FCPX vs Premiere Render/Export speed tests by Alex4D

[Daniel Mcclintock]

I think speed tests are not completely reliable. There are many variables in such tests.

I ran the same test on my Spring 2009 Mac Pro with 16 GB ram and a 2.9 quad core processor. My graphics card is a nVidia 4000 with CUDA.

I don’t have FCPX so I can’t compare, but the results for Premiere Pro 6.0.2 were:

Preview: No time. The CUDA card previews in real time with Gaussian blur so there is no render.

Export: 1 minute through the Adobe Media Encoder with ProRes 422. The source video was H.264 1080p.

I think the reason FCPX did well on the tests is because of Open CL, which I don’t think Premiere supports yet. (I could be wrong on that, so don’t crucify me.)

———————–

“Sometimes Life Needs a Cmd-Z!”

[Craig Seeman]

[Daniel McClintock] “I think speed tests are not completely reliable.”

I think they’re reliable for the specific machine.

Also one should do comparable tests if you’re going to compare. Alex listed exactly what he did.
The source clip is a 1 minute 1080p25 ProRes 422(HQ) movie.

He noted differences in speed between 6.0 and 6.0.2 as well since he had tested with 6.0.

Regarding Premiere Pro and OpenCL support
https://blogs.adobe.com/premierepro/2012/05/opencl-and-premiere-pro-cs6.html

AMD Radeon HD 6750M and AMD Radeon HD 6770M graphics card with 1GB VRAM in MacBook Pro computers running Mac OSX v10.7 or later

(See the system requirements page.)

[UPDATE: The Premiere Pro CS6 (6.0.2) update added OpenCL and CUDA functionality for the GT 650M GPU, which is in the newer MacBook Pro computers.]

Alex used
This was on an 3.4 GHz i7 iMac with 16GB of RAM and 2MB Radeon 6970M.
His typo should be 2GB. I’m not sure that the 6970M is supported because Adobe somehow leaves that out.

[Bret Williams]

The 6970m is not supported. You can hack the preference file and add it and it appears to work, but in my experience it did nothing but hamper performance until ultimately crashing a few minutes later. Over and over. YMMV, but supported, no.

The new iMac are going to be wonderful for those of us that do anything Adobe on their systems. After Effects especially. Without CUDA, the CS6 upgrade to AE was nearly pointless. Ok, I’ve made good use of the new camera tracker for sure.

[John-michael Seng-wheeler]

This isn’t really a fair test at all. Any test comparing FCPX against Premiere without CUDA/Open CL isn’t really an direct comparison.

I would love to see this test done by someone with a new Mac that has a Premiere supported GPU in it.

[Craig Seeman]

[Bret Williams] “The new iMac are going to be wonderful for those of us that do anything Adobe on their systems.”

That will be when a “real” comparison test can be done. It will be a system that both NLEs should take advantage of.

What is clear is that Apple is moving to nVidia for all independent GPUs so one might expect (or hope) that it’ll be the standard in the MacPro replacements as well.

[Tim Kolb]

As Craig says, it’s a reliable test for the machine it was tested on. If you have that configuration, these are likely the results you’ll see.

It’s no secret that Mac users without Adobe-approved GPU hardware don’t have the performance of those who do…and if the GPU you have is supported by FCPX and not Adobe, then FCPX will run faster.

Since NVIDIA cards also support OpenCL, I’m not sure there would be a way to test the reverse scenario as FCPX would also be able to use any card that PPro would as far as I know.

TimK,
Director, Consultant
Kolb Productions,

Adobe Certified Instructor

[Jakub Vomacka]

i tried to replicate the test here on my hardware, going with CUDA for Premiere. i have hackintosh:
i7 920 – 3.8Ghz
6 GB RAM
Nvidia 560 Ti

test made on 10.8.2
FCPX 10.0.6
PPro CS6 v6.0.0
1 min of ProresHQ 1080p25

Gaussian Blur Export ProresHQ:
PPro – 0:49 (export through AME 0:41)
FCPX – 0:22

to try something else, apart from Prores Codec
Gaussian Blur Export H264 (960x540p25 – “better quality” settings in fcpx):
PPro – 1:00
FCPX – 1:06

One thing to note here can be the quality of encodes, with blur there is not much to visually compare.
Personally i think (tried many tests for various apps) that CUDA downscaling in PPro (especially for interlaced footage) cannot be matched in quality in any other NLE. Its pretty fast too! FCPX comes very blurry compared to PPro on this task. Therefore, CUDA gives “sometimes” access to better algorithms for certain tasks, so possibly a better quality.

PS: I tried activating OpenCL on my Nvidia card in Premiere and that gave me 4:50 on ProresHQ export. Might be different on AMD cards though, because Nvidia doesnt like OpenCL (still pretty new on their platform) as much as AMD.

[Walter Soyka]

[Daniel McClintock] “I think speed tests are not completely reliable.””

[Craig Seeman] “I think they’re reliable for the specific machine.”

I’d add one additional qualification: they’re reliable for the specific workflow on the specific machine. Premiere Pro has a performance disadvantage decoding and encoding ProRes media that doesn’t exist with other more open formats.

Tests like these are very useful, but they more meaningful (in my opinion) for deciding which hardware and formats work best with a given NLE than they are for deciding which NLE to use. They don’t touch on any of the workflow issues where the real efficiency gains for either FCPX or Pr would lie.

Walter Soyka
Principal & Designer at Keen Live
Motion Graphics, Widescreen Events, Presentation Design, and Consulting
RenderBreak Blog – What I’m thinking when my workstation’s thinking
Creative Cow Forum Host: Live & Stage Events