color recction class: what is your eye most sensitive too?

[Sithxace]

for hw, i have 2 questions:

1. what is your eye most sensitive to in relation to film/video? (

(My answer I think is light, however I dont know how that can be applied to the next question.)

2. How does that affect the way film/video is processed?

 

[PDR]

for hw, i have 2 questions:

1. what is your eye most sensitive to in relation to film/video? (

(My answer I think is light, however I dont know how that can be applied to the next question.)

2. How does that affect the way film/video is processed?

Our eyes are much more sensitive to luminance than chroma. Here are some articles and background reading:

http://provideocoalition.com/index.php/cmg_keyframes/story/luma_processing/

http://broadcastengineering.com/hdtv/avch-encoding/index.html


This affects the way video is processed, because final distribution formats (e.g. blu-ray, dvd, flash for web) are 8-bit , 4:2:0 Y'CbCr to conserve bandwidth (instead of higher color sampling like 4:4:4, or bit depth like 10-bit (bits per channel) video), because you can get away with using lower bandwidth for encoding the color information.

If you look on page 3 or 4 in that 2nd article, there are PSNR comparisons for 4:2:2 vs. 4:2:0 (not that PSNR is that great of a measurement... but I digress)

However, you should strive to use higher sampling and bit depth for intermediate calculations and color correction (you get better results, fewer errors accrue), even if your sources were only 8-bit and 4:2:0 to begin with, and the higher values were interpolated

 

[Amr Rahmy]

Our eyes are much more sensitive to luminance than chroma. Here are some articles and background reading:

http://provideocoalition.com/index.php/cmg_keyframes/story/luma_processing/

http://broadcastengineering.com/hdtv/avch-encoding/index.html


This affects the way video is processed, because final distribution formats (e.g. blu-ray, dvd, flash for web) are 8-bit , 4:2:0 Y'CbCr to conserve bandwidth (instead of higher color sampling like 4:4:4, or bit depth like 10-bit (bits per channel) video), because you can get away with using lower bandwidth for encoding the color information.

If you look on page 3 or 4 in that 2nd article, there are PSNR comparisons for 4:2:2 vs. 4:2:0 (not that PSNR is that great of a measurement... but I digress)

However, you should strive to use higher sampling and bit depth for intermediate calculations and color correction (you get better results, fewer errors accrue), even if your sources were only 8-bit and 4:2:0 to begin with, and the higher values were interpolated

for your class, use the previous answer.

for your own knowledge, the most sensitive thing to video is contrast(maybe contrast is not the word am looking for) changes between frames, it's not something available or similar to looking at one frame, or still picture.

if part or all the image goes from light to dark or from color to color, or the camera movement from slow to black to suddenly cut to fast moving camera and object.