[Joe Marler]

[Jenn Lindsay] “how I can come to better understand the difference between system frequencies, frame rates and bit rates and how they interact with shutter speed….but I’m getting really confused about what 29.97i or 23.98, or 30p refer to (these are the three frame rates [?]…I was shooting with the iphone occasionally (30p), and because a friend there recommended that since I was shooting AVCHD at 59.94 that 23.98 was better, so I changed from 29.97 about mid-shoot….)”

As first approximation you only need to remember a few things:

(1) Set shutter speed to 2x the frame rate. E.g, for 29.97 (or 30 fps) use 1/60th. For 25 fps use 1/50th sec.
(2) Normally the frame rate for your region will be a multiple of that region’s power line frequency. E.g, countries with 50Hz AC power usually have 25 fps video. Countries with 60Hz AC power usually have 29.97 fps video. Provided you do this and adhere to rule #1, it solves many problems.
(3) Do not change frame rates part way through a production. It is more difficult to match the “motion look” and cadence when cutting between those.

There is no direct interaction between data bit rates and frame rate. However some cameras only provide certain bit rates for a given codec, which may also constrain frame rates. There is usually a table in the back of each camera manual which lists the codec, bit rates and frame rates for each shooting mode.

The frame rate nomenclature is confusing since many cameras abbreviate 29.97 as 30 and 59.94 as 60 to save display space. The iPhone shoots true 30 fps not 29.97 but it’s easy to conform those since they are so close.

[Joe Marler]

[Jenn Lindsay] “C100 Mk II to shoot a bunch of footage and interviews in rooms with halogen lights. I saw the flicker in the monitor and did some on-site googling, and turned my shutter speed to 1/120 and put on the “clear scan”…I was shooting in Greece….My footage is shot at a 23.98 frame rate”

Greece’s utility grid is 50Hz, so your shutter speed should have been 1/100. That will take care of many problems. However I think Clear Scan might override this and allow dialing the shutter speed smoothly up and down. Depending on how that was set it could help or hurt.

If the lighting was halogen they were probably 50Hz. Increasingly a problem is LED lighting which uses Pulse Width Modulated (PWM) dimming that can cause problems at various shutter speeds. For those Clear Scan might be appropriate, assuming you have time on set to shoot test footage and review it.

To fix the problem after it’s happened is difficult. The only tool I’ve seen that fixes a wide range of flicker and banding problems is Digital Anarchy’s Flicker Free. It’s expensive but relative to the cost of re-shooting, it’s cheap. It works well but is very compute-intensive and slow to run. https://www.digitalanarchy.com/Flicker/main.html

[Joe Marler]

[Ken Bennett] “project that is using 4K 3840 x 2160, 59.94p and HD 1440 x 1080, 29.97i. When I create a Combined Clip, that clip becomes 1440 x 1080i. When I Share out an MP4 file, all my motion blurs have that ugly interlace line shift around the edges of my moving people in the shots….How can I change from interlace to progressive for my output?”

Your output is interlaced because your project is. Interlaced output is intended for playback on hardware or software players that deinterlace. E.g, if you played it back in VLC with deinterlace enabled, it would probably look much better.

If you want progressive output the project must be progressive. Your project may have been interlaced due to having an “auto select” project (which is the default) and dragging the interlaced clip as the first item. That makes the project interlaced. You can’t deinterlace interlaced content in an interlaced project. You must use a progressive project or change the project characteristics to progressive.

To change the characteristics, click on the project, then in Inspector click Modify.

To create a project that is progressive to which you’ll be adding both interlaced and non-interlaced content, File>New Project, click the “Use Custom Settings” button and select your desired parameters.

In some cases (depending on media characteristics) interlaced content added to a progressive project may require deinterlacing. In other cases FCPX does it automatically. To manually deinterlace, select the clip in the timeline, the in Inspector at the bottom 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 which FCPX does not automatically deinterlace this can help.

Assessing the material visually is not always reliable because you never know the behavior of the playback tool. E.g, FCPX will often deinterlace in the Viewer, but Quicktime Player may or may not, depending on the clip metadata. VLC by default does not deinterlace but has various configurable deinterlace modes. The most reliable method is manually inspect the clip metadata using a tool like Invisor, which can also do spreadsheet-style side-by-side comparisons: https://itunes.apple.com/us/app/invisor-media-file-inspector/id442947586?mt=12

[Joe Marler]

I used this for years on Windows — it works very well. IMO it may be the best stabilizer out there, but I haven’t tested it back to back with the various Mac alternatives. I look forward to having it on macOS.

[Joe Marler]

[David Mathis] “Currently looking to upgrade the GPU on MaC Pro for use with Resolve along side FCPX as well.”

For FCPX, AMD GPUs often provide better performance. But since you are also using Resolve that makes the decision more complicated.

https://9to5mac.com/2017/05/10/hackintosh-amd-rx-480-vs-nvidia-1080-ti-final-cut-pro-x/

[Joe Marler]

[Lucy Slavinsky] “All of the footage is from Youtube or other online sources. I am looking for the fastest way to edit. Quality of the footage is not important, we will buy needed footage later in the original format. There are about 15 hours of footage. What transcoding should I use when I import the files? “

If this is 1080p or 720p content from Youtube or other online source, IMO you don’t need to transcode to proxy or anything else. On a halfway decent machine, e.g. 2012 iMac 27, 2013 MacBook Air, etc, FCPX is fast enough to edit that natively.

You can even import it with “leave files in place” which further accelerates the import and reduces disk consumption. Just make sure neither “optimized media” nor “proxy media” are selected during import and those are also deselected in FCPX preferences>import.

If it is 4k you may need to transcode it to proxy for best editing performance.

However there is no automated way to “match back” clips on a timeline to the original content you plan on later acquiring for the final version. Therefore I suggest you use the Youtube version for a very rough assembly and don’t use any effects or transitions. That will be a small fraction of the original 15 hr. Then acquire the original format, and manually replace each clip, after which you can continue the edit, add transitions, effects, correct color/exposure, etc.

OTOH, if you add effects and color/exposure correction to the Youtube version, this could help non-editors better visualize the potential quality. You might be able to cut/paste those effects to the original content, but you never know how different the image and color characteristics will be. So that might not work.

[Joe Marler]

[chris walker] “I had asked earlier how to speed up the final export of a project, using an underpowered computer, when using proxies to edit 4k material on a 1080p timeline. “

For reference, that was this thread: https://forums.creativecow.net/thread/344/46986#46991

[chris walker] “One solution was to periodically switch the view to optimized/original and render out already edited portions so that less transcoding has to be done upon final export. The problem with that, though, is that many of my render files keep disappearing!”

By “render out” you mean render the timeline section with CTRL-R. What you describe is not the render files disappearing (although it looks like that) but they have been invalidated by some operation or action. The render files are a cache of rendered timeline content. Like any cache they can be invalidated by various events. This can be related to certain effect, but FCPX’s decision to re-render is not under control of the effect code. IOW FCPX is responsible for data integrity and it decides the render cache no longer accurately represents the timeline and must be rebuilt. It is very common to see cases where some seemingly innocuous action causes the render bar to go orange. It’s conceivable some of these might be over-conservatism on the part of FCPX and maybe given further algorithmic refinement it could be improved, but that is only speculation.

On my machine the render cache is not invalidated by restarting FCPX — for straight video with no effects, or just built-in color correction. But adding stabilization then rendering the timeline, then restarting FCPX will invalidate the cache necessitating a re-render. I don’t know why that is. Nothing changed on the timeline. Maybe they are using some in-memory cache for certain effects that is not backed by disk storage.

I understand your situation but I don’t think you can totally rely on the render cache being 100% predictable. From my previous tests it seems that exporting 4k H264 content from a 1080p timeline to 1080p H264 was about 25% faster than exporting to 1080p ProRes, at least on my 2015 iMac 27. Of course if you are facing a 48 hour export that might not be sufficient even if H264 was acceptable.

I recollect you were exporting 80 min. of 4k material from a 1080p timeline to 1080p ProRes. Are you saying without the render cache, that takes 48 hours? My 2015 iMac 27 would take about 33 minutes to do that, which implies it’s not your machine but possibly computationally-intensive effects on that timeline. E.g, if I used stabilization, Neat Video and Digital Anarchy’s Flicker Free on 80 min. of 4k it might also take 48 hr.

Is it possible you can do a test export with all effects removed to see what % of the problem is from effects vs the basic video? Starting with 10.3.x, there is a new feature to remove all effects from all clips. So you can duplicate the project, then remove all effects from the duplicate then add them back selectively: 

https://support.apple.com/kb/PH12615?locale=en_US

You previously indicated that using 80 min of 1080p content the export time was “reasonable”. I’d expect 4k to be at least 4x your previous export time — if all other conditions were maintained constant. It’s conceivable some effect is having a non-linear performance degradation at the higher resolution. You’ll have to test this to isolate it.

[Joe Marler]

[Larry Brill] “…shooting a number of different shots in a school classroom. Most captured fine (start/stop) but a few times I hit the record button, it started to record and then stopped. I got a warning message: “recording stopped” with a little hour glass animation. It did that for a few seconds before returning to normal….”

We’ve never had that happen on our DVX200. We only use Lexar 256GB 1000x cards, and we only record UHD 29.97p at 100 Mbps. Maybe your card is intermittent?

[Larry Brill] “….I can open those video clips directly from the SD card in both Quicktime and Photos on the iMac.
So the problem seems to be getting them directly from the SD card into FCPx. Oh, and if I right-click the package contents of the library, the imported video that doesn’t show thumbnails, is there and will play fine when you expand and play by hitting the space bar….”

OTOH this could indicate a problem with FCPX, the library itself or related cache or plist files. I suggest (1) Trashing FCPX user preferences, then (2) Creating a test library and importing the problematic content from the SD card to the test library. See item 11 on this page: https://support.apple.com/en-us/HT203477

[Joe Marler]

[Michael Hancock] “If I add metadata to EventA (notes, scene info, favorites, etc…), can I transfer and merge that with EventA on the other editor’s system, without having duplicate clips added to their event? “

Yes this can be done by exporting the event XML then importing it on the destination machine. Select the event, then do File>Export XML. On the destination machine, do File>Import>XML. At import time it will give you the option to overwrite the existing metadata or create duplicate clips. There is no intra-clip overlay-type metadata merge. It either uses the source machine’s metadata or it makes duplicate clips, preserving both original and new metadata. See https://support.apple.com/kb/PH12654?locale=en_US

You should definitely make a backup copy of the library before doing this.

Even if it seems to work, give the library a close look and verify several clips work correctly, the metadata works as expected and the clips are in the correct event. If you have other events, check that those were not changed. You don’t want to load the XML, believe it’s OK, then commit lots of work only to find out something went wrong that you didn’t spot.

[Joe Marler]

[Larry Brill] “FCP 10.2.3 importing video from Panasonic DVX 200….Suddenly…When I import from the video card into iMac running Yosemite (plugged into back of iMac) The video imports but when I select a clip, the clip browser shows either a weird blue icon boxes where the clip should be (not the video) or a freeze frame of the video. In freeze frame I can get audio, but no audio with the weird blue icon….screen shot…I tried off-loading the clips from the video card directly to an external backup and import from there. It worked for a couple but not for others…Help?”

My team has a DVX200 (great camera) and I’ve imported huge amounts of UHD 29.97/p .MP4 to FCPX 10.2x and 10.3x without any problem. I normally copy to disk then import from there with “leave files in place”. Are you using AVCHD? Is it possible your hard drive is out of space or the SD card is failing? Can you play the problematic video files using Quicktime or VLC? If not will those files play from the camera? If they will play from the camera and from Quicktime/VLC but you can’t import them all, can you temporarily close your library, create a test library and try to import to that?