moon1234
5th December 2011, 11:07
I have been reading for hours and I am getting quite confused with x264 and SAR.
I have several hundred analog Hi-8 tapes that were captured with an ADVC-1394 DV card. When the tapes are captured there is an 8-10 pixel black pad on the right and left of the actual video image. There is also approximately an 8 pixel black pad to to top of the image and 8 pixels of tracking and other analog garbage on the bottom of the image.
The captured resolution is 720x480. Most of the material is 4x3 with some analog 16x9 material (I tried to be forward thinking in the 90's).
Now I would like to crop 8 pixels all the way around to yield a nice clean center image with no crap. Most of this bound for youtube and other sites so people can watch in other countries (family).
I know I can resize to square pixel for most things, but I would really like to do this anamorphic and leave the cropped res alone and not resize if I can help it.
The problem is that I really do not understand how SAR works with x264. It would seem that if I change any of the crop dimensions in either the vertical or horizontal planes that I need to recalculate SAR.
For 4x3 material this usually means I crop to 704x464. Now I know 704 is considered the "normal" image area and the extra 16 pixels are just pad. So when I am looking at 4x3 DAR the 704 would be fine for the horizontal axis. 480 would be the normal resolution for the vertical axis to maintain the proper 4x3 DAR.
Now that I have cropped the vertical axis to 464 I no longer have a proper 4x3 aspect ratio.
Maybe someone can help me understand this a little better? This has to easier than I am making it out to be.
I have several hundred analog Hi-8 tapes that were captured with an ADVC-1394 DV card. When the tapes are captured there is an 8-10 pixel black pad on the right and left of the actual video image. There is also approximately an 8 pixel black pad to to top of the image and 8 pixels of tracking and other analog garbage on the bottom of the image.
The captured resolution is 720x480. Most of the material is 4x3 with some analog 16x9 material (I tried to be forward thinking in the 90's).
Now I would like to crop 8 pixels all the way around to yield a nice clean center image with no crap. Most of this bound for youtube and other sites so people can watch in other countries (family).
I know I can resize to square pixel for most things, but I would really like to do this anamorphic and leave the cropped res alone and not resize if I can help it.
The problem is that I really do not understand how SAR works with x264. It would seem that if I change any of the crop dimensions in either the vertical or horizontal planes that I need to recalculate SAR.
For 4x3 material this usually means I crop to 704x464. Now I know 704 is considered the "normal" image area and the extra 16 pixels are just pad. So when I am looking at 4x3 DAR the 704 would be fine for the horizontal axis. 480 would be the normal resolution for the vertical axis to maintain the proper 4x3 DAR.
Now that I have cropped the vertical axis to 464 I no longer have a proper 4x3 aspect ratio.
Maybe someone can help me understand this a little better? This has to easier than I am making it out to be.