Uncompressed Field Order at DVinfo.net

Ben Gurvich · December 17th, 2006, 01:06 AM

Having been editing with hdv i would like to go uncompressed for color correcting in after effects.

If i was going to deinterlace in Final Cut Pro or After Effects, which field or would i use for 720x576 uncompressed 8 bit 4:2:2 25fps
Or would i use no fields?

I did a quick test from final cut striaght to compressor and the results were very pleasing (much better than when i had shot with an fx1). I just needed to de-interlace my footage.

Cheers,
Ben Gurvich

Graham Hickling · December 17th, 2006, 08:32 PM

You are starting with 1080i60, right?

If so, you would specify upper field first for the source footage, and no fields for the deinterlaced output.

Edit: Oops - Melbourne - 1080i50 perhaps?

Ben Gurvich · December 18th, 2006, 03:42 AM

Correct Graham,

1080i50! Why can we all just shoot 50, half that is 25 which is closer to 24 anway. But then 60p is a lot more than 50p for overcrankn.

Ive never fuly understood outputting to no fields. Does it mean that you drop one of the fields or does it mean that it will jsut show in frames, and on a video device it willonly show the first field in that frame repeating?

Graham Hickling · December 18th, 2006, 12:11 PM

Tell me about it - I'm from New Zealand originally!

I don't really know the data structure of avi and mov files, etc, but my assumption of the way it works is that there are 'chucks' of data that represent an image, together with an instruction for how fast those images are played.

So 720p25 HDV1 footage has chucks of data representing a 1280x720 image, and these chunks are displayed at a rate of 25 per second.

Whereas 1080i50 HDV2 has chunks of data that are 960x1440, and these chucks are played at a rate of 50 per second (with the extra twist that alternating chucks are displaced 1-scanline up or down from the previous one.

If the above is correct (?) then ... to answer your question ... I assume that when 720p25 is played on a device expecting fields (or when you specify 'separate fields' in a app like Virtualdub) then the device/application goes to the 1280x720 'chunk', extracts half of it, plays that for 1/50th, then grabs the remaining half and displays that for the next 1/50.

So the fields don't really exist per se in the raw data of a progressive file, rather they are created by the process of interpreting that file? Perhaps?

December 17th, 2006, 01:06 AM	#1
Ben Gurvich Major Player Join Date: Dec 2003 Location: Australia Posts: 329	Uncompressed Field Order Having been editing with hdv i would like to go uncompressed for color correcting in after effects. If i was going to deinterlace in Final Cut Pro or After Effects, which field or would i use for 720x576 uncompressed 8 bit 4:2:2 25fps Or would i use no fields? I did a quick test from final cut striaght to compressor and the results were very pleasing (much better than when i had shot with an fx1). I just needed to de-interlace my footage. Cheers, Ben Gurvich

December 17th, 2006, 08:32 PM	#2
Graham Hickling Trustee Join Date: May 2004 Location: Knoxville, Tennessee Posts: 1,669	You are starting with 1080i60, right? If so, you would specify upper field first for the source footage, and no fields for the deinterlaced output. Edit: Oops - Melbourne - 1080i50 perhaps?

December 18th, 2006, 03:42 AM	#3
Ben Gurvich Major Player Join Date: Dec 2003 Location: Australia Posts: 329	Correct Graham, 1080i50! Why can we all just shoot 50, half that is 25 which is closer to 24 anway. But then 60p is a lot more than 50p for overcrankn. Ive never fuly understood outputting to no fields. Does it mean that you drop one of the fields or does it mean that it will jsut show in frames, and on a video device it willonly show the first field in that frame repeating?

December 18th, 2006, 12:11 PM	#4
Graham Hickling Trustee Join Date: May 2004 Location: Knoxville, Tennessee Posts: 1,669	Tell me about it - I'm from New Zealand originally! I don't really know the data structure of avi and mov files, etc, but my assumption of the way it works is that there are 'chucks' of data that represent an image, together with an instruction for how fast those images are played. So 720p25 HDV1 footage has chucks of data representing a 1280x720 image, and these chunks are displayed at a rate of 25 per second. Whereas 1080i50 HDV2 has chunks of data that are 960x1440, and these chucks are played at a rate of 50 per second (with the extra twist that alternating chucks are displaced 1-scanline up or down from the previous one. If the above is correct (?) then ... to answer your question ... I assume that when 720p25 is played on a device expecting fields (or when you specify 'separate fields' in a app like Virtualdub) then the device/application goes to the 1280x720 'chunk', extracts half of it, plays that for 1/50th, then grabs the remaining half and displays that for the next 1/50. So the fields don't really exist per se in the raw data of a progressive file, rather they are created by the process of interpreting that file? Perhaps?