[Rob Ainscough]

Shane,

If there are no handles (aka media) I have no choice, I can’t go back and get it re-shot with enough “handles” so that my transition is covered … just not going to happen.

If you have some other way to make it work, please let me know. On very short transitions a freeze is rarely that obvious, but like I said if you have some method other than asking for a longer clip, then please do fill me in.

My suggestion was to make freeze or move back in clip or just show warning as a option in Preferences … pick what you prefer to happen, I try NOT to speak for everyone as neither you or I really know what “everyone” does. So give them options to meet how they like to flow.

Rob.

[Rob Ainscough]

Yeah, it looks like I’ll have to make a “Freeze Frame” in order to provide enough “handles” for the transition. I keep forgetting that if my Clip In/Out points are at the very end of the source clip FCP will toss out this error.

I really hate the way FCP 7 “handles” this … no pun intended. Rather than toss out the error it should just either generate the freeze frames needed for the transition or move back/forward in the clips as needed and do the transition. But just tossing out an error and moving the transition to the start of the clip is never EVER something an editor would want. Or at the very least make this a Preference options somewhere. Grrrr.

Thanks for the help.

Rob

[Rob Ainscough]

Rafael,

I don’t disagree it can help fix problems on a Mac … it has helped out your G5 and I’m not saying DW is a bad product.

What I am saying and have tested with and without DW use, is that DW will NOT speed up FCP7 sequence renders — the ONLY thing that seems to improve FCP7 sequence renders is a faster CPU (just one CPU since that’s all FCP7 will use during a sequence render).

FCPX is a different product, it’s listed as being able to use all available CPUs and even GPUs to help render sequences faster. Jerry said FCPX is 2X faster renders than FCP7 — this has me puzzled for Jerry’s 8 core setup … I would have expected 6-7X faster, not 2X.

But either way, DW will not speed up FCP7 sequence renders.

Cheers, Rob.

[Rob Ainscough]

Rafael,

Dial it back, no need for that language.

Maybe you are thinking of a different product, but this is the product I’ve used called “DiskWarrior” by Alsoft … Link Here As you can see, it’s primary purpose is to repair disk problems.

Perhaps if you could provide some real world tests of FCP7 sequence render times before and after you used “DW” and not some generic information that “it’s faster”. Note: we’re talking about FCP7 sequence render times, not Export times, not Compressor times.

Rob

[Rob Ainscough]

Testing I’ve done with DiskWorrior has show zero benefit in terms of FCP7 performance. It’s a handy tool to recover files but it doesn’t do anything for FCP7??

Depends on the SATA controller, but typically you want to point the scratch disk to some other fast SATA drive, but in your case if you only have one SSD then I’m assuming it’s the faster of your drives so I’d point the scratch disk to your SSD even if it’s your main boot HD.

But test it out for yourself, easy enough to change the path of the scratch files. But I think you’ll find it doesn’t make much difference in terms of rendering in FCP7.

But just so ya know, the 3.4Ghz Intel i7 processors (in the iMac) do video encoding considerably faster than 2.93 Ghz MacPro Xeon processors. Things apple probably don’t want you to know, 4 core i7 iMac will render a lot faster than a 12 core MacPro (in FCP sequence). If you run compressor and setup Qmaster then the 12 core will be faster, but in FCP the i7 processor is considerably faster in almost ALL cases (especially sequence rendering).

What you don’t get with i7 is ECC memory … but with current RAM technology, ECC is becoming less and less of an issue … in fact pretty much just non-existant problem these days with DDR3.

Now toss in two Thunderbolt ports in the iMac and you have a very competent FCP base and lots of room to expand.

My hunch is the MacPro line will die and evolve into an iMacPro … just makes sense IMHO.

Rob

[Rob Ainscough]

Render times don’t seem to be tied to HD performance in FCP7. I’ve noticed the same, SSD, software RAID, standard HD all produce (within 2%) the same render time.

Applications will load a little faster, but nothing that significant. But as far as render times, no difference regardless of where I set the scratch disk. Maybe I haven’t done a render larger enough to exceed 2GB limit of FCP7 32bit address space?

Cinema4D can render pretty fast on my same MacPro and utilizes 4 cores, FCP7 sequence renders only use 1 core. Qmaster only works with compressor which doesn’t help FCP7 sequences.

[Rob Ainscough]

Twice as fast is still too slow for me … at least compared to Adobe on a Windows 7 box anyway. Does FCPX support any 3rd party accelerators when it comes to rendering on the timeline?

I’ll certainly keep and eye on FCPX and what Apple do with their MacPro line … I like how FCP 7 flows, sure there are some features I’d like that are missing … good example is being able to change parameters on a video generator that is used in several clips by selecting those clips and open in editor so that changing an Opacity value will be reflected on all the selected clips not just what is determine to be the “current” one.

Also seems like my ATI 5770 is being under utilized … it has a lot of processing power sitting idle. Does FCPX support nVidia CUDA or ATI Stream?

I’m not trying to sound Pro Adobe/Windows either, it’s just that similar projects render considerably faster and in some cases I don’t have to render. But I suppose on the plus side I can surf the web while I wait 😉

[Rob Ainscough]

Hi Jerry,

Thanks for response. I have a Matrox HD H.264 card installed and a Black Magic Intensity card installed. If I select the Matrox H.264 as the codec and try to render it just flashes an error and returns me to the timeline … the Matrox H.264 seems to only work on Exporting of sequence.

However, if I select the Black Magic 10bit RGB for codec I get 22 second sequence down to 8:05, about a 25% improvement. However output is not as good but can live with it and then switch to final codec once the sequence is complete.

Oh yeah, nothing to fancy, Boris Vector Shape and Boris Title 3D, some 3D world animation done in Cinema 4D imported as QT and some basic transitions.

Rob

[Rob Ainscough]

Thanks for the posts all.

Ran some testing: when from ProRes 422 HQ to ProRes 422 LT and (RGB and 10bit YUV) here are my results (timed on with stopwatch)

22 second sequence (min:secs)
ProRes 422 LT = 13:12.7
ProRes 422 LT RGB = 13:09.5
ProRes 422 HQ = 13.24.2
ProRes 422 HQ RGB = 13.19.6

5 second sequence
ProRes 422 LT = 6:04.9
ProRes 422 LT RGB = 6:03.1
ProRes 422 HQ = 6:08.3
ProRes 422 HQ RGB = 6:06.2

Test sequence has 24 video tracks. It might be that I need to move this sequence over to Motion and do it from there … but I’m not sure Motion will do it any faster and it adds a layer in my workflow.

As far as RAID — I ran the same 5 second sequence tests onto SSD, software RAID HD’s, and a single standard HD and the difference across them was 1-2% variance … so obviously the rendering isn’t HD dependent, seems to be purely CPU dependent.

Future of the MacPro line is definitely “in doubt” and hence why I’m trying to push thru with what I have right now, but it’s painfully slow (especially when compared to Adobe on a my Windows 7 box running on a i7 4 core processor). Can’t really “wait until spring” for a “possible” new MacPro lineup from Apple or lack there of a MacPro lineup — I know how Apple operate when it comes to providing future information … someone needs to give the Apple execs a hint that their paranoia is NOT what made them return to success 😉

Rob

[Rob Ainscough]

Excellent, exactly what I was looking for, saved me $300 🙂

Thank you sir.

Rob