[Joe Marler]

If you just do a CTRL-A in the Event Browser to select all, at the bottom of that pane will be the cumulative duration of all clips.

Other methods of getting various clip durations:

– Select clip OR range in browser, at bottom of browser is duration
– Select Project, at bottom is duration of entire project
– Select multiple projects in browser, at bottom is total duration
– Select multiple clips in browser (can be non-contiguous), at bottom is total duration
– At bottom of timeline, duration for current project is shown (no clicking required)
– In timeline, doing a range select (R tool) will show duration of range at bottom
– CTRL-D: shows duration of selected clip in timecode window

[Joe Marler]

[Brett Sherman] “…would really save me a ton of time editing, it’s obvious. Automated Speech Recognition. It would quite frankly be revolutionary for my editing process. The thing is, where is it? The technology is there….Yes I know it won’t be even 90% reliable at this point.”

I agree. The technology is apparently available for non-trained, speaker-independent automated transcription at a sufficient quality level for various purposes. My Comcast phone service has automated transcription of voice mail, and it works fairly well. https://www.comcast.com/readablevoicemail. The underlying patent for this technology is apparently stated here: https://www.google.com/patents/US6775651

Nuance also has voicemail-to-text; I don’t know if the underlying technology is shared with Comcast or not: https://www.nuance.com/for-business/mobile-solutions/voice-to-text-services/index.htm

Voicemail to text is speaker-independent automated transcription. It can obviously be done on longer input files.

At one time the web service https://www.voicebase.com/ did free automated transcription but it’s no longer free.

The previous Premiere Speech Analysis yielded generally poor accuracy for pure transcription. It was mainly designed to sync a written script with narrated audio, not stand-alone speech to text.

I tested Dragon Dictate 4 for Mac, which can transcribe pre-recorded audio files. It had better accuracy than VoiceBase (back when VoiceBase was free) but is designed for a single speaker and requires a brief training per voice. It also does not do auto-punctuation.

It would seem there’s a significant market for a stand-alone speaker-independent transcription product which works with pre-recorded audio and does not require training. For documentaries the conversion accuracy could be quite modest, yet still be useful. The fact that Comcast and Nuance already use this shows it is technically feasible.

I don’t know what the impediment is to implementing this more widely on personal computer platforms.

[Joe Marler]

I have imported a lot of UHD 4K XAVC-S from a Sony A7RII with no problems. That is on 10.2.2.

The easiest thing to try is delete and re-create your library, open only that one library and try again the import. First try import with “leave files in place” and de-select all other analysis and transcoding options.

If that doesn’t work trash the FCPX preferences by holding down the OPT & CMD keys when launching FCPX.

Preference Manager is a tool to assist with this: https://www.digitalrebellion.com/prefman/

[Joe Marler]

[Walter Soyka] “What’s the biggest bottleneck or frustration in your workflow? “

I’ll list deficiencies unique to PP CC and FCPX that slow me down. This may sound negative but it’s based on actual experience.

PP CC 2015: The playback engine is just too slow on H264 4K material on normal hardware. By “slow” I mean responsiveness to JKL commands in the timeline, and Program Monitor lag time when fast forwarding through the timeline. This is ironic since traditionally Mercury Playback was so fast. However the performance falls off a cliff on H264 4K. This is on a 2015 iMac 27, 32GB GB, 1TB SSD, M395X and Pegasus R4 Thunderbolt RAID. I have also tested on similar Windows machines. You could argue “it’s 4x the data, of course it’s slower”, but FCPX handles the same 4K material on the same machine much faster.

The playback engine apparently buffers around the current playhead position so after a while it becomes pretty quick making small movements — say using a jog wheel in the Source Monitor. But if you rapidly scrub down the timeline, it quickly exhausts that buffer and major lag appears — both in frame rate and response lag to JKL commands. This issue is further magnified since Premiere has no built-in proxy capability. In the 4K era it needs either a greatly reworked playback or built-in proxy or both.

FCPX: While lauded for great multicam, I find the deficiencies in FCPX’s multicam are hugely irritating and time-consuming to work around. It is amazing to me these are not more widely discussed.

E.g, you can’t stabilize or apply optical flow smoothing to a multicam clip. You can’t use a tracking plugin like SliceX on multicam. You can’t use Auditions on multicam. You can’t even use FCPX stabilization on a single-cam synchronized clip, although 3rd party Lock&Load stabilization works in that case.

The workaround for all the multicam limitations is cumbersome and error-prone. It involves back-tracking to the original clip, writing down the time code, then opening that clip separately, jumping to the written-down timecode, blading the range and applying stabilization/optical flow there, then going back to the multicam clip. Even after applying the effect, you cannot adjust the parameters in the multicam timeline. So getting the right result often involves multiple iterations of this for each clip.

Premiere CC has no problem applying stabilization or optical flow time remapping directly in the multicam timeline.

[Joe Marler]

[Stacy Wells] “If you guys have any pros/cons to offer concerning my decision, please let me know. Much obliged…”

It’s already been stated but just remember even RAID-5 must be backed up. You are safe from a single drive failure but other human, software and machine failure modes can jeopardize your data.

[Joe Marler]

[Ed Stevens] “I’m a wedding videographer, after editing on FCPX, I deliver my clients Blu-rays. I also offer standard definition dvds if they ask for it. I’d like to offer my clients the .mov master file also….I could provide them with a slim hard drive, formatted for a Mac OS Extended (journaled), loaded with their file(s). But most clients don’t have a Mac to read the hard drive…A 50 minute .mov video is about 30GB”

As optical media dies out this can be a difficult transition for smaller content providers. Increasingly customers use mobile devices which cannot even read an optical disc or (with Apple) a USB stick.

There’s also no universal streaming equivalent of the DVD/Blu-Ray presentation format, with splash screens, background music, menus, etc.

Yet another problem is the perceived loss of value from not having a physical product. Customers may be willing to pay more for a shrink-wrapped jewel case with nice artwork and a custom printed disc.

Even though some households have internet-enabled TVs or ability to AirPlay/Miracast from mobile devices, in general the communal viewing experience is not as readily available for streaming content from smaller providers.

But none of this matters because optical media is on a fast track to total obsolescence. Content providers must find some solution. Here are some possibilities (some already mentioned). Some of these assume you’ve already handled the transaction separately, others allow e-commerce.

(1) Use custom-printed USB sticks, possibly mounted in a jewel case or laminated gift card
(2) Host the content on line and give the customer a polished, artistic gift card with QR code and URL to the site.
(3) For Roku you can make your own private channel which the customer can view: https://developer.roku.com/home They don’t host the content but you can host it elsewhere (like Vimeo Pro)
(4) Use Vimeo Pro which allows presentation portfolios you can build without any programming: https://vimeo.com/pro This also allows you to optionally enable file download to your customer.
(5) Use Vimeo On Demand which allows ecommerce: https://tinyurl.com/zdd62yg
(6) Use unlisted or private Youtube video.

If you’ve already handled the transaction via other means you can give them an H264-encoded video file on USB devices or (depending on bandwidth, resolution and program length) send it via file transfer. My group is testing the Infinit.io peer-to-peer file transfer system. It has worked pretty well so far: https://infinit.io/

As already stated you don’t need to give them a ProRes file. A 1080p H264 file is likely sufficient, and for many devices and scenarios it’s hard to see the difference between that and 720p. In fact many broadcast and cable networks use exclusively 720p/60, including ABC, Fox, ESPN Networks, A&E Networks, and Disney Channels. Even if you watch those on a 1080p or 4k TV, you are actually seeing upscaled 720p content.

[Joe Marler]

[Phil Lowe] “Avid also provides a filmstrip view. I’ve never used it because, again, it affects performance. As someone else noted, a large project in FCPX will also suffer from performance issues having to refresh all those thumbnails and waveforms in real-time”

I just imported a 2.2 terabyte project to FCP X, which comprised 5,667 clips, 117 hours of mixed 1080p and 4k material. After import the Event Browser filmstrip view was a little sluggish until I scrolled to the bottom; after that it was fast.

I then connected my second iMac in Target Display Mode and configured FCPX to fill the entire second screen with thumbnails. It again got a little sluggish until I scrolled to the bottom, then it was fast as I scrolled up and down marking ranges.

In general I don’t see any major performance problem from refreshing thumbnails. This is on a 2015 iMac and an 8TB Thunderbolt RAID.

OTOH if you want to put all your unedited material on a timeline, FCP X can do that also. In this test on a Mac Pro it handled 1,600 lanes (ie tracks) and a timeline length of 13,000 hours: https://www.fcp.co/final-cut-pro/news/1372-pushing-apple-s-new-mac-pro-and-final-cut-pro-x-to-their-limits

[Joe Marler]

[Walter Soyka] “What’s the biggest bottleneck or frustration in your workflow? “

Good question. Some things we blame on the editing phase or try to fix with editing tools are actually planning and procedural issues in the pre-production and production phases.

E.g, with documentaries (except “Cinema Direct” where it doesn’t exist) you generally want to minimize talking head time, hence need dialog or narration to cover b-roll. If the shooting plan did not specifically capture targeted b-roll for each interview, this produces a huge time wastage as the editor fishes around looking for something usable which matches the dialog. Even worse is having a great visual and trying to find matching dialog, since scrubbing through audio is slower. Of course a transcript is nice — if available. Yes FCP X and some other tools let you do that fishing very fast and organize whatever was shot, but targeted shooting is much more efficient.

Another big frustration is when inattention or error during production causes time-consuming fixes in post. A common case is audio problems. Yes you can fix lots of that in Audition or Izotope but it takes extra time, sometimes lots of time.

[Joe Marler]

[James Culbertson] “[Walter Soyka] “If you really need top-line system speed, I don’t see how the nMP is a consideration. Is this the architecture you would want to build a high-end finishing system on?

…Macs and PCs have been leap-frogging each other for as long as I have been buying computers. If a nMP is not high enough powered to do the “high-end finishing” you need for your niche than buy something else. Next year perhaps a n(ewer)MP will be the high-end system you need to buy because it leap-frogs PCs…
“

Bingo. While impressive, high-end finishing is a tiny market niche. Yes it’s fun to see what Quantel’s Pablo Rio can do. There are PC manufacturers who will build you a 48-core, overclocked, liquid-cooled Xeon workstation. The now two-year-old nMP cannot really compete with those.

However IMO the nMP was never designed to occupy the absolute highest echelon of the workstation market. It is a generalized design. Chasing rarified niche market segments for “biggest/fastest” prestige often does not yield good business results. A good example is the supercomputer market segment. It is exotic and impressive but many companies have gone bankrupt pursuing that.

The next Mac Pro update will probably bump it up the performance ladder. When that is released it will be interesting to see how it compares to its upper-end competition.

[Joe Marler]

[Walter Soyka] “You can hit about 5 GB/s in a Z840 with an internal RAID of 16 SSDs…the fact remains that there is simply no way to get it on the nMP.
“
The nMP has three TB2 buses. Even if you leave one bus for a 4k display, you could theoretically put multiple large RAID arrays on the other buses. That could potentially be six Pegasus R8s or SSD RAIDs. AnandTech achieved 1.38 GB/sec on one TB2 bus of the nMP. That was an actual real-world test, not a theoretical calculation. That would imply that around 3 GB/sec total should be achievable. I wonder has anyone ever tested that?