Mac Studio and iMovie

[bogdanw]

Today (12 April 2022) iMovie has been updated (10.3.2) with improvements for Apple silicon.
I intend to buy a Mac Studio for iMovie editing and I would be interested in some real life tests.
If you can, please post screenshots of Activity Monitor with GPU & CPU history during iMovie exports. (720p/1080p/4K)
I’m not interested in FCP or other video editor, only iMovie.

 

[F-Train]

Today (12 April 2022) iMovie has been updated (10.3.2) with improvements for Apple silicon.
I intend to buy a Mac Studio for iMovie editing and I would be interested in some real life tests.
If you can, please post screenshots of Activity Monitor with GPU & CPU history during iMovie exports. (720p/1080p/4K)
I’m not interested in FCP or other video editor, only iMovie.

I don't use iMovie, but I assume that, like Final Cut Pro, it uses Compressor presets for export. If that's correct, I would think that the question is what the update does to improve Compressor.

 

[bogdanw]

I don't use iMovie, but I assume that, like Final Cut Pro, it uses Compressor presets for export. If that's correct, I would think that the question is what the update does to improve Compressor.

I don’t have time to compare the files (/Applications/iMovie.app/Contents/PlugIns/Compressor/CompressorKit.bundle/Contents/Frameworks/Compressor.framework/Versions/A/Frameworks/CompressorKit.framework/Versions/A/Resources/Settings/), but I don’t think that the presets for export are relevant.
I’m interested in what ProTranscoderTool_sb does. That’s why I’m asking for GPU & CPU history data to see if iMovie manages to use the resources to the max. There is no point in buying a 48-core GPU if it’s not properly used. I’ll just buy the 32-core model.

 

[eddie_ducking]

I'm kind of in the middle of testing other things on my base Studio Ultra at the moment, but give me a couple of days and sure .... I'm not an iMovie person, does it make any difference to the results from your perspective what the source and encoded files use for codex or have you a preference of H.264 over H.265 ? (I'm guessing it'll make a difference to encode time but possibly not to CPU/GPU utilisation). Never having used iMovie I might need a bit of hand holding unless it's defaults being used.

 

[Solenoid289]

I'm not a video guy but I've tested with latest iMovie 10.3.2 on M1 Max 32GB/24GPU/1TB

Here's a screenshot during exporting simple 4K project with iMovie (High Quality preset, H.264).
I'm from Japan, sorry if you can't understand the details.
Hope this helps.

 

[bogdanw]

I'm kind of in the middle of testing other things on my base Studio Ultra at the moment, but give me a couple of days and sure

Thank you, there is no rush. The source doesn’t matter. 1080p/High quality/fast setting for export.

I'm not a video guy but I've tested with latest iMovie 10.3.2 on M1 Max 32GB/24GPU/1TB

Thank you, it seems the GPU and CPU are not used to their full potential.

 

[eddie_ducking]

Thank you, there is no rush. The source doesn’t matter. 1080p/High quality/fast setting for export.

Here you are
Mac Studio Ultra 20/48/64Gb



iMovie 10.3.2


Input files used :-
bbb_sunflower_1080p_30fps_normal.mp4 (Duration 10min 34sec)
bbb_sunflower_2160p_30fps_normal.mp4 (Duration 10min 35sec)

Test 1:
1080p source, 1080p destination, High / Faster settings
Time taken: 1min 19sec



Test 2:
1080p source, 1080p destination, Highest (ProRes) / Better settings
Time taken: 32sec



Test 3:
2160p source, 1080p destination, High / Faster settings
Time taken: 1min 20sec


Test 4:
2160p source, 1080p destination, Highest (ProRes) / Better settings
Time taken: 39sec


Test 5:
2160p source, 2160p destination, High / Faster settings
Time taken: 3min 41sec


Test 6:
2160p source, 2160p destination, Highest (ProRes) / Better settings
Time taken: 56sec

 

[bogdanw]

Test 1:
1080p source, 1080p destination, High / Faster settings
Time taken: 1min 19sec
Test 2:
1080p source, 1080p destination, Highest (ProRes) / Better settings
Time taken: 32sec

Thank you very much for all the tests. Those are encouraging results.
Just to confirm, the “Best (ProRes)” tests completed faster than the “High” ones?

 

[eddie_ducking]

Thank you very much for all the tests. Those are encouraging results.
Just to confirm, the “Best (ProRes)” tests completed faster than the “High” ones?

Yes, in all cases. And quite significantly too. I guess that's the beauty of the hardware encoders cause the CPU usage really isn't that high at all.

I'm assuming hardware encoders were being used, I couldn't find anything to enable/disable them.

Also, I'm presently booting from a TB3 NVMe SSD to test MacOS 12.4 Beta. As such, my disk performance is less than half of the internal SSD ie. "only" 1800MB/sec or so. Can't imagine it'd have an effect on the "High" exports, but may have had a marginal one on the "Best (ProRes)" given the 41Gb size of the 4k->4k exported file.