DV Info Net

DV Info Net (https://www.dvinfo.net/forum/)
-   Final Cut Pro X (https://www.dvinfo.net/forum/final-cut-pro-x/)
-   -   Mini DV NTSC Interlace to Final Cut pro 10 HD 1080i (https://www.dvinfo.net/forum/final-cut-pro-x/517196-mini-dv-ntsc-interlace-final-cut-pro-10-hd-1080i.html)

David Bastarache June 14th, 2013 11:31 AM

Mini DV NTSC Interlace to Final Cut pro 10 HD 1080i
 
I have hundreds of small video productions from a while ago that are on Mini DV NTSC standard def. When I import them into Vegas, no problem look great, but when I import them into Final Cut Pro 10, I get combing. I set the field order and it cleans it up a little, but it still looks terrible.
Has anyone got a good workflow? Yes I can use Vegas 12, but need to do stuff with Final Cut Pro 10.

William Hohauser June 14th, 2013 12:32 PM

Re: Mini DV NTSC Interlace to Final Cut pro 10 HD 1080i
 
How did you import them? By capturing from tape in FCPX or transferring previously captured files? Files captured in Vegas may not import into FCPX without some adjustments.

And why do you want them in 1080i now? Up-resing SD footage will never look great but it can be worked with. While I think FCPX does a very good job of up-resing footage, some people feel they get a better result by using Compressor.

And by "combing" do you mean aliasing on edges in the images?

Bill Davis June 15th, 2013 12:04 AM

Re: Mini DV NTSC Interlace to Final Cut pro 10 HD 1080i
 
Quote:

Originally Posted by David Bastarache (Post 1800295)
I have hundreds of small video productions from a while ago that are on Mini DV NTSC standard def. When I import them into Vegas, no problem look great, but when I import them into Final Cut Pro 10, I get combing. I set the field order and it cleans it up a little, but it still looks terrible.
Has anyone got a good workflow? Yes I can use Vegas 12, but need to do stuff with Final Cut Pro 10.

I suspect your problem is related to this.

X is resolution independent at it's core. It's happy to create project rasters of whatever you ask it to.

Your SD footage is a coarse raster with little temporal resoultion being interlaced at 720 X 480. So each field is actually 720 x 240 not a very dense raster.

One thing about X is that since it's built to handle any resolution up to 4k and beyond (provided your computer has enough horsepower to handle big files) it tries to do all it's transcoding and image math in high precision.

So if you feed it a coarse raster, and ask it to output a fine raster (what happens during uprezing) it's going to spend a LOT of time trying to map and move your big pixels onto the finer array.

Think of it like having oversized checkers and placing them on a checkerboard with standard size squares. No matter how you position them, you'll be blocking out more of the squares than the one to one relationship of properly sized checkers on the same board.

This is where morire and edge crawl and similar issues are born.

The software can alias everything - but that may lead to nasty picture softening as the edges are blurred to prevent alias issues.

In the end, there are two processes that can help.

The easiest is to just keep working in SD so that the old footage maps at 1 to 1.

The next is harder, and that's to figure out the raster differences between your SD source and the HD world you want to fit it into - and calculate how to set your X timeline to a raster that the original SD Footage scales into easily.

1440 x 960 would double both the horizontal and vertical rasters and map each SD pixel on to 4 HD pixels - which should be pretty clean. But you've got to live with that and if you need to re-size your SD issues, you're right back where you started. Plus you may not be able to deliver in 1440 x 960. So you might face downrezing compression if you need to output a standard file size like 1280 x 720 p.

The problem is that you have two different rasters and mapping one into the other can be a pain.

Hope that gives you some food for thought.

Good luck.

David Bastarache June 20th, 2013 01:21 PM

Re: Mini DV NTSC Interlace to Final Cut pro 10 HD 1080i
 
Quote:

Originally Posted by William Hohauser (Post 1800300)
How did you import them? By capturing from tape in FCPX or transferring previously captured files? Files captured in Vegas may not import into FCPX without some adjustments.

And why do you want them in 1080i now? Up-resing SD footage will never look great but it can be worked with. While I think FCPX does a very good job of up-resing footage, some people feel they get a better result by using Compressor.

And by "combing" do you mean aliasing on edges in the images?

Hi,
I have imported them both ways, direct from MiniDV tape into Final Cut Pro X and going through Vegas 12.
By combing I mean use can see the spaces between the lines of resolution, I guess one field or the other is dropped because I select either upper or lower in an attempt to clean up the video.

Dave

David Bastarache June 20th, 2013 01:25 PM

Re: Mini DV NTSC Interlace to Final Cut pro 10 HD 1080i
 
Quote:

Originally Posted by Bill Davis (Post 1800365)
I suspect your problem is related to this.

X is resolution independent at it's core. It's happy to create project rasters of whatever you ask it to.

Your SD footage is a coarse raster with little temporal resoultion being interlaced at 720 X 480. So each field is actually 720 x 240 not a very dense raster.

One thing about X is that since it's built to handle any resolution up to 4k and beyond (provided your computer has enough horsepower to handle big files) it tries to do all it's transcoding and image math in high precision.

So if you feed it a coarse raster, and ask it to output a fine raster (what happens during uprezing) it's going to spend a LOT of time trying to map and move your big pixels onto the finer array.

Think of it like having oversized checkers and placing them on a checkerboard with standard size squares. No matter how you position them, you'll be blocking out more of the squares than the one to one relationship of properly sized checkers on the same board.

This is where morire and edge crawl and similar issues are born.

The software can alias everything - but that may lead to nasty picture softening as the edges are blurred to prevent alias issues.

In the end, there are two processes that can help.

The easiest is to just keep working in SD so that the old footage maps at 1 to 1.

The next is harder, and that's to figure out the raster differences between your SD source and the HD world you want to fit it into - and calculate how to set your X timeline to a raster that the original SD Footage scales into easily.

1440 x 960 would double both the horizontal and vertical rasters and map each SD pixel on to 4 HD pixels - which should be pretty clean. But you've got to live with that and if you need to re-size your SD issues, you're right back where you started. Plus you may not be able to deliver in 1440 x 960. So you might face downrezing compression if you need to output a standard file size like 1280 x 720 p.

The problem is that you have two different rasters and mapping one into the other can be a pain.

Hope that gives you some food for thought.

Good luck.

I did use 1440 x 960, It seemed to be the least offensive to look at. I will go back and try an SD project and see if that is any better, but I may have tried that as well. I may have to stay with Vegas for the old SD video's.
Thanks for your help,

Dave


All times are GMT -6. The time now is 04:12 AM.

DV Info Net -- Real Names, Real People, Real Info!
1998-2024 The Digital Video Information Network