[Alan Okey]

Use ProRes 422, not ProRes 422 HQ).

[Alan Okey]

The best investment for optimizing performance on an FCP workstation is to make sure you have a balanced system, i,e. don’t overspend in any one area to the detriment of another. Fast, reliable storage is of critical importance to FCP’s performance. And don’t skimp on equipment for proper monitoring.

Consider the following:

– Fast RAID 5 external storage from a reputable vendor (CalDigit, Dulce Systems, etc.). Storage speed is the single most critical factor influencing general system responsiveness in editing, and directly impacts RT and multicam performance more than any other factor.

– For the new Mac Pros, the ideal RAM configuration is for 3 (not 4 ) slots to be filled. Configure your RAM accordingly , i.e. 3x2GB or 3x4GB DIMMs.

– Rendering speed in FCP is CPU-bound. If your main concern is how fast FCP renders, then buy the fastest CPUs you can afford.

– Choice of graphics card is not critical for Final Cut Pro, but it is critical for Motion and Color. If you plan on using Motion and Color to any great extent, choose the ATI Radeon HD 4870, not the nVidia cards. Apple’s Pro Apps perform better on ATI cards due to driver optimization.

– Invest in a proper I/O card like the AJA Kona series, and use an external broadcast monitor to monitor your work, not a computer LCD. A proper broadcast monitor would be a Sony PVM or BVM series CRT or a broadcast-quality LCD from a company like Flanders Scientific.

[Alan Okey]

[Brent Stanley] “Recently I’ve started working with footage shot on a Sony EX3 camera (XDCAM EX 1080p30 35 Mb/s) With a FCP timeline set to 720×480 (NTSC 3:2), CCIR 601 Pixel ratio, and a DV/DVCPRO- NTSC Compressor, I then created a QT movie using the XDCAM HD 1080i60 settings (This was requested by the client) “

Let me get this straight:

The footage was shot on a Sony EX3 at 1080p/30. Your client is requesting an HD Quicktime file @ 1080i/60. And you’re editing in an SD 720×480 sequence? Why on earth would you do that? Why not simply use the XDCAM 1080p/30 Easy Setup?

[Brent Stanley] “I decided it would probably make more sense to edit in HD, so I started a new sequence and then brought one of the XDCAM clips in, and said ‘yes’ to the ‘would you like to match?’ prompt. So now I’ve got a timeline with the following sequence settings:”

Are you dragging content from the browser into the new XDCAM EX sequence, or from your existing SD sequence?

If you’re dragging in footage from an SD sequence, once you’ve dropped it into your new HD sequence, highlight the clip in the timeline, right-click on it, choose “remove attributes” and check “distort.” This should remove any distortion that was added when you dropped the HD source clips into an SD sequence.

Also, you need to change your Audio/Video settings so that your external video output is set to 1080i/60, otherwise you won’t be seeing the proper aspect ratio on your video preview monitor.

[Alan Okey]

[Craig Shields] “hen you say “not likely to happen in the near future” are you talking NAB or later than that? “

I have no clue. I’m a skeptic at heart, so I wouldn’t hold my breath for any NAB revelations. John Siracusa, the author of the article, might have more insight into your question than I do.

I’m not convinced that FCP is constrained by RAM usage at this point, however. Rendering is CPU-constrained much more than it is RAM-constrained. I think there’s more performance to be eked out of FCP through multithreading/multicore optimization than from maximum RAM usage. Technologies like Grand Central and OpenCL need to be tapped by FCP just as much as the increased RAM made available through 64-bit addressing.

There’s no magic 64-bit fairy dust that immediately makes everything run faster and better if an app is natively 64-bit. There are certain kinds of applications that use large data sets that can benefit from 64-bit memory addressing, but I’m not sure that FCP is one of them. Considering that video is frame-based, and even the very largest formats have frame sizes that fit well within the 4GB limit of a 32-bit app, I’m not sure how 64-bit memory addressing would speed up rendering beyond allowing for the rendering of multiple concurrent frames, something that would have to be supported at the thread level as well. But this is just conjecture on my part. I don’t have a degree in computer science, and I’m sure someone with the proper credentials could shed more light on the subject, if only in theory.

[Alan Okey]

[David Roth Weiss] “$1800 would get you three 24″ Dell monitors.”

David, I was referring to the Apple LED display, not the 30″. I was also referring to a cheaper Dell display, but the point still stands. For video editing, there’s little incentive to go with an Apple display. Prepress might be a different story.

[Alan Okey]

[William Carr] ” Is there any advantage at all to an Apple display? “

Some clients might think they look sexier.

They’re made of easily recycled materials (aluminum, glass).

They are mercury and arsenic free.

Apart from that? No, no real reason to buy an Apple display as a GUI monitor for editing unless you think they are attractive enough to pay the price premium.

It’s a pretty hard sell to convince me that a single Apple LED display with a 3′ cable, a single signal input and an exotic (read: annoyingly uncommon and inflexible) connector type is worth the price of two 23″ 1920×1080 Dell displays.

[Alan Okey]

“Would the FW400/HVRM25 chained to the back end of the drive effect performance?”

Quite possibly. Test playback performance without the deck attached.

[Alan Okey]

Try reducing the contrast level of the logo, i.e. dip the white level down to 85 IRE or so and raise the black level slightly. Areas of high contrast are particularly problematic. Also try a very slight Gaussian blur on the logo, perhaps .025 or so.

[Alan Okey]

[Greg Ball] “I’m not just dropping the wide screen into an 8 bit uncompressed timeline, I’m blowing it up to fill the frame. I don’t believe you can do that in compressor. “

Yes, you can. Compressor is an extremely versatile Swiss army knife of an application that allows you to transcode, compress, and perform advanced format conversions, including converting frame rates using optical flow technology for smoothly interpolated results.

I implore you to do some research by opening the Compressor user manual and reading about its capabilities. You can very easily transcode footage from HD to SD, and there are many options for adjusting aspect ratio, cropping, scaling, etc. Whether or not you choose to use it on this particular project, it is a valuable asset that can really expand your capabilities.

[Greg Ball] “BTW how long would it take to transcode 4 90-minute presentations?

That’s entirely dependent upon the speed of your computer and whether or not you’ve set up a virtual cluster with Qmaster. For detailed information on how to accomplish this, read my posts in this thread:

https://forums.creativecow.net/thread/8/1043558#1043564

[Greg Ball] I also have a LHe, so I can down convet easily.”

I have an LHe as well, and I still use Compressor extensively. The Kona is only helpful for downconversion on capture or output, not for internal transcoding. Its downconversion hardware isn’t accessed internally for format conversion unless you’re ingesting or outputting footage using the SDI or analog I/O jacks. If you merely want to transcode already-captured files for use in a project, it doesn’t really help.

[Alan Okey]

Be careful – FCP’s spill suppressor can product some nasty results, sometimes visible as a dark ring around keys. You’re better off rolling your own spill suppressor using the 3-way CC and limiting the effect.