FORUMS: list search recent posts

2013 MacPro 6 core VS 2012 3.2 MacPro

COW Forums : Apple Final Cut Pro X Debates

<< PREVIOUS   •   VIEW ALL   •   PRINT   •   NEXT >>
Dave Jenkins
2013 MacPro 6 core VS 2012 3.2 MacPro
on Jan 10, 2014 at 8:06:12 pm

Here's some test I've done with my new 2013 MacPro stock 6 core VS 2012 3.2 MacPro.

Exported a 5 minute timeline unrendered with Canon XF C300 footage, DVCPRO HD 23.98, H264, ProRes 1080 422 & 720 59.94

2013 3.5GHz 6-Core stock - Pegasus R6 T2 raid 800MB's
Time = 1:26

2012 3.2 quad core with 32gig of ram - 4 hard drive internal raid 400MB's
Time = 2:47

Could be the HD speed difference.

To Render a 5 minute timeline:
2013 = 1:58

2012 = 2:30

Dajen Productions, Santa Barbara, CA
MacPro 3.2GHz Quad Core - AJA Kona LHe+
FCS 3 OS X 10.7.4
FCP X, Logic Pro, Squeeze, Filemaker 10.8.3


Return to posts index

Steve Connor
Re: 2013 MacPro 6 core VS 2012 3.2 MacPro
on Jan 10, 2014 at 8:46:42 pm

Thanks for posting that, how is FCPX is day to day operation, is it noticeably faster?

Steve Connor

There's nothing we can't argue about on the FCPX COW Forum


Return to posts index

Michael Sanders
Re: 2013 MacPro 6 core VS 2012 3.2 MacPro
on Jan 10, 2014 at 8:53:28 pm

Personally I'm not bothered about short things like a quick render, I want to know how the big tasks - the ones where you want big iron, like exporting a 30 min show or compressing to H264 take.

But then see my mini review above..

Michael Sanders
London Based DP/Editor


Return to posts index


John Heagy
Re: 2013 MacPro 6 core VS 2012 3.2 MacPro
on Jan 10, 2014 at 9:23:06 pm

[Michael Sanders] "I want to know how the big tasks - the ones where you want big iron, like exporting a 30 min show or compressing to H264 take"


This is why I'm lamenting the loss of QT reference movie export. In FCP7 I can export a 1hr ref.mov in less than 1 min and let Episode start encoding. With FCPX one has to wait for a self contained .mov to finish.

John


Return to posts index

Bret Williams
Re: 2013 MacPro 6 core VS 2012 3.2 MacPro
on Jan 11, 2014 at 4:25:08 am

Well, not if it wasn't rendered you couldn't. And the audio always had to be recompressed. It didn't reference audio. A 1hr video would had taken a little bit just for the audio alone.


Return to posts index

John Heagy
Re: 2013 MacPro 6 core VS 2012 3.2 MacPro
on Jan 11, 2014 at 3:28:50 pm

[Bret Williams] "Well, not if it wasn't rendered you couldn't. And the audio always had to be recompressed. It didn't reference audio. A 1hr video would had taken a little bit just for the audio alone."

Correct… In our case all media is converted to ProRes so no rendering except transitions, titles, etc. Yes it does flatten the audio, no re-compression though. Even with nothing rendered a ref movie takes less than 5min to create typically. There's no way around render time. One pays that price whether ref or self contained.

Apple wants to kill ref.mov because of security concerns. Basically the treat of a ref mov pointing to suspect files. Not sure how a FCP xml or any project file from any media app which points to source files is any safer.

John


Return to posts index


Bret Williams
Re: 2013 MacPro 6 core VS 2012 3.2 MacPro
on Jan 12, 2014 at 3:55:31 am

Ref was nice, but since it was referencing render files, that disappeared easily if changes were made, that I found if I used reference files that apps like AE or DVDSP were often trying to search for render files that were gone. So I quit using them a long time ago. Too much trouble and drive space was too cheap.

Now we don't have to spend the time transcoding to ProRes, so there's the time saving on the front end.


Return to posts index

John Heagy
Re: 2013 MacPro 6 core VS 2012 3.2 MacPro
on Jan 12, 2014 at 8:11:56 pm

[Bret Williams] "I used reference files that apps like AE or DVDSP were often trying to search for render files that were gone."

True enough.. FCP ref.mov are not good long term. We use them in conjuction with our Episode engine to offload flattening or encoding, after which they are deleted. Besides saving the editor time it allows quick re-dos.

[Bret Williams] "Now we don't have to spend the time transcoding to ProRes, so there's the time saving on the front end."

True again… but we find saving time on the back end far more valuable. It takes time for all the media to accumulate, time we use to encode to ProRes and tag with embedded metadata. As a project proceeds, workflow pipelines start sprouting invoking other apps, and the final timeline gets more complex as the deadline nears. It's all this later work we strive to lubricate so there are fewer issues and renders as deadline pressure and project complexity increases.

John


Return to posts index

Jeremy Garchow
Re: 2013 MacPro 6 core VS 2012 3.2 MacPro
on Jan 13, 2014 at 5:27:08 pm

[John Heagy] "True again… but we find saving time on the back end far more valuable. It takes time for all the media to accumulate, time we use to encode to ProRes and tag with embedded metadata. As a project proceeds, workflow pipelines start sprouting invoking other apps, and the final timeline gets more complex as the deadline nears. It's all this later work we strive to lubricate so there are fewer issues and renders as deadline pressure and project complexity increases."

Here's hoping that developers will plug in to FCPX so that ref files won't be as necessary?

I miss ref files, too.

"Export integration with third-party apps

Developers can plug directly into the Share interface within Final Cut Pro, allowing for seamless export to third-party systems such as Sienna, Vizrt, Primestream FORK, axle, and Cantemo Portal."


Return to posts index


Keith Koby
Re: 2013 MacPro 6 core VS 2012 3.2 MacPro
on Jan 13, 2014 at 7:25:25 pm

[Jeremy Garchow] "Here's hoping that developers will plug in to FCPX so that ref files won't be as necessary?

I miss ref files, too.

"Export integration with third-party apps

Developers can plug directly into the Share interface within Final Cut Pro, allowing for seamless export to third-party systems such as Sienna, Vizrt, Primestream FORK, axle, and Cantemo Portal.""


I think that means exporting the file from fcpx and having a MAM be aware of it instantly. What John would want would be to export an xml and have episode engine read it like an edl and use it as a source.

Does the fcpxml point to render files? I'm under the impression that it does not. Otherwise, you could in theory convert fcpxml to edl and feed edl to episode engine.

I'm curious to see what they (telestream) have in store for nab.


Return to posts index

Jeremy Garchow
Re: 2013 MacPro 6 core VS 2012 3.2 MacPro
on Jan 13, 2014 at 7:37:31 pm

[Keith Koby] "What John would want would be to export an xml and have episode engine read it like an edl and use it as a source. "

That's what I would want, too.

Compressor can do it without render files so I was hoping telestream (or anyone) can do it. Making multiple encodes of one file is a much slower process with FCPX as you have to wait for the master to export first.


Return to posts index

John Heagy
Re: 2013 MacPro 6 core VS 2012 3.2 MacPro
on Jan 13, 2014 at 10:47:09 pm

I like discussing what John wants ;)

So no, fcpxmls do not point to render files so that won't do unless the option of including them was added so it did link to current render files.

Some sort of aaf variant might do, but the glacial progress of aaf.xml says that isn't likely.

There is the mxf reference format OP1b, but I don't know if that supports other wrappers besides .mxf

I understand Apple wants to kill QT ref due to security concerns. Apparently Apple is concerned that the referenced files could point to a suspect file and be opened. What I don't understand is how files listed in a fcpmxl, or really any project file from any other app, would not have the same issue?

John


Return to posts index


Dave Jenkins
Re: 2013 MacPro 6 core VS 2012 3.2 MacPro
on Jan 10, 2014 at 9:32:30 pm
Last Edited By Dave Jenkins on Jan 10, 2014 at 9:38:13 pm

9 min video to H264 for Blu-ray with Compressor:

2013 = 6 minutes
2012 = 8 minutes

Dajen Productions, Santa Barbara, CA
MacPro 3.2GHz Quad Core - AJA Kona LHe+
FCS 3 OS X 10.7.4
FCP X, Logic Pro, Squeeze, Filemaker 10.8.3


Return to posts index

Keith Koby
Re: 2013 MacPro 6 core VS 2012 3.2 MacPro
on Jan 10, 2014 at 10:49:32 pm

You wouldn't happen to have access to an i5 or i7 mac book pro or iMac for comparative purposes?


Return to posts index

Dave Jenkins
Re: 2013 MacPro 6 core VS 2012 3.2 MacPro
on Jan 10, 2014 at 11:21:56 pm

Yeah, I'll hook up the promise R6 raid to my laptop i7 2.3 and run the export.

Dajen Productions, Santa Barbara, CA
MacPro 3.2GHz Quad Core - AJA Kona LHe+
FCS 3 OS X 10.7.4
FCP X, Logic Pro, Squeeze, Filemaker 10.8.3


Return to posts index


Dave Jenkins
Re: 2013 MacPro 6 core VS 2012 2.3 GHz i7 Macbook Pro
on Jan 10, 2014 at 11:53:20 pm
Last Edited By Dave Jenkins on Jan 10, 2014 at 11:55:41 pm

Export 5 minutes Timeline Macbook Pro:
1:52
Export 5 minutes Timeline MacPro
1:26

Render 5 minutes Timeline:
Macbook Pro:
1:32
Mac Pro:
1:58

I ran this test twice on both machines. Same hard drive and the Macbook Pro had a faster render.
The MacPro has the BlackMagic Mini Monitor with th 9.93 drivers. Maybe that make a difference.

Promise R6 raid Speeds:
TB 1 = 699 write 549 read
TB 2 = 830 write 700 read

Dajen Productions, Santa Barbara, CA
MacPro 3.2GHz Quad Core - AJA Kona LHe+
FCS 3 OS X 10.7.4
FCP X, Logic Pro, Squeeze, Filemaker 10.8.3


Return to posts index

Keith Koby
Re: 2013 MacPro 6 core VS 2012 2.3 GHz i7 Macbook Pro
on Jan 11, 2014 at 12:10:05 am

Yeah. This is where I expect the results to get interesting. We need to see the 8 core d700 and the 2 GB vram iMac. But the performance is pretty good on the laptop.

Render vs export. I always preferred turning off background rendering and trying to work realtime as much as possible. This is what I tell my editors to attempt to do. I'm really excited to see how the d700 does. However, there are definitely times where third party plugins won't play back in realtime (some of the edge stuff for example) and you need to render it to see it.

For a bigger facility that turns out lots of material in fcpx, the time savings over a year or two makes the mac pro worth it. The retina mac book pro is a great choice for some based on that performance. The imac with the nvidia card is yet another great choice if you need Adobe performance as well.

I haven't got to test an imac on mavericks and 10.1, but from a GUI responsiveness standpoint, the 6 core mac pro we have was much, much better than the ML iMac. This was on a complex project. I can't wait to get an imac up to mavericks and 10.1 to see just how much of that is OS and software vs the hardware.


Return to posts index

Dave Jenkins
Re: 2013 MacPro 6 core VS 2012 2.3 GHz i7 Macbook Pro
on Jan 11, 2014 at 12:27:08 am

However the playback of the unrendered footage on the MacBook Pro was very bad compared to the Mac Pro which played perfectly unrendered with the better performance quality setting.

Dajen Productions, Santa Barbara, CA
MacPro 3.2GHz Quad Core - AJA Kona LHe+
FCS 3 OS X 10.7.4
FCP X, Logic Pro, Squeeze, Filemaker 10.8.3


Return to posts index


Erik Lindahl
Re: 2013 MacPro 6 core VS 2012 2.3 GHz i7 Macbook Pro
on Jan 11, 2014 at 3:37:42 pm

Would be more valid with a compression test from a proper encoder like Episode Pro or even Adobe Media Encoder. Apples Compressor is quite sub-par (or has been before).


Return to posts index

Gary Huff
Re: 2013 MacPro 6 core VS 2012 2.3 GHz i7 Macbook Pro
on Jan 12, 2014 at 5:04:26 pm

[Erik Lindahl] " Would be more valid with a compression test from a proper encoder like Episode Pro or even Adobe Media Encoder. Apples Compressor is quite sub-par (or has been before)."

Agreed. Compressor isn't the best way to check this.


Return to posts index

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