Larry Asbell
Forum Replies Created
-
Couldn’t you make an NTSC project in FCP, import the converted file, put it in a sequence, set the sequence TC to what you want, and export again?
-
Well for starters, you could check to see if I correctly fingered the cause. Re-import your export into FCP, put it in a sequence, set playback to Safe RT so it won’t drop frames and compare the original and the export. Watch them at play speed and advance them frame by frame. I believe you’ll find that the original has “finer slices” of the action and twice as many of them.
What do you do?
1) you could export your good looking original at at it’s original frame rate. Sure, it will double file size but you can lower the date rate by adjusting the amount of compression if you need to.
2) Less effective would be to shoot in 30P and make sure the shutter speed was long enough so each frame had a more blurred action rather than stop-action look. Somewhere in the neighborhood of 1/60th sec, i.e. 180 degree shutter. This will naturally blur fast action and allow the viewer’s eyes to see the action of each frame “meld” with the one before. I don’t really believe you will like the result anywhere near as well as option 1. I mostly suggest it to help illustrate the principles involved.
Remember both broadcast specs for HD sample action 60 times a second. 1080i is 30 frames per second, with two interlaced fields in each frame, i.e. 60 fields per second, and 720p is 60 frames per second. There’s a reason they chose those rates.
You may say, Yea but film has been shot at 24 fps since the invention of talkies. That’s correct but cinematographers keep the shutter speed slow enough to blur the action, avoiding the look you are getting.
– Larry Asbell
-
You are shooting at 60P. When you export at 29.97 your export is throwing away every other frame.
You can’t do this and expect the result to look the same.
-
Today I was capturing from a PAL HDV tape and could not get FCP to recognize my deck. My Sony HDV deck is normally used with NTSC tapes so it puts out a 60i signal when the tape is not playing. So I played the 50i tape while FCP was booting and the deck was recognized over Firewire.
Had this not worked, I would have changed the 60i/50i setting in the settings menu, but for one PAL tape this got me by.
– Larry Asbell
-
The fact that your source is 1080 at 30 frame/sec puts it just over the line for what a Quad G5 can capture in ProRes. A Quad G5 will capture ProRes at 1080 at 24 fps.
An alternative, if the amount you have to capture is not too much, would be to capture in Uncompressed 10-bit and transcode the files to ProRes. Your machine can handle that because the computation for compressing to ProRes can proceed in less than real time.
-
Rick nailed it exactly.
For years I’ve done moves on stills in both Avid and FCP. It’s a very common problem that images have too much hard edged detail to make clean moves in video. Whether from digital cameras, scans or computer graphics, edges that are too hard, especially horizontal lines, will jitter as they jump from pixel to pixel across the video raster. Don’t worry that you are throwing away sharpness. We are talking about detail that is beyond what will make it to the video anyhow.
I have found no formula for how much blur works but if the still dimensions are a fraction larger than the video you may find that less than one pixel is enough. If it’s three times the video dimension than a couple of pixels may be enough.
The detail in the still has to be softened before it hits the move effect. You cannot reduce the flicker at all if you try to apply a blur or any other fix after the move is done.
– Larry Asbell
-
Michael, I’m glad you asked this because I was recently banging my head against a similar problem:
How do you mask the area of an image that a filter is to apply to? (Think animating a glow across an object.) My testing and reading between the lines of the Motion 3 manual seems to say that you also have to duplicate the original layer also.
Boris and AE both allow you to mask a filter directly so I wasn’t expecting this.
In the Motion 3 manual your situation of needing to mask a key is discussed for several pages in the Motion 3 manual starting on page 1108.
But the general question of masking a filter, something you need to do for many common effects, is discussed only obliquely as far as I see. Not only for animating a glow across an object, but applying a blob of blur/mosaic to block a face, or color treating just an area of a picture. The section on Applying a Filter to a Mask on page 1000 is not the same as applying a mask to a filter. In other words, I don’t want to add a ripple edge to the creature’s cut-out eye, I want to apply my filter to the eye and see the rest of the face normally. This need and the fact that you must duplicate the original layer is not mentioned.
Once I got clear on the solution, the only efficiency I could bring to the method is:
Rather than duplicate the layer, clone it so it automatically keeps the clone updated as you modify the original layer.
-
[Kent Beeson] “If this is true, then why would anyone transcode ever again?”
1. To edit footage into a program of a different standard,
2. to distribute on tape,
3. to broadcast it,
4. to distribute a PAL program in NTSC countries where DVD players don’t play PAL.
-
When I have to do textless shots at the end I do Zero frame handles. Handles in this case will only confuse or slow down someone who has to drop those clips back in cut to cut. And unless you blow off an effect or a color correction there will be no video mismatch what so ever. When dissolves are involved the clip I extract is from the first frame after the dissolve in to first frame after the next dissolve. This is the best place to hide a discrepancy if there was one.
-
You know that CDs are mastered at a level that uses all available headroom. Think about what that means. It means there are no higher levels on the digital audio scale. There is no numerical value to represent anything hotter than that (insert reference to “This is Spinal Tap” here).
If Final Cut, Avid or any other audio mixing system worked at that level you simply COULD NOT mix the volume of anything further up!
You wouldn’t like that.
So, long ago designers of audio production systems set the working level inside these systems to be lower than that. The mix faders on both Final Cut and Avid allow you to turn any sound up 12dB. It may not be documented anywhere but isn’t it obvious that the capture process must drop all audio levels -12db or else there wouldn’t be anywhere UP to go? For some reason they didn’t design it to also drop the level of audio on import. And so you and everyone else has to.
P.S. I remember the very early versions of Avid Media Composer (about 1993) that could not mix audio any hotter, only lower.