Can't Export FCPX Project without Re-encoding?

For those finding this post looking for the ffmpeg solution I used this is how you do it:

First, figure out the desired start times for your a/v streams. I did this by syncing the clips in FCPX and cutting any unnecessary media at the beginning, and then by selecting the clips you can get your clip's "Start" value in the Info Inspector which we'll use in ffmpeg.

Next open up terminal and use the following command, replace any [VARIABLE] as needed:

ffmpeg —ss [VIDEO START TIME] -i [VIDEOFILENAME.EXT] —ss [AUDIO START TIME] -i [AUDIOFILENAME.EXT] -c copy [OUTPUTFILE.EXT]

• If you have never used/installed ffmpeg as a command line tool just google "homebrew", install that, and use it to install ffmpeg with the terminal command "brew install ffmpeg"
• Remember to use 'cd' to navigate to directory with source media or use absolute paths when defining input and output files
• Time format for seek (-ss) argument: HH:MM:SS.MS (hours:minutes:seconds.milliseconds)
• MS will be = (1000/framerate)*frames

And voila, no unnecessary intermediate transcoding to combine your offset a/v files.

Here's what it looks like:

Note if you have a video file that already has mixed a/v streams but out of sync audio there is a similar but different process for fixing that. It involves using ffmpegg's -itsoffset and -map arguments to selectively offset the start timestamp of just one of the streams (usually audio by ms as it gets you most accurate syncing).

Thank you for the help and quick response!

I mistook/assumed exporting uncompressed with source media that's already compressed wouldn't literally "un-compress" it while creating the output file but rather just write the existing media streams to a new file. I don't get why the GOP structure needs to be completely rewritten given that ffmpeg is able to do this mux operation just fine, even when trimming the beginning or end of either file. I'm assuming for the video it only re-encodes input stream data for any frames before the first and after the last i-frame that will still be in the output stream.

I guess my confusion is that QuickTime and FCPX are both Apple developed, and whatever QT did to encode my screen capture compressed ~90 mins of ~4K footage into a ~5GB file in less than a minute after ending the recording, yet FCPX can't use whatever encoding option QT used? My only export options are ProRes, H264, and Uncompressed; is there something I'm supposed to be installing to get more options? Better yet there should just be a "don't encode" option when rendering files that produces an export warning only if the project media has mixed codecs.

I have compressor and Adobe Media Encoder and all that installed my issue was just not wanting to unnecessarily transcode a bunch of times when the source media is already well compressed with great quality. Was able to do it with ffmpeg just took a bit of toying around to figure out the right arguments to give it.

Looked into this but even H264 was still making the file more than 10x the size from 5 to over 60GB. Source media was already well compressed with AVC I didn't want to have to re-encode a bunch so I just wound up using ffmpeg to mux the files instead.