FCP11 keeps dropping frames for no apparent reason

This is a complex topic, but here’s likely what’s happening: There are two methods for recording video camera material intended for slow motion:

1. Higher frame rate recording: Each frame is recorded at a high frame rate (e.g., 120 fps), and the file metadata reflects that same frame rate. When imported into an NLE and slowed down (e.g., retimed to play back on a 30 fps timeline), the result is 25% speed or 4:1 slow motion. For example, a 1-second recording at 120 fps, if retimed to 25% speed in Final Cut Pro, would have a 4-second duration at 30 fps. On a 30 fps timeline, all frames of the original 120 fps clip would play back, providing smooth 4:1 slow motion.
2. "In-Camera Retimed" Slow Motion: Often referred to as “HFR mode” (Blackmagic), “S&Q mode” (Sony), or “High Frame Rate Movie” (Canon), this method samples the sensor at a high frame rate (e.g., 120 fps) but retimes the clip in-camera to a lower playback frame rate (e.g., 30 fps). The resulting file’s metadata reflects the lower frame rate. Using the same 1-second recording example, the out-of-camera clip has a 4-second duration at 30 fps, with "baked-in" 25% slow motion.

Why Manufacturers Use This Method:

1. Reduced Data Bandwidth: Retiming in-camera reduces the encoder's data bandwidth and the SD card’s performance requirements. For example, a 1-second recording at 120 fps does not need to write 120 fps per second but might spread the data over 2 seconds.
2. User Convenience: It provides users with an out-of-camera slow-motion clip without requiring retiming in a “complicated” NLE.

Downsides:

1. Non-Standard Terminology: Different manufacturers use inconsistent terminology, such as “Off-Speed” or “HFR” (Blackmagic), and include settings like “project frame rate” or “sensor frame rate.” Camera manuals rarely explain this well, and NLE documentation often doesn’t address it because it’s a varying camera-specific feature.
2. Accidental Activation: On Blackmagic cameras, for instance, the hardware "HFR" button is poorly placed, leading to accidental activation. This frequently results in footage that’s difficult to manage in post.
3. Metadata Issues and NLE Confusion: Non-standard clip metadata can confuse NLEs, producing unpredictable results. Traditional 120 fps clips can be retimed predictably in any NLE. However, with “baked-in” slow motion (e.g., 120 fps recorded as 30 fps), playback at 30 fps generally works fine, but attempting to retime such clips may cause erratic behavior due to the NLE’s rate-conforming algorithm.
4. Camera-Specific Limitations: Certain cameras impose restrictions in HFR mode, such as disabling audio or limiting codec options. For example, the Canon R3 only supports "All-Intra" in HFR mode.

Regarding Dropped Frames:

It’s important to clarify the term “dropped frame.” Dropped frames during playback differ from a persistently missing frame in a clip or frames dropped during encoding.

In Final Cut Pro (Settings > Playback), the default setting is “If a frame drops, stop playback and warn.” This is an aesthetic courtesy, not an indicator of a critical issue. Many users, including myself, disable this setting.

If the "dropped frame" warning is being triggered because the setting is enabled, simply disabling it resolves the issue.

The fact that a “dropped frame” warning now appears in FCP (when it didn’t previously) or that one NLE flags the issue while another does not might be caused by non-standard method combined with a metadata “kludge." We can investigate further if this issue proves significant enough to warrant time-consuming analysis by experienced volunteers.

Will that file play smoothly at 4x speed using Quicktime Player? If not, your "dropped frame" problem may have nothing to do with FCP.

I understand the file is on an external USB drive. Open the file in Quicktime Player with playback halted. Press the "L" key once; that is 1x playback speed. Press the "L" key a second time, that is 2x playback speed. Play the "L" key a third time; that is 4x playback speed. Is that smooth, or does it behave like FCP?

It's important to separate a media or decode issue from an effects issue. As a test, select the 120 fps HFR file in Event Browser (not on the timeline) and do the above 4x playback test. Is that smooth or sluggish as you described?

If the file itself shows sluggish playback at 4x speed in Quicktime Player, this isn't an FCP issue. If the file is smooth when played at 4x speed in the FCP Event Browser but slow if played at 4x speed in the timeline, it's possibly an effects issue.

To play a 25% speed (30 fps) clip at 4x speed is a significant I/O load. Keep in mind the Canon HFR recording uses only All-Intra, which is much less compression than H.264 or H.265. A 30 fps 4k H.264 clip might require 12 megabytes per sec at 1x playback rate.

By contrast, a 4k All-Intra clip may require 30 megabytes per second -- just to play at 30 fps. At 120 fps in FCP (required to obtain 1x subject motion on the slowed-down clip), that is 4x the data rate or about 120 megabytes per sec -- sustained.

USB SSDs can be very sensitive to cable type and hubs. It's easily possible the SSD is not delivering full-rated speed. Get the AmorphousDiskMark utility from the App Store. Select your external USB drive. Configure the utility to use 4 threads and 4 GB test file size. Compare the results to the attached file of me running those parameters on a Sandisk USB-C SSD, formatted APFS. You can also use Blackmagic Disk Speed Test, but that isn't as thorough. In the below tests SEQ1M = 1 megabyte sequential I/Os, QD = n means Queue Depth of n outstanding I/Os, RND4k = 4 KB random I/Os.

There is something seriously wrong with your Samsung T7 or the interconnect. I have the same drive -- just a ruggedized case, but the interior electronics are the same. See below results from the same test on my T7, which is about 67% full, formatted APFS.

The numbers on the right side are the achievable frame rate at a given resolution and codec. Of the three codec types listed in Disk Speed Test, your 4k Canon All-I codec is most similar to ProRes. The data rate you need for 1x normal speed playback is 30 fps slow motion at 4x forward speed (120 fps equivalent). A normal T7 can do that -- see my below results. Your drive can only do 11 fps at ProRes 422 HQ 2160 4k.

The Canon All-I codec used for HFR recording is actually somewhat less bitrate than ProRes 422 HQ, but it is roughly in the ballpark. If it's about 1/2 the bitrate of ProRes, that still only equates to about 22 fps on your T7.

The reason it can play at 120 fps for about two seconds, then lags is likely due to caching effects, either within the drive or within FCP, or both.

You can verify this by taking the HFR Canon file or a smaller clip of the same material, copying it to your M2 Max internal SSD, then importing it from that location to FCP and immediately trying a playback test. If it works on the internal SSD but not the external Samsung T7 with slow performance results, it's not an FCP issue.

Even though it says "not loaded", there are google things in there:

  [Not Loaded] com.google.GoogleUpdater.wake.plist (Google LLC - installed 2024-11-11)
    Command: ~/Library/Application Support/Google/GoogleUpdater/Current/GoogleUpdater.app/Contents/MacOS/GoogleUpdater --wake-all --enable-logging --vmodule=*/components/update_client/*=2,*/chrome/enterprise_companion/*=2,*/chrome/updater/*=2

  [Not Loaded] com.google.keystone.agent.plist (Not signed - installed 2024-05-10)
    <Empty>

  [Not Loaded] com.google.keystone.xpcservice.plist (Not signed - installed 2024-05-10)
    <Empty>

These should be completely removed.

Also: the samsung software should not be installed, your mac can use external drives just fine without manufacturer software, which often causes trouble.

MalwareBytes need not be running permanently.

The vpn stuff should probably go. Unless used as instructed as a tunnel into the work or school network, vpn should usually not be used. It is not the security thing that vendors purport it to be.

And finally, I see a crash of some PFS plugin. I and many around here have very bad perception of this company, both in terms of quality and business practices.

Can you clarify that you have used similar 120fps 4K footage in 30fps projects with no playback problems for some time and this has suddenly started?

Just check that you have selected View/Better Performance and not Better Quality.

If you create a new "test" project you should also use different clips as there may be a problem with specific footage.

Lastly, how many other apps and browser windows do you have open?

Could you share a clip using WeTransfer or Dropbox. If you post a link here it will appear in email notifications. Thanks.

Can you give complete specs of the media? What shot it? When did this start happening? Was it right after an update of the system or the application?

I have been puzzled by the terminology all along and assumed it was a typo.

What is footage shot at 120fps and recorded at 30fps? I thought recording and shooting were the same?

Does the camera somehow shoot 120 frames and interpolate(???) them down to 30fps? If so, what is the point?

Or do you mean you are putting 120fps media in a 30fps timeline which would have the same effect?

Or do you mean you were shooting 30fps at a shutter speed of one hundred and twentieth of a second?

Dropped Frames VS Missing Frames

I agree with joema that I always turn off the "Dropped Frame" (on playback) warnings. 99.999% of the time is a nuisance and a waste of time/energy. It means nothing as you can't really see the dropped frames on playback, anyway.

If you have "Missing Frames" physically absent in the original video file, that's going to be in your camera. Totally different subject.