J.R. Hudson
Spirit Animal
Re: DVX100a Squeeze Grabs
Tell your wife not to go into the backyard
Tell your wife not to go into the backyard
DVX100a Squeeze Grabs
- Thread starter donbarzini
- Start date
donbarzini
Well-known member
Re: DVX100a Squeeze Grabs
WTF? It's 5:00 AM here and I don't even want to go out there
WTF? It's 5:00 AM here and I don't even want to go out there
J.R. Hudson
Spirit Animal
Re: DVX100a Squeeze Grabs
He's going to get the Wedding Ring!
He's going to get the Wedding Ring!
Re: DVX100a Squeeze Grabs
..the correct square pixel output for 16:9 from an anamorphic (this includes "squeeze mode" images which are essentially crop and stretch electronic anamorphic images) DV frame is either 864x480 or 864x486- either one is correct for all practical purposes. the second is the mathamatically correct numbers which comes from a full D1 res frame which is 720x486. the DV spec drops these 6 extra pixels for compression reasons to make it 720x480. when you take the horizontal pixel count, and multiply it * 1.2 to account for the pixel aspect ratio being 1.2 you get 864 the vertical pixel count is unaffected. so you get 864x486 and 864x480.
864x480 is not mathematically correct, but what its doing is dropping the last 6 pixel lines in vertical resolution for compression reasons. .. some programs call for a 853x480 frame size to represent 16:9 in a square pixel aspect but IMHO its sort of incorrect, because the DV frame size should be based off of all the same calculations as a full D1 frame which is 720x486 and processed the same.. only dropping the 6 pixel lines in the vert resolution.. also 853 is not correct for multiplying 720 * the 1.2 pixel aspect ... you can only get 864.
youll notice that the 720x480 frame size is also not mathematically correct 4:3 (remember the horizontal pixels are at a .9 pixel aspect so you need to mutliply 720 * .9)..
but 720x486 IS correct 4:3 ..
so the DV spec is not mathematically correct. it just uses full D1 resolution specs for it calculations , and drops the 6 pixel lines from the vertical for compression..
anyway ..so if you want correct 16:9 DV spec in terms of sqaure pixels.. it is 864x480
and if you want correct 4:3 DV spec in terms of sqaure pixels.. it is 648x480
the square pixel sizes of 864x486 and 648x486 are mathematically correct but will stretch the DV image by 6 pixels vertically which is incorrect.. but you dont notice, and its fine to use for all practicasl reasons.. once you use the image in a video project in your NLE you will have to crop or squeeze back to 720 x 480... ..of course squeezing would be the best since you dont lose your image info and keep the 6 pixel stretched image. if you shot in full D1 of course you would use these figuers with no distortion since that is the actual frame size.
anyway
..the correct square pixel output for 16:9 from an anamorphic (this includes "squeeze mode" images which are essentially crop and stretch electronic anamorphic images) DV frame is either 864x480 or 864x486- either one is correct for all practical purposes. the second is the mathamatically correct numbers which comes from a full D1 res frame which is 720x486. the DV spec drops these 6 extra pixels for compression reasons to make it 720x480. when you take the horizontal pixel count, and multiply it * 1.2 to account for the pixel aspect ratio being 1.2 you get 864 the vertical pixel count is unaffected. so you get 864x486 and 864x480.
864x480 is not mathematically correct, but what its doing is dropping the last 6 pixel lines in vertical resolution for compression reasons. .. some programs call for a 853x480 frame size to represent 16:9 in a square pixel aspect but IMHO its sort of incorrect, because the DV frame size should be based off of all the same calculations as a full D1 frame which is 720x486 and processed the same.. only dropping the 6 pixel lines in the vert resolution.. also 853 is not correct for multiplying 720 * the 1.2 pixel aspect ... you can only get 864.
youll notice that the 720x480 frame size is also not mathematically correct 4:3 (remember the horizontal pixels are at a .9 pixel aspect so you need to mutliply 720 * .9)..
but 720x486 IS correct 4:3 ..
so the DV spec is not mathematically correct. it just uses full D1 resolution specs for it calculations , and drops the 6 pixel lines from the vertical for compression..
anyway ..so if you want correct 16:9 DV spec in terms of sqaure pixels.. it is 864x480
and if you want correct 4:3 DV spec in terms of sqaure pixels.. it is 648x480
the square pixel sizes of 864x486 and 648x486 are mathematically correct but will stretch the DV image by 6 pixels vertically which is incorrect.. but you dont notice, and its fine to use for all practicasl reasons.. once you use the image in a video project in your NLE you will have to crop or squeeze back to 720 x 480... ..of course squeezing would be the best since you dont lose your image info and keep the 6 pixel stretched image. if you shot in full D1 of course you would use these figuers with no distortion since that is the actual frame size.
anyway
donbarzini
Well-known member
Re: DVX100a Squeeze Grabs
iamloser,
Thank you for the information, very informative and useful.
iamloser,
Thank you for the information, very informative and useful.
CaptainMench
Carbonite Member
Re: DVX100a Squeeze Grabs
Still don't get it... why would you squeeze out the width? Isn't THAT constant? Wouldn't you want to drop the heigth?
CaptM
Still don't get it... why would you squeeze out the width? Isn't THAT constant? Wouldn't you want to drop the heigth?
CaptM
Re: DVX100a Squeeze Grabs
well, without being too technical
..its widescreen.. the height dosent change. the width does.. its wider. you dont get a shorter image at the same width.. you get a wider image at the same height.
the pixels are 1.2 ratio meaing they are 1.2 times as wide as a square pixel ,, when regular dv is only .9 which is less wide than a square pixel. the pixel aspect ration does not change the height.. only width of the pixels.
when you put on an anamorphic lense it sees a wider horizontal field of view..but same height. and squeezes the image onto the less wide ccd. then you display those pixels at 1.2 width ratio to widen the squeezed image back out to its native wideness 16:9
again.. squeeze mode is fake anamorphic.. so you do the same thing.
if you shoot real naitive 16:9 DV then its full 16:9 height, and full 16:9 width to begin with from the CCD. and is squeezed into a 4:3 image to match DV spec of 720 x 480 and usually flagged as 1.2 pixel aspect so that the NLE will widen it out again.
hope this helps :-/
well, without being too technical
..its widescreen.. the height dosent change. the width does.. its wider. you dont get a shorter image at the same width.. you get a wider image at the same height.
the pixels are 1.2 ratio meaing they are 1.2 times as wide as a square pixel ,, when regular dv is only .9 which is less wide than a square pixel. the pixel aspect ration does not change the height.. only width of the pixels.
when you put on an anamorphic lense it sees a wider horizontal field of view..but same height. and squeezes the image onto the less wide ccd. then you display those pixels at 1.2 width ratio to widen the squeezed image back out to its native wideness 16:9
again.. squeeze mode is fake anamorphic.. so you do the same thing.
if you shoot real naitive 16:9 DV then its full 16:9 height, and full 16:9 width to begin with from the CCD. and is squeezed into a 4:3 image to match DV spec of 720 x 480 and usually flagged as 1.2 pixel aspect so that the NLE will widen it out again.
hope this helps :-/