Fast 4K Editing on OSX and Windows 10

[jcs]

A while ago I noted that FCP X was faster editing 4K material on a 2014 MBP and GT 750M GPU vs. Premiere Pro CC 2015 (latest) on a 12 Core MacPro and GTX 980ti. Today I tried FCP X on the MacPro in FCP X with C300 II 4K files. 4K editing was blazing fast- scrubbing instantaneously. This is full 4K resolution, native files (dual 4K displays). Curious to find the limit, I began stacking 4K clips and was able to get 6 4K clips (cropped so all 6 clips had visible elements) before there was visible slowdown (still usable!). 5 4K clips were fully real-time.

On the same hardware & OSX, Premiere Pro CC 2015 can play one 4K clip in real-time. On the same hardware and Windows 10, playback resolution must be set to 1/2 to get near real-time performance.

Resolve 12 (latest) wasn't able to play the C300 II 4K file in real-time on the MacPro in OSX.

For fast 4K editing, FCP X is currently the champ.

[gethin]

whats the data rate of those files?  (I'm not a premiere fan, I remember version #1 and thought it was a mess - and still think it is, but I'm invested in it now).  

100Mbit gh4 4k plays fine on my 5 year old windows (8) box in PP, whilst all 6 cores are in use rendering!  Can't believe a single 4k file wouldn't play back on mac pro hardware - only if there were drivers not installed, or some other hardware or OS snafu.

BTW I think this kind of test would be incredibly useful, I'd love to see more of them. Very curious how other software stacks up on the same hardware in different OSs.

 

 

 

 

 

 

[jcs]

It's possible there's a Windows 10 issue, though I have tried all the latest drivers from NVidia since October. On the same hardware in OSX, PPro is only slightly faster (can play 1 4K C300 II clip at full playback resolution vs. 1/2). FCPX can play 5 4K clips in real-time at the same time.

Testing on a 2014 MBP Retina 2.6GHz GTX 750M 16GB: PPro barely plays C300 II 4K in real-time at 1/2 resolution, FCPX plays the same clip in real-time, with instant scrubbing at full quality.

It seems like more of a PPro design issue vs. OS or drivers.

[joema]

It seems like more of a PPro design issue vs. OS or drivers.

I agree. I have tested Premiere CC 2015 on Windows 10 and Mac OS X 10.11.2, and timeline responsiveness is slow on both with H264 4K. Running on identical hardware (a top-spec 2015 iMac 27) FCP X is much faster than PP CC at various things. This difference is much less noticeable at 1080p. In my experience PP CC timeline performance falls off a cliff at 4K. By contrast FCP X slows down some but it is still usable and generally responsive. That is with CC on 1/4 resolution and FCPX viewer on "performance".

By "various things" I mean the most common timeline editing operations that convey an impression of responsiveness. IOW how rapidly it responds to JKL input. How much lag to keyboard input when changing directions from fast forward to reverse. How many frames per sec the Program Monitor updates when fast forwarding through the timeline in both forward and reverse. PP is almost a slide show relative to FCPX, especially on multicam 4K.

OTOH I have talked to other experienced filmmakers who claim they have no problems with PP CC performance on H264 4k -- on similar iMac hardware. Part of this may be editing style. If you never use JKL to rapidly survey material but just drag the playhead and wait a second, you can get work done. PP seems to eventually cache a buffer around the current playhead location so that smaller trim or jog movements within that buffer are pretty quick. But larger timeline movements exhaust that buffer and become very slow and laggy.

Also if you never compare FCPX and CC back-to-back on the same hardware with the same material, you get accustomed to the editor's responsiveness and that becomes "the new normal". You can adapt to almost anything.

However CPU utilization is much higher in CC than FCPX when moving at the same speed through the timeline on the same hardware. This indicates the CC playback code path is far less efficient. Encode/decode of Long-GOP codecs is inherently CPU bound unless algorithm-specific specific hardware acceleration is used. A general purpose GPU cannot do this, maybe FCPX is using Quick Sync. OTOH others observe FCPX is still very fast on Mac Pros which use Xeon that does not have Quick Sync. That implies a code path optimization.

With FCPX I never used proxy files on HD but I often use those on 4K and always on multicam 4K. It is built-in, seamless and easy. With CC there is no built-in proxy feature and ironically it needs it much more than FCPX due to the slower performance on H264 4K.

PP CC really needs a lot more optimization on the playback engine. This is ironic since traditionally "Mercury Playback" has been so fast, but it is not handling 4K well.

 

[Xavier Plagaro Mussard]

Imho, Adobe keeps adding features to their programs, without really building a new version from scratch. I remember a lot of years ago, I was on a PC and Premiere gave me an error in a window that was from two versions of Windows before!!!

I now use a 1920x1200 external monitor and I am thinking about getting a 4K one, but I have one doubt, will my setup be faster or slower??

[gethin]

I agree about the 4K performance and cliff: editing anything 4K on my system becomes a nightmare.  Sure files play fine, but as soon as you add a grade or transition you are constantly having to render in order to just to look at the footage play without skipping or stuttering.   I'm not convinced (and the comments here seem to support) that any hardware upgrades wouldn't do much to help.  

I think back to the happy times I spent mucking about with media 100 on the mac and discreet logic effect on pc.  Such fine bits of software, why didn't autodesk look to have a broad user base model for its business?

[jcs]

When writing plugins for PP you see just how ancient their software is. Apple wrote FCPX from scratch (perhaps with some bits from iMovie ;)), whereas PP's architecture dates back to the time of FCP with the recent Mercury engine tacked on. It bought them some time with 1080p, however it shows its limitations with 4K.

While FCPX has limitations and gotchas different from PP, basic editing and grading of 4K footage is fast and fluid which helps the creative process.

A 4K monitor should not slow a system down with a decent GPU rendering 1080p. Rendering or monitoring 4K is another story- depends on the software.

[Cassius McGowan]

Well I have Mac and PC Systems and I really don't have any problems editing on either systems. FCPX or PP CC.

[jcs]

Well I have Mac and PC Systems and I really don't have any problems editing on either systems. FCPX or PP CC.

Cool- can you tell us more details regarding PP CC- version, OS versions, video card, and driver versions? This is for 4K sequences, at full display resolution (vs. 1/2 or 1080p sequences).

[tellure]

I recently upgraded my PC for 4K editing with Premiere Pro CC 2015 and was dismayed to find that although it's usable, it's not what I would call "fast" or responsive.

I have an i7 5820K overclocked to 4Ghz, GTX970 graphics card, Samsung SSD, 16GB of DDR4 RAM (more coming, had to send back some bad RAM sticks, although the project I'm editing is very small and doesn't even use 5GB).  I have to keep the playback window at 1/4 resolution to avoid dropping frames and the lag when positioning the play head at a new location and pressing play is pretty long (about 2/3rds of a second before the clip starts playing).  Scrubbing is extremely choppy.. probably around 5-8fps if the timeline hasn't been rendered.

For more info check out Don Kotlos's thread "A story about 4K XAVC-S, Premiere and transcoding" where he tries various transcoding options to speed up editing in Premiere.  Some of the thoughts there were that FCP's use of Intel's Quick Sync processor may indeed be playing a big role in its speed.  Whatever the solution, it's an abysmal state of affairs that the flagship editing app on the PC is so far behind the flagship app on the Mac in performance.

[jcs]

Changing Premiere Pro CC 2015's Renderer to Mercury - OpenCL (from CUDA) resulted in real-time playback of 4K material at full resolution! This is in OSX El Capitan, GTX980, latest drivers. Simultaneous playback on a 2nd 4K display didn't slow down.

Adding a third display (1080p HDTV) continued to run without slowdown.

[joema]

Changing Premiere Pro CC 2015's Renderer to Mercury - OpenCL (from CUDA) resulted in real-time playback of 4K material at full resolution! This is in OSX El Capitan, GTX980, latest drivers. Simultaneous playback on a 2nd 4K display didn't slow down.

I just re-tested PP 2015.1 (9.1.0) on a 2015 top-spec iMac 27. I verified Renderer was set to Mercury Playback - GPU Acceleration (OpenCL). It was sluggish on UHD 4K H264 material from a Sony A7RII and Panasonic AG-DVX200. This is all at 1/4 playback resolution. By "sluggish" I mean frame rate during FF and REW, and responsiveness to JKL commands. It *could* play two-camera 4K multicam at 1/4 res with no dropped frames. Relative to FCPX viewer update rate on the same hardware and same playhead speed, it's about 20x slower -- about 1/2 Hz update vs 10 Hz.

I then tried it on a 4K H265 clip from an NX1 -- program monitor frame rate was quite fast, despite the compute-intensive nature. It felt like editing 1080P material. That doesn't make sense, unless the hardware assist was more effective or somehow not working properly on H264 4K.

I also tested the same H264 4K material on PP 2015.1 on a top-spec 2013 iMac 27 having a GTX-780m, using both CUDA and OpenCL. That is using the latest 7.5.22 CUDA drivers. It was similarly sluggish on both OpenCL and CUDA (by above definition), although it would play at normal speed at 1/4 res with no dropped frames.

I've also tested the same H264 4K material on PP 2015.1 on a 4Ghz Windows PC with a GTX-660. It was sluggish there also, so it's not a Windows vs OS X thing.

Part of this is definitional based on your editing style and what response characteristics define "fast" vs "slow". I am accustomed to the lighting-fast viewer update rate in FCPX, and before that PP on 1080P material. If using JKL to scrub through the timeline, PP CC 2015 feels sluggish on H264 4K on any platform I have tested, whether CUDA or OpenCL. It's not unusable but forces a different editing style of dragging the playhead and being patient. Overall the responsiveness of FCPX on unrendered H264 4K is about like PP CC 2015 on a fully-rendered 1080p timeline.

It is highly suspicious that PP is more responsive on 4K H265 than H264. That's with both CUDA and OpenCL. IMO there is some inefficiency in the code path of the H264 renderer which degrades drastically on 4K.

[Don Kotlos]

It is highly suspicious that PP is more responsive on 4K H265 than H264. That's with both CUDA and OpenCL. IMO there is some inefficiency in the code path of the H264 renderer which degrades drastically on 4K.

I do feel the same. From what I read H264 4K from GH4 does not have the same issue (can anyone verify this?), so it might be related to Sony's XAVC-S implementation.

[Sekhar]

No problems here on Windows 10 with Premiere Pro (CC 2015.1). Smooth playback at full resolution with or without effects, usually SpeedGrade (except stuff like Neat Video, obviously) with H.264 or H.265. Hardware is an older i7-2600 3.4 GHz, 32 GB main memory, GTX 980, and SSD storage.

Do you see any contention on the performance meters? Storage (if you have a slow hard disk) might be bogging you down.

[jpfilmz]

Changing Premiere Pro CC 2015's Renderer to Mercury - OpenCL (from CUDA) resulted in real-time playback of 4K material at full resolution! This is in OSX El Capitan, GTX980, latest drivers. Simultaneous playback on a 2nd 4K display didn't slow down.

Adding a third display (1080p HDTV) continued to run without slowdown.

Are you using an external thunderbolt2/USB3 RAID storage setup or is it all internal ssd?

[Don Kotlos]

No problems here on Windows 10 with Premiere Pro (CC 2015.1). Smooth playback at full resolution with or without effects, usually SpeedGrade (except stuff like Neat Video, obviously) with H.264 or H.265. Hardware is an older i7-2600 3.4 GHz, 32 GB main memory, GTX 980, and SSD storage.

Do you see any contention on the performance meters? Storage (if you have a slow hard disk) might be bogging you down.

I also don't have a problem with playback. Its the scrubbing (or placing the playhead at different timepoints) that is slowing me down a lot. 

Is it with a7rii 4K files? 

[joema]

Are you using an external thunderbolt2/USB3 RAID storage setup or is it all internal ssd?

This is using an 8TB Promise Pegasus R4 Thunderbolt RAID-5 array. However the I/O system wouldn't make much difference since the data rate for H264 is pretty low. It is CPU and GPU limited, not I/O limited.

[Sekhar]

I also don't have a problem with playback. Its the scrubbing (or placing the playhead at different timepoints) that is slowing me down a lot. 

Is it with a7rii 4K files? 

Well, the OP mentioned slowdown with playback, which is why I mentioned that. WRT scrubbing, I guess it depends on your expectation. I usually use J-K-L editing, so never had any issues going back/forth. But if you're talking about scrubbing rapidly/wildly on the timeline to go to a random point, a compressed format like H.264/H.265 is never going to be as smooth as say ProRes (because it needs to do more than work on just that frame). However, even that seems acceptable on my setup, but then it's hard to compare unless we can compare side/side. Bottom line, I don't see it as an issue on my system.

[Don Kotlos]

Well, the OP mentioned slowdown with playback, which is why I mentioned that. WRT scrubbing, I guess it depends on your expectation. I usually use J-K-L editing, so never had any issues going back/forth. But if you're talking about scrubbing rapidly/wildly on the timeline to go to a random point, a compressed format like H.264/H.265 is never going to be as smooth as say ProRes (because it needs to do more than work on just that frame). However, even that seems acceptable on my setup, but then it's hard to compare unless we can compare side/side. Bottom line, I don't see it as an issue on my system.

Is this with A7rii 4K files? 

[joema]

I do feel the same. From what I read H264 4K from GH4 does not have the same issue (can anyone verify this?), so it might be related to Sony's XAVC-S implementation.

I just re-tested H264 4K from a A7RII, Panasonic AG-DVX200 and GH4 on PP 2015.1 on a top-spec 2015 iMac 27 with media on a Thunderbolt array. I don't see much if any difference regarding performance. I previously thought maybe the GH4 material was a little quicker, but upon retesting it feels about the same.

Testing additional H265 4K NX1 material, there is no question PP CC 2015.1 is much more responsive on it than on H264 4K from the other cameras. That doesn't make sense, considering H265 is much more compute-intensive to decode.

Re scrubbing with JKL, there are two factors: (1) The program monitor update rate when scrubbing, and (2) The responsiveness of JKL.

On 1080P material it's all pretty fast. You can FF, hit stop, back up -- it responds quickly. Program monitor frame rate is fast, whether in FF or dragging the playhead. On H264 4K, when fast forwarding, the program monitor update rate can drop to about once per second or slower, and JKL lag to stop and reverse increases to several seconds. It almost feels like the keyboard is broken.

This doesn't make it unusable, you just have to adopt different methods. On H265 4K from the NX1, it is a lot faster, though not equal to 1080P. Using FCPX, 4K is about as fast as 1080P on Premiere.