[Joe Marler]

Your content is being transcoded at least once, and probably something is degrading the image. Do you have an optical drive so you can play the Blu-Ray disk on your Mac? That would verify whether the degradation is visible there or only on the big screen.

What is the FCP X output format that you’re feeding Toast? Can you can play that intermediate file to your projector (say using QuickTime or VLC on a MacBook) and verify it looks OK? Basically you want to see if it went wrong at the output of FCP X or during Toast processing. If the problem happens *before* Toast that’s also easier to experiment with, as you aren’t burning a Blu Ray disk every time and you can experiment with a shorter clip.

Toast has various bit rate and quality options. Did you pick some custom options or just “automatic encoding”?

[Joe Marler]

I know this doesn’t directly address your “trim unused media” issue, but have you tried Final Cut Library Manager? https://www.arcticwhiteness.com/finalcutlibrarymanager/

[Joe Marler]

Since encoding performance for web delivery is your top concern, evaluate whether the version of Sorenson Squeeze uses Intel’s Quick Sync on-chip transcoder. The latest version does but you must pick the right encoding format. Your 2013 iMac’s i7 has the CPU hardware, the issue is whether the software is using that.

Quick Sync is only available for single-pass MPEG-2 and H.264, but it is vastly faster than software or even GPU-accelerated encoding. In most cases for web-destined content, single-pass encoding may be good enough. In most cases I can’t see any difference.

This is a key issue because the Xeon CPU in Mac Pros do not have Quick Sync. A Mac Pro may do other things faster but it definitely will not do single-pass H.264 faster.

I think in general an 8-core Mac Pro with D700s is somewhat better for a full-time professional video editor but there are many types of workflow. If you are editing less and rendering more, then maybe the Quick Sync-equipped iMac would be better.

The 4Ghz retina iMac is about 15% faster from a CPU standpoint than your current iMac. It’s the same Hasell CPU generation, so the instructions per clock will be identical. The M295X GPU performance is debated but most benchmarks show it a modestly faster on some tests than the 780m in the 2013 iMac.

There is some debate about fan noise on the retina iMac when pushed really hard. Peak power consumption in the 2013 iMac i7 is 229w, vs 288w on the 4Ghz retina iMac, so the fan will probably run a bit more often. This isn’t an issue on the Mac Pro.

You already have an iMac that supports Quick Sync, so I’d suggest doing some tests with your current machine and Sorenson Squeeze. You want to evaluate whether you are already using it, or if not can you. If using it is the output quality OK? If you are using it, try multi-pass H.264 which will *not* use it and evaluate the performance difference.

[Joe Marler]

[Mark Morache] “Any faves here? We need a way for our show to receive large files (possibly a couple of gigs) from any number of people. “

I have paid versions of both WeTransfer and transferbigfiles.com. They both work well and are simple to use. Both allow users to send you files by uploading them to a URL (aka personal drop box) without logging in or using an app. I tried HighTail, but it was just too slow.

[Joe Marler]

Maybe these instructional videos will help: “Warp-Speed Key-Wording in FCP X”: https://www.youtube.com/watch?v=azJ4J41JaZk

Used Media Range Improvements: https://www.youtube.com/watch?v=aS4o0nVlGas

Power of the timeline index: https://www.youtube.com/watch?v=soN3NtPWW-E

Using the range tool: https://www.youtube.com/watch?v=oT8HuQ4gLZw

Smarter smart collections: https://www.youtube.com/watch?v=gLPnK_vvk0w

Some contents or functionalities here are not available due to your cookie preferences!

This happens because the functionality/content marked as “Google Youtube” uses cookies that you choosed to keep disabled. In order to view this content or use this functionality, please enable cookies: click here to open your cookie preferences.

[Joe Marler]

Here an experienced FCP X editor does a complete explanatory walk through of an actual production job. He emphasizes the metadata organizing features of FCP and need to organize material before editing. Maybe some of what he says would be beneficial.

Note he was using FCP 10.0.8, which is before the new library scheme. However most of what he says applies to the current version.

https://fcpxposure.com/video-editing-with-fcpx-at-the-aipp-2013/

[Joe Marler]

In general you want the AVCHD bundle to stay together and not extract .mts files. See this excellent instructional video on codecs. He also discusses the AVCHD situation: https://www.youtube.com/watch?v=WpBjGUlBTHU

In Mac OS, AVCHD .mts files will not normally display in Finder (or by pressing the Quick Look space bar). However this free utility will show the first frame of an .mts file as a thumbnail. I tested it and it works OK.

https://shedworx.com/hdquicklook

In general my group tries to avoid AVCHD, but where this isn’t possible we convert them using ClipWrap: https://www.divergentmedia.com/clipwrap This isn’t really necessary (see below) but if you have ingrained habits of organizing and selecting clips at the file level, this makes it easier.

Instead of renaming camera files, the best approach might be to import everything into FCP X and use the tagging/logging features to organize the material. I agree renaming files is a long established practice but it is very limited and cumbersome. FCP X has tremendous metadata management tools, that is the future.

However FCP X has recently integrated Finder tags, and here’s a video on that: https://www.youtube.com/watch?v=g175fguKnKw

Some contents or functionalities here are not available due to your cookie preferences!

This happens because the functionality/content marked as “Google Youtube” uses cookies that you choosed to keep disabled. In order to view this content or use this functionality, please enable cookies: click here to open your cookie preferences.

[Joe Marler]

I have the 8TB Thunderbolt G-Raid and it’s definitely not hot. The top, sides and bottom are mildly warm. There’s a small intake fan in the rear which runs continuously. You may need a flashlight to see it turning.

I wonder if the lower capacity units try to get by without a cooling fan, just using conductive transfer to the metal case?

It’s good to be suspicious about things like this because a failed cooling fan may cause eventual drive failure.

It’s a pretty good unit, well made and appearance is good. You can stack them vertically, which is nice. Acoustic noise isn’t very loud. Mine is in RAID 0, yet it’s not nearly as fast as my 8TB Pegasus R4 RAID 5, but the G-Raid is a good value for the money.

[Joe Marler]

[Noah Kadner] “Even a top end 2014 iMac doesn’t really come close for GPU heavy tasks like well pretty much everything FCPX and Resolve do…”

FCP X is very render-oriented. I am constantly looking at red render bars for almost any editing change. The good news is FCP does this smoothly in the background. The bad news is it happens frequently. By contrast Premiere does many of these instantly in hardware via the GPU — they appear as yellow render bars in the timeline. Yellow means GPU implemented.

Examining FCP X in iStat Menus shows these common rendering operations are CPU-bound, not GPU or disk bound. IOW FCP spins up a pool of worker threads which effectively harnesses all available cores. The more cores, the faster it gets done. It’s a good argument for 8-core and higher new Mac Pros.

It’s true sometimes I see lower CPU utilization and higher GPU. However the most common case is heavy CPU activity on all cores. In these cases no matter how fast the disk or GPU, you are waiting on the CPU.

Based on these observed behaviors, I don’t see how extreme GPU performance would help. I understand how it’s theoretically possible using OpenCL to further leverage the GPU, but I don’t see this happening very frequently in 10.1.3.

In my personal tests of a base nMP vs a max’d out 2013 iMac, I don’t see the superior nMP GPU performance translating to any huge real-world advantage on common FCP X editing operations. By contrast the iMac has a big advantage in single-pass H.264 export due to Quick Sync. I know that’s not Apple’s fault — it was an Intel design decision but that’s how it is.

I’d be interested in any further feedback or discussion on this point. Specifically I don’t see the nMP’s faster GPU translating to improved FCP X performance in many common editing operations.

2013 iMac 3.5 Ghz i7, 32GB, GTX780M, Thunderbolt Pegasus R4 8TB RAID 5

[Joe Marler]

[Robin S. Kurz] “Just curious: is there a particular reason why you chose to take the (usually, from experience, unnecessary and bug prone) detour through PE?”

It’s generally because the FCP X audio sync only works in simple cases. If it’s a simple two camera shoot and you know exactly what clips go together, yes X will sync that. I recently shot several takes of a musical event using four cameras and two additional audio recorders, and X just would not sync that.

Likewise I am sometimes handed a hard drive with hundreds of .wav files from several days of documentary shooting (sometimes with incorrect date/time stamps), and have to sync that with hundreds of video clips from multiple cameras. Plural Eyes does that quickly and without error. FCP X cannot remotely do that.

In an extreme case, I have used PE to search across many gigabytes of .wav files to look for a match to a video file, and it worked perfectly.