FORUMS: list search recent posts

FCPX performance - codecs and external IO hardware

COW Forums : Apple Final Cut Pro X Debates

<< PREVIOUS   •   VIEW ALL   •   PRINT   •   NEXT >>
Oliver Peters
FCPX performance - codecs and external IO hardware
on Mar 6, 2017 at 11:39:11 pm

I've complained a number of times about JKL play performance in FCPX. Over the past 6 years, I have run FCPX on a wide variety of Macs, including 2009/2010 cheese graters, newer MBPs and iMacs and now a new 2013 trash can Mac Pro. What I've generally found is that across the board, if I'm working with ProRes, performance is OK. If I work with external i/o turned off and/or not installed, performance is OK. When I work with external i/o engaged, especially with non-ProRes codecs (C300, DSLRs, XDCAM, RED, etc.) performance is awful. For example, skimming is fine, but fast reverse play causes the viewer to freeze up until I stop. The same media on the same hardware works better in Premiere.

This leads me to believe that FCPX is by and large ONLY optimized for ProRes, even though other codecs show up as "optimized". I know many of you will dispute this and talk about wonderful performance. So for you, I would like to ask several things.

1) Are you optimizing media first in FCPX?
2) Are you generating proxies first in FCPX?
3) Are you monitoring through external i/o hardware?
4) Are you working with only ProRes files?
5) If you are running other camera formats with external I/O and you are NOT optimizing or creating proxies, what is your configurations? And do you use JKL, or just skim and play?

Just curious.

- Oliver

Oliver Peters Post Production Services, LLC
Orlando, FL
http://www.oliverpeters.com


Return to posts index

Gabe Strong
Re: FCPX performance - codecs and external IO hardware
on Mar 7, 2017 at 12:06:43 am

For me, no optimizing media or generating proxies.
Monitoring via HDMI on my 2013 MacBook Pro and
via my 980Ti (6GB) GPU (flashed for Mac) on my 2009
(flashed to 2010) MacPro. In both cases I go to a
42 inch Vizio screen for full screen monitoring.
And in both cases its way, way, way, WAY faster than
trying to do the same thing in Premiere Pro.
But I'm not using hugely complex video files, just AVCHD
from FS700/100, and VG20, some XAVC from mirrorless A6000 series,
some GoPro footageand so on.

Gabe Strong
G-Force Productions
http://www.gforcevideo.com


Return to posts index

Oliver Peters
Re: FCPX performance - codecs and external IO hardware
on Mar 7, 2017 at 12:15:03 am

[Gabe Strong] "Monitoring via HDMI on my 2013 MacBook Pro and"

So no external io hardware. Yes, I see good performance with HDMI, too. But that ONLY works well for FCPX.

Oliver

Oliver Peters Post Production Services, LLC
Orlando, FL
http://www.oliverpeters.com


Return to posts index


Gabe Strong
Re: FCPX performance - codecs and external IO hardware
on Mar 7, 2017 at 4:50:59 pm

[Oliver Peters] "So no external io hardware. Yes, I see good performance with HDMI, too. But that ONLY works well for FCPX.
"


Sorry, I was thinking of the external monitor as 'external hardware'. I see what you
are saying here, a different 'external output'.

Gabe Strong
G-Force Productions
http://www.gforcevideo.com


Return to posts index

Joe Marler
Re: FCPX performance - codecs and external IO hardware
on Mar 7, 2017 at 12:28:24 am

[Oliver Peters] "...complained a number of times about JKL play performance in FCPX...What I've generally found is that across the board, if I'm working with ProRes, performance is OK. If I work with external i/o turned off and/or not installed, performance is OK. When I work with external i/o engaged, especially with non-ProRes codecs (C300, DSLRs, XDCAM, RED, etc.) performance is awful...."

I've tested H264 4k editing using Premiere CC vs FCPX many times back-to-back on a top-spec 2015 iMac 27. I use JKL heavily. In general FCPX is much faster without proxy or optimized media -- on this machine, and when not using external I/O. However it seems Premiere has gotten faster within the past year or so, but it's still significantly less responsive, especially to JKL. Using proxy they are both very fast on H264 4k.

I rarely use external I/O, and when I do it's only Thunderbolt to another iMac. It's interesting your case seems mostly confined to external I/O. Are other FCPX users with your I/O hardware and device driver also reporting problems? Can you test it with different I/O hardware?

I'd be frustrated if FCPX was slow. In a sense Apple has less leeway on this than Adobe, since a core design principle of FCPX is slippery-fast speed. You are talking mostly about timeline performance but if the Event Browser got sluggish that would deactivate one of the main reasons for using FCPX.


Return to posts index

Oliver Peters
Re: FCPX performance - codecs and external IO hardware
on Mar 7, 2017 at 12:50:47 am

[Joe Marler] ". Are other FCPX users with your I/O hardware and device driver also reporting problems? Can you test it with different I/O hardware"

My experience includes a Kona LHi, BMD Mini Monitor card and a BMD UltraStudio Express. The first two are cards and generally the Kona is best. However, there I'm mainly running Proxies from RED camera files. So, I'm typically never running RED natively, except at the last step.

The UltraStudio is the worst with X, because this is also on a MacPro via Thunderbolt. It has to live on a shared port and reacts the worst that way. If I put it on a bus by itself, it's better but still not as good as with Premiere or Resolve. The really odd thing with X is that the Viewer in the UI is affected. Not just the video output through the device.

At home I'm running a 2009 with the Mini Monitor card. I usually don't have a video monitor connected. I noticed that even there, when I turn off a/v output in X, the performance improves.

Oliver

Oliver Peters Post Production Services, LLC
Orlando, FL
http://www.oliverpeters.com


Return to posts index


Dave Jenkins
Re: FCPX performance - codecs and external IO hardware
on Mar 7, 2017 at 4:08:06 pm

[Oliver Peters] "The UltraStudio is the worst with X, because this is also on a MacPro via Thunderbolt. It has to live on a shared port and reacts the worst that way. If I put it on a bus by itself, it's better but still not as good as with Premiere or Resolve. The really odd thing with X is that the Viewer in the UI is affected. Not just the video output through the device."

I have the same problem. I dumped the BMD products because of this. AJA T-TAP is a little better but I often get freezes on my I/O monitor and the FCPx viewer keeps playing using JKL.

Dajen Productions, Santa Barbara, CA
Mac Pro 3.5MHz 6-Core Late 2013
FCP X


Return to posts index

Oliver Peters
Re: FCPX performance - codecs and external IO hardware
on Mar 7, 2017 at 4:16:18 pm

[Dave Jenkins] "I have the same problem. I dumped the BMD products because of this. AJA T-TAP is a little better but I often get freezes on my I/O monitor and the FCPx viewer keeps playing using JKL."

Yes, AJA is a bit better. Right now, the only acceptable performance that I'm seeing (unless I'm using PR Proxy) is to either turn off a/v output or use HDMI. As a quick test this morning, I connected HDMI to a large Panasonic flat panel. This is connected by way of a CalDigit dock on one of the Thunderbolt ports that I have on a 2013 MacPro. Performance is good and no hang-ups on JKL.

I could use HDMI direct from the Mac, but then I'd have to give up the second external Mac display. Using the dock, I get all three. But, by doing this, I lose the color critical monitor connected via SDI when I'm in X. Of course, if I use AJA, then I have to swap between AJA and BMD every time I need to use Resolve. Argh!

- Oliver

Oliver Peters Post Production Services, LLC
Orlando, FL
http://www.oliverpeters.com


Return to posts index

Brian Seegmiller
Re: FCPX performance - codecs and external IO hardware
on Mar 8, 2017 at 5:43:27 pm

Could you use a thunderbolt to HDMI adapter? We connected a Ninja Blade to HDMI to record the screen and lost my second monitor. When I used the adapter I had 3 monitors including the Ninja. This is from a 2013 mac pro.


Return to posts index


Oliver Peters
Re: FCPX performance - codecs and external IO hardware
on Mar 8, 2017 at 5:59:24 pm

[Brian Seegmiller] "Could you use a thunderbolt to HDMI adapter?"

That function is what the CalDigit dock that I was using provides. It's also what the BMD UltraStudio does, but of course, drivers certainly affects its function. The dock is a pass-through.

- Oliver

Oliver Peters Post Production Services, LLC
Orlando, FL
http://www.oliverpeters.com


Return to posts index

Steve King
Re: FCPX performance - codecs and external IO hardware
on Jun 21, 2017 at 1:45:15 pm

Tom
I was thinking about getting the Caldigit 2 dock. Have you been happy with it? My application was to go from my MBP 2011 TBolt first edition to Caldigit 2 dock. Esata out to an SSD for capture ProRes LT from a BM Ultrastudio mini recorder that is connected from the Tbolt loop through on the Caldigit. Use the HDMI out to feed a BM HDMI to SDI adapter to feed a broadcast switcher for quick playback for live event. Your thoughts? 1080i 59.94

2011 MBP has a 1TB SSD for system drive but only a single first get Tbolt port...arrgh

Steve King
AVS Media Group


Return to posts index

andy patterson
Re: FCPX performance - codecs and external IO hardware
on Mar 7, 2017 at 9:00:49 pm

[Dave Jenkins] "I have the same problem. I dumped the BMD products because of this. AJA T-TAP is a little better but I often get freezes on my I/O monitor and the FCPx viewer keeps playing using JKL."

It seems like you would want to contact BMD and Apple both to correct the problem.


Return to posts index


Oliver Peters
Re: FCPX performance - codecs and external IO hardware
on Mar 7, 2017 at 10:56:54 pm

[andy patterson] "It seems like you would want to contact BMD and Apple both to correct the problem."

That won't fix the problem in the short term.

- Oliver

Oliver Peters Post Production Services, LLC
Orlando, FL
http://www.oliverpeters.com


Return to posts index

Helge Tjelta
Re: FCPX performance - codecs and external IO hardware
on Mar 7, 2017 at 5:00:03 pm

DNxHD and ProRes are almost equal in performance. They are both heavy multithread and works very well.

ProRes is so efficient that besides DNxHD, it is almost much better to transcode rather than use native codecs throughout the editing/grade/online/mastering.

i.e. Resolve uses 5-6 times longer for grade/output of a H.264 than ProRes (I would guess DNxHD is the same as ProRes).

The point is, not all codecs support segmentation or multi-core use. And some programs utilise this very much, hence FCPX is so insane efficient with ProRes.

just my guess :)

Helge Tjelta
Creative apps for the FCPX ecosystem.


Return to posts index

andy patterson
Re: FCPX performance - codecs and external IO hardware
on Mar 7, 2017 at 9:03:31 pm

[Helge Tjelta] "The point is, not all codecs support segmentation or multi-core use. And some programs utilise this very much, hence FCPX is so insane efficient with ProRes."

But FCPX shouldn't take a performance hit when using 3rd party I/O gear. Something is wrong with FCPX's real-time playback engine.


Return to posts index


Oliver Peters
Re: FCPX performance - codecs and external IO hardware
on Mar 7, 2017 at 10:56:23 pm

[andy patterson] "But FCPX shouldn't take a performance hit when using 3rd party I/O gear. Something is wrong with FCPX's real-time playback engine."

I think a lot more is being done internally by FCPX when presenting a signal to the device than other apps do. That's likely part of the issue here. For example, I have 4K video in a 4K timeline, but the UltraStudio is a 1080 HD device. FCPX takes care of the scaling to send a full frame image out to my 1080 monitors. The same situation using Premiere Pro results in a black screen. In Resolve I have to set the output of the timeline to 1080 in order to see full screen video. Otherwise I see a 1080 window of the 4K image.

- Oliver

Oliver Peters Post Production Services, LLC
Orlando, FL
http://www.oliverpeters.com


Return to posts index

andy patterson
Re: FCPX performance - codecs and external IO hardware
on Mar 8, 2017 at 1:12:27 am

[Oliver Peters] "I think a lot more is being done internally by FCPX when presenting a signal to the device than other apps do. That's likely part of the issue here. For example, I have 4K video in a 4K timeline, but the UltraStudio is a 1080 HD device. FCPX takes care of the scaling to send a full frame image out to my 1080 monitors. The same situation using Premiere Pro results in a black screen. In Resolve I have to set the output of the timeline to 1080 in order to see full screen video. Otherwise I see a 1080 window of the 4K image."

You should not get a black screen because Premiere Pro does the exact same thing with 4K timelines to a 1080P monitor using the Intensity Shuttle. There is no real-time performance loss when using Premiere Pro on a PC. You have to go into the settings to scale down. Watch my video link below at 5:30. Premiere Pro can out put 4K timelines to IEEE DV converter as well. Let me know if it works for your Mac after watching the video.







Return to posts index

Oliver Peters
Re: FCPX performance - codecs and external IO hardware
on Mar 8, 2017 at 2:18:33 pm

[andy patterson] "Watch my video link below at 5:30. Premiere Pro can out put 4K timelines to IEEE DV converter as well. Let me know if it works for your Mac after watching the video. "

You are correct. The "scale down" in the set-up menu fixes that. Thanks.

- Oliver

Oliver Peters Post Production Services, LLC
Orlando, FL
http://www.oliverpeters.com


Return to posts index


andy patterson
Re: FCPX performance - codecs and external IO hardware
on Mar 8, 2017 at 7:46:39 pm

[Oliver Peters] "You are correct. The "scale down" in the set-up menu fixes that. Thanks."

Thanks for letting me know it worked.


Return to posts index

andy patterson
Re: FCPX performance - codecs and external IO hardware
on Mar 7, 2017 at 8:59:17 pm

I don't doubt FCPX does handle Pro Res better than Premiere Pro. It is interesting that when using 3rd party I/O gear FCPX chokes up on your systems. My PC takes not performance hit when using the Blackmagic Design Intensity Shuttle USB 3.0 version?


Return to posts index

Tom Sefton
Re: FCPX performance - codecs and external IO hardware
on Mar 7, 2017 at 11:29:06 pm

New Mac Pro, 8 core, 64GB Ram, d700 GPU and a blackmagic ultrastudio 4k via thunderbolt 2, all media on gtech studio xl at raid 5 via thunderbolt 2. Using Red epic-w raw media in 4K, performance is ok in FCPX. JKL is sluggish and stutters. same footage as prores LT, performance is snappy and quick. JKL is ok, but not as good as via hdmi. Anything above 4K, performance with Red footage slows to a crawl. With prores, audio sync via playback out of a separate USB audio I/O device is fine.

In premiere cc, Red media doesn't play at realtime via same device. Frame rate suffers unless quality is dropped to 1/8, which leaves it almost useless as images are blocky and poor. With prores, performance improves but not as snappy as FCPX. JKL with prores has stuttering performance in reverse. With premiere, audio sync is lost - weird and frustrating - means we can't output a 10bit image with audio sync via an external audio device in premiere. Hdmi is much better.

I/O performance needs to be better via thunderbolt with raw, and with prores - agreed. Perhaps expectedly, best performance is found with davinci, especially when you customise your Red debayer settings.

Co-owner at Pollen Studio
http://www.pollenstudio.co.uk


Return to posts index

Erik Lindahl
Re: FCPX performance - codecs and external IO hardware
on Mar 8, 2017 at 9:05:38 pm

I can confirm the issue of FCPX performance when using AV-output via a video interface (BMD / AJA). No machine, no software or OS-version makes any difference for me. It's the number one issue I've had with the app since day 1.

I talked to one of the guys behind "that shared storeage" for FCPX at IBC last year and he confirmed my issue also. Even one of the presenters there editing a feature in FCPX noted "outstanding performance" (as long as AV-output was turned off).

It's really crap and defiantly is a deal breaker for me. It astounds me they haven't fixed it by now as the FCPX engine aside from this massive omission is an amazing piece of technology.


Return to posts index

andy patterson
Re: FCPX performance - codecs and external IO hardware
on Mar 8, 2017 at 9:31:48 pm

[Erik Lindahl] "I can confirm the issue of FCPX performance when using AV-output via a video interface (BMD / AJA). No machine, no software or OS-version makes any difference for me. It's the number one issue I've had with the app since day 1.

I talked to one of the guys behind "that shared storeage" for FCPX at IBC last year and he confirmed my issue also. Even one of the presenters there editing a feature in FCPX noted "outstanding performance" (as long as AV-output was turned off).

It's really crap and defiantly is a deal breaker for me. It astounds me they haven't fixed it by now as the FCPX engine aside from this massive omission is an amazing piece of technology."


I think the reason might be that FCXP was originally never intended to be used with broadcast equipment. Perhaps there is nothing that can be done other than a total rewrite of the program. Does FCPX take a performance hit if you use a second computer monitor using HDMI out for client previews? Just curious.


Return to posts index

Oliver Peters
Re: FCPX performance - codecs and external IO hardware
on Mar 9, 2017 at 12:33:19 am

[andy patterson] "Does FCPX take a performance hit if you use a second computer monitor using HDMI out for client previews? Just curious"

No. In fact I can run 2 displays via Thunderbolt/MiniDisplayPort and still run a 3rd display as HDMI via a Thunderbolt dock. No performance hit that I can tell.

- Oliver

Oliver Peters Post Production Services, LLC
Orlando, FL
http://www.oliverpeters.com


Return to posts index

andy patterson
Re: FCPX performance - codecs and external IO hardware
on Mar 9, 2017 at 1:56:10 am

[Oliver Peters] "No. In fact I can run 2 displays via Thunderbolt/MiniDisplayPort and still run a 3rd display as HDMI via a Thunderbolt dock. No performance hit that I can tell."

Thanks for the info. I don't doubt you can run three monitors with FCPX but is one of the monitors being used as a client/Preview monitor as opposed to just GUI screen real estate for FCPX?


Return to posts index

Oliver Peters
Re: FCPX performance - codecs and external IO hardware
on Mar 9, 2017 at 1:58:28 pm

[andy patterson] "is one of the monitors being used as a client/Preview monitor as opposed to just GUI screen real estate for FCPX?"

When you are not in FCPX, all 3 monitors are the macOS GUI. When you are in FCPX, the HDMI-connected display is the video output for client/preview. It's not simply the viewer blown up to fill the screen. No different than going through an AJA or BMD device. You can also enable the same function in Premiere.

- Oliver

Oliver Peters Post Production Services, LLC
Orlando, FL
http://www.oliverpeters.com


Return to posts index

Dave Jenkins
Re: FCPX performance - codecs and external IO hardware
on Mar 9, 2017 at 3:29:58 pm

[Oliver Peters] "It's not simply the viewer blown up to fill the screen. No different than going through an AJA or BMD device. You can also enable the same function in Premiere."

Except you don't get the same color spectrum (Rec709) when using the HDMI output as opposed to an AJA I/o device.
I can't use the HDMI for critical viewing or color correction.

Dajen Productions, Santa Barbara, CA
Mac Pro 3.5MHz 6-Core Late 2013
FCP X


Return to posts index

Oliver Peters
Re: FCPX performance - codecs and external IO hardware
on Mar 9, 2017 at 3:39:20 pm

[Dave Jenkins] "Except you don't get the same color spectrum (Rec709) when using the HDMI output as opposed to an AJA I/o device.
I can't use the HDMI for critical viewing or color correction."


I've heard that said, and I would tend to agree. But honestly, I don't know the correct answer as it applies to FCPX. Plus, when you see companies like Flanders offering HDMI, it tends to cloud the waters. Especially now that Rec709 is no longer a universal target.

Given that the majority of work that most of us do is for outlets other than broadcast, I have to question whether or not Rec709 accuracy is of paramount concern any longer. Especially when you are performing editing and not necessarily final color correction, for which FCPX isn't the ideal tool anyway.

- Oliver

Oliver Peters Post Production Services, LLC
Orlando, FL
http://www.oliverpeters.com


Return to posts index

andy patterson
Re: FCPX performance - codecs and external IO hardware
on Mar 9, 2017 at 4:57:21 pm

[Oliver Peters] "When you are not in FCPX, all 3 monitors are the macOS GUI. When you are in FCPX, the HDMI-connected display is the video output for client/preview. It's not simply the viewer blown up to fill the screen. No different than going through an AJA or BMD device. You can also enable the same function in Premiere."

I know what Premiere Pro can do. I was curious if FCPX was being used in the same way without taking a performance hit. I guess FCPX only takes a hit in performance with video capture cards. I am thinking Apple did not plan for 3rd party hardware when creating FCPX. I think they tried to add 3rd party support the best that they could.


Return to posts index

Erik Lindahl
Re: FCPX performance - codecs and external IO hardware
on Mar 9, 2017 at 7:00:37 am

Dual GPU monitors isn't a problem with FCPX. I think others have talked about having a third monitor vi for example the MP's HDMI-port.

I only think it's a matter of priority and optimization. Premiere was virtually useless when we started testing it back in I think CS 5.5. Performance died when using a video interface. Now that isn't a problem at all most of the time (I say this since I have tested 4K monitoring).


Return to posts index

Erik Lindahl
Re: FCPX performance - codecs and external IO hardware
on Mar 9, 2017 at 7:12:28 am

Looking at it from my point of view and a "top OS" point of view they should build the i/o suppport into Core Video / AV Foundation rather than just FCPX. This would potentially mean any app could have the same output features as an NLE.

In theory this should be virtual interfaces as well such as a h264 realtime encoder and having a virtual screen as your output (i.e. stream your AV Output to another device).


Return to posts index

andy patterson
Re: FCPX performance - codecs and external IO hardware
on Mar 9, 2017 at 5:05:24 pm

[Erik Lindahl] "I only think it's a matter of priority and optimization. Premiere was virtually useless when we started testing it back in I think CS 5.5. Performance died when using a video interface. Now that isn't a problem at all most of the time (I say this since I have tested 4K monitoring)."

I have Premiere Pro CS 5.5. Premiere Pro CS 5.5 has better real-time performance than Premiere Pro CC 2017 when using 3rd party hardware. You did have to use a 3rd party timelines from AJA, BMD or Matrox. Other than that it worked fine. Premiere Pro CS 6 implemented the Mercury Transmit feature that allowed just bout any timeline to make use of 3rd party hardware.


Return to posts index

<< PREVIOUS   •   VIEW ALL   •   PRINT   •   NEXT >>
© 2017 CreativeCOW.net All Rights Reserved
[TOP]