Lu Nelson
Forum Replies Created
-
DV footage looks soft in many monitoring situations because of how FCP and Quicktime display it. Open your files in Quicktime Player and hit Command-J for options. Click on the video track and go to the Visual Settings tab, you will see there is an option for ‘High Quality’ display. The look should change substantially.
NB: This won’t help playback in FCP but it will allow you to verify that your footage is indeed sharp. You need a proper output card and to see this stuff on a video monitor while using FCP, or at least on a TV if you want to know what you’re editing.
-
Go in to your Qmaster settings (assuming you are running Final Cut Studio 2) in your System Preferences and I believe there are further options for the Compressor services that let you run multiple instances of the service. On a quad core machine the optimal number is 3. On an eight core it is 5. This will improve your times quite a bit as it will process your job in parallel segments; but yes it is slow.
However, I don’t think it’s too ridiculous when you consider it’s free as part of FCS2 and the comparable hardware-based realtime alternative is *much* more expensive.
-
Actually you should try compressor. it will take a long time but on a new computer it might work for you, and actually it does not look like crap it looks pretty good.
-
Hi,
Sorry, if you’re using Motion 2 this is difficult to do. There is I believe a “Basic 3D” filter which can mimic this to some extent but if you want to fly through multiple layers you need them to appear appropriately in relation to each other as well and you will encounter a lot of frustration in an app that doesn’t really support 3D space.
Ramping, as you have noticed, doesn’t do the trick because objects need to scale exponentially to look like they are ‘flying’ toward you., and ramps are linear.
However in Motion 3 you can ramp the Z position and you get precisely the right effect. Better yet, you simply do exactly what you described, which is position the layers in 3D space and fly the camera through them.
Check out ripple training’s Motion 3 Fast Forward DVD — they have a lesson which shows precisely a camera fly-through of layers in 3D Space
LMN
-
Yeah I’m not sure about that — I tried importing a PDF shape from illustrator and moving it in 3D space and I got some stair stepping.
Motion 3’s own shape objects DO of course rasterize perfectly at any scale in 2D and 3D…
I heard however that you can paste Illustrator shapes directly in to Motion. I believe the guy at motionsmarts.com has a note about this on his downloads page.
-
You may also find that thought the X1900XT has good specs, it is limited to graphics of 2048×2048 while Nvidia cards can support graphics of 4096×4096. At least this is the way it used to be.
My machine came with the X1900XT so I can’t compare, just say that it’s pretty good.
LMN
-
Hi,
the “Preparing Video for Display” is mainly a RAM thing. When you say graphics you mean Still Images in the timeline right?
If so, I can tell you it remains a problem even on my new MacPro Quad 2.66 with 8GB of RAM. With still images, if you start loading a lot of them in to the timeline it slows you down. Final Cut’s architecture isn’t really meant for it. Actually I haven’t heard of any editing prog that really does this well.
Working in an HD timeline however, on a G5 with only 2GB of RAM I’m surprised you’re getting much work done. Short answer: yes more RAM will help and you should have at least 4GB if you are editing in HD. Try that first.
-
Don’t shoot in 1/30 mode. Probably you will just lose resolution, if you are shooting on a camera that isn’t mean to do 30p properly (if you were of course, you don’t need deinterlacing in the first place); but MAINLY, you will find that all deinterlacers on the market are designed to deal with 60i or 50i footage, interlaced fields, and to try to create an appropriate motion blur as well, from that information. Shooting at 1/30s shutter will screw this up and your deinterlacing won’t work properly.
However, that being said if you are using Final Cut Studio 1 or 2 you have excellent deinterlacing tools in Compressor as well, you could batch process your footage through there, even convert it to 24p. Results will look better than Nattress’ filter (sorry Graeme) and for that matter better than ReVision or Boris, because it’s based on the Shake optical flow code; but if you max out the quality settings it takes a very long time to process.
Lu Nelson
Berlin, Germany -
I believe its possible using the sw from spherico.de — you export a sequence xml and their product pulls text from that (marker names, probably also clip names) and then creates a sequence full of text generators which you can paste in to your original sequence as an overlay. It’s meant for the kind of thing you’re talking about; but I don’t know more than that, haven’t tried it
Lu Nelson
Berlin, Germany -
This config should be OK as long as you distribute each set of four as 2 above, 2 below. I would however move your 2GB sticks in to the 1 and 2 slots on each riser (where the 512s will be when you get your Octocore), and let the 512s be the second set.
This is because the 512MB units are typically single lane and 1GB and above are typically dual lane, therefore they’ll run a bit slower than your new ones from Crucial so by putting the crucial ones first most of your memory usage will operate within the first 8GB or so and you’ll be running at top speed. In my case, I sold the 512s that came with my Quadcore 2.66 and replaced all slots with identical 1GB chips.
However, Apple doesn’t really acknowledge this or else they wouldn’t ship an Octo with 512MB modules. There have been articles on anandtech and on diglloyd.com about the fact that even a full roster of current DDR2 modules for the MacPro (of which ony CAS-5 speeds are on the market) is not really enough throughput to properly support the 8-core machine…but that’s really splitting hairs. And BTW 8GB should be plenty of RAM until we all start running Leopard
Lu Nelson
Berlin, Germany