilization, Neat Video, or Digital Anarchy Flicker Free. However the i7 would possibly help in those cases. Identifying these cases it easy — just use iStat Menus or Activity Monitor, and if all CPU cores are pegged, it’s not waiting on I/O or GPU. In all those situations more CPU horsepower (whether additional cores, higher clock speed, better IPC from a newer CPU or hyperthreading) would likely help. Whether it would be noticeable or worthwhile depends on the magnitude of improvement and how often those cases are encountered.
Obtaining smooth 4k editing is often difficult on almost any hardware. However this varies widely based on the codec. If the camera captures ProRes or has an external ProRes recorder, editing that is a lot easier. If it is H264 ( including containers such as XAVC-S), then it’s a lot harder. If you never edit multicam or apply any compute-intensive effects, it’s easier. Multicam and/or compute-intensive effects make 4k editing much harder. But if you are willing and able to transcode everything to proxy, it can be edited on a MacBook Air.
[Julius Jonas] “…when you would ONLY use Final Cut Pro X then you can “ignore” the processor, go for the best graphics card you can have…”
You definitely cannot “ignore” the CPU in FCPX. If so a quad-core 2013 Mac Pro with dual D700 GPUs would be as good on FCPX as a 12-core machine with the same GPUs.
-
[Daniel Hernández] “Sony A7SII (mp4 1080p XAVCS). After I import the files creating optimized media and proxies, the event starts working fine but after a while some of the files start showing the red ! Modified Files message. If I close the FCPX and reopen it they appeared fine but then some other files go to the same message.”
I have seen this a few times and am not sure what causes it. FCPX maintains internal metadata about attributes of each file, at the time that file was imported. This includes audio channel config, audio sample rate, codec, file size, etc. Upon opening a library it compares that stored metadata to the current file attributes on disk. If any of them differ it shows a red box and says the file was modified.
For a group of files it won’t tell you which attributes are the problem. However if you pick a single file and try to relink it, it will then tell you.
Re how to fix this, it may make sense to run Disk Utility First Aid on all drive volumes. BTW any drive with FCPX content should ideally be HFS+ formatted not exFAT or anything else. After that I’d suggest rebuilding Spotlight indexes, since FCPX uses those for file access. It is conceivable some cached or corrupt data in the indexes are causing the problem: https://support.apple.com/en-us/HT201716
Digital Rebellion has a utility called QT Edit which allows editing the file metadata. I’ve never used it but it supposedly can fix this problem in some cases: https://www.digitalrebellion.com/promedia/